**Towards Bridging Worldviews in Biodiversity Conservation: Exploring the Tsonga Concept of**  *Ntumbuloko* **in South Africa**

Brandon P. Anthony1, Sylvia Abonyi2, Petra Terblanche3 and Alan Watt1 *1Department of Environmental Sciences & Policy, Central European University 2Department of Community Health & Epidemiology, University of Saskatchewan 3Department of Sports, Arts, and Culture, Limpopo Province 1Hungary 2Canada 3South Africa* 

### **1. Introduction**

Many scholars and managers now question the traditional top-down, and often hegemonic, approaches of excluding local participation and ignoring local interests in management of biodiversity both within and outside formal protected areas (Johnston, 1995; Kiss, 1990). Greater participatory planning is believed to enhance local support for biodiversity conservation goals and decrease conflicts between local people and conservation authorities (Happold, 1995; Heinen, 1996; Manfredo et al., 2004).

Efforts worldwide to integrate biodiversity conservation and rural development objectives have had mixed results, indicating that synergies between the two are not inherent, and they are not a panacea (Alpert, 1996; Barrett et al., 2005; Hughes & Flintan, 2001; Newmark & Hough, 2000). We argue here that they must more fully incorporate local worldviews in their design and implementation if they ever hope to succeed. For institutions responsible for conservation, detailed knowledge of the people whose lives are affected by conservation policies can be as important as information about the biodiversity to be conserved (Anthony & Bellinger, 2007; Brechin et al., 2002; Veech, 2003). Moreover, it has been noted that in addition to playing a key role in human-environment interactions (Nietschmann, 1992; Smith, 2001), cultural elements of nature protection can be a resource providing insight into development of conservation plans (Kuriyan, 2002; Stevens, 1997) while also reinforcing community identity and, promoting community cohesion and adaptability (Goodland, 1991; Kleymeyer, 1992; Robinson & Redford, 1994). Thus, recognition and understanding of different local cultural systems permits a broader, more appropriate overall policy toward natural resource use (Maffi, 2004).

Towards Bridging Worldviews in Biodiversity Conservation:

(Ntsebeza & Hendricks, 1998).

**1.2.1 Tsonga** 

(1UpInfo, 1996).

Africa (1UpInfo, 1996).

**1.2 The people of Limpopo Province** 

Exploring the Tsonga Concept of *Ntumbuloko* in South Africa 5

centers of authority, their actual rights and responsibilities are not clearly spelled out

Communities in the study area comprise almost exclusively (96.2 – 99.1%) people from the Tsonga people group (Statistics South Africa 2003). Tsonga are a diverse population, and in the mid-1990s numbered about 1.5 million in South Africa, and at least 4.5 million in southern Mozambique and Zimbabwe (1UpInfo, 1996). In the 18th century, ancestors of the Tsonga lived in small, independent chiefdoms. Most Tsonga relied on fishing for subsistence, although goats, chickens, and crop cultivation were also important. Because their coastal lowland habitat was tsetse-fly infested, cattle were rare in their economies

During the *mfecane*1 and subsequent turmoil of the 19th century the history of the Tsonga was dominated by invasions of Zulu conquerors who left Chaka and enslaved the Ama-Thonga of the coast (Junod, 1912). Many Tsonga emigrated inland to the Transvaal from 1835 to 1840. Some successfully maintained their independence from the Zulu, while others were conquered by Zulu warriors even after they had fled. One Zulu military leader, Soshangane, established his authority over a large Tsonga population in the northern Transvaal in the mid-19th century (1UpInfo, 1996). The descendants of some of the

Tsonga who migrated inland brought new sources of food into the Transvaal, including cassava, certain kinds of groundnuts, potatoes and sorghum. Particularly important were the maize and fowls introduced in their new settlement areas. Agricultural work was performed almost exclusively by women, except for initial land clearing which was the men's responsibility (Magubane, 1998). Even today, labour division along gender lines still exists: men are traditionally hunters, herdsmen, fishermen, housing constructors, as well as traders; women are agriculturalists, gatherers, and collect water and fuelwood (Ombe, 2003). Crop harvesting was usually cooperative, done on a rotational basis, with area

By the early 1920s, the Tsonga-speakers constituted about 4% of the total South African population (Magubane, 1998). In the north, large chiefdoms, including Xikunda, Mhinga, Xigalo, and Makuleke occupied distinct reserves adjacent to the KNP. The Tsonga-Shangaan homeland, Gazankulu, was carved out of northern Transvaal Province during the 1960s and granted self-governing status in 1973. In the 1980s, the government of Gazankulu established a legislative assembly made up mostly of traditional chiefs. The chiefs opposed homeland independence but favored a federal arrangement with South

Communities were torn apart as families were moved to the Tsonga homeland, and the resulting taxation and overpopulation made people increasingly dependent on migrant labour. This caused men to leave their families for long periods, and today even women in rural areas seek seasonal work on nearby citrus farms (Mathebula, pers. comm.).

1 'The Crushing' - a series of Zulu and other Nguni wars and forced migrations in the early 19th century that changed the demographic, social, and political configuration of southern and central Africa.

conquered populations are known as the Shangaan, or Tsonga-Shangaan.

communities gathering to harvest each family's crop in turn.

We report here on research that included a focus on conflicts and synergies between local Tsonga people and conservation authorities in and around the boundaries of Kruger National Park (KNP) in Limpopo Province, South Africa. We begin below with a brief description of official conservation in Limpopo Province, followed by an introduction to *ntumboloko*, the Tsonga worldview that shapes the perspectives on the environment and conservation that was the focus of the research we report on here. We subsequently outline our theoretical approach, and our methods. Our results present perspectives on conservation and its context in the *ntumbuloko* worldview of the Tsonga. In our discussion we identify potential conservation synergies and conflicts between the two worldviews: that of 'western science', which predominantly influences official conservation management and practices in Limpopo Province, and that of Tsonga communities bordering the KNP. We conclude with some thoughts about the importance for conservation agencies to philosophically and practically understand and integrate local worldviews into their biodiversity conservation and socio-economic objectives.

#### **1.1 Official conservation in Limpopo Province**

#### **1.1.1 Within protected areas: Kruger National Park**

The Kruger National Park (KNP), situated in the Republic of South Africa (Figure 1), is approximately 350 km from north to south and covers nearly 2 million ha (Mabunda et al., 2003). Established in 1926, KNP is home to an unparalleled diversity of wildlife and maintained by a very sophisticated management system (Braack, 2000). Internationally, KNP functions as a major tourism destination with up to one million visitors annually, and serves as an important socioeconomic and ecological component of the Great Limpopo Transfrontier Park. Traditionally, KNP management has strictly followed 'western' scientific principles, often dismissing other forms of knowledge systems (see Wolpert, 1993), with the aim of single species management and manipulating ecosystems to meet prescribed goals (Carruthers, 1995). Following the dynamic economic and political transformations within South Africa since 1994, South African National Parks, including KNP, has witnessed a transformation in its policies which seek to integrate conservation and socioeconomic objectives of neighboring communities, including community resource use policies and the establishment of community fora (Mabunda et al., 2003).

#### **1.1.2 Outside protected areas**

Outside protected areas in Limpopo Province, environmental management is primarily the responsibility of the Department of Finance and Economic Development – Environmental Affairs (DFED) (Limpopo Provincial Government, 2005). DFED is operationally sub-divided into municipal districts which provide conservation extension services, control damagecausing animals outside KNP, and monitor and regulate the use of natural resources. DFED activities are largely governed by the *Limpopo Environmental Management Act* (LEMA) No. 7 of 2003*,* which is analogous with national legislation*.* DFED is also mandated to promote sustainable development by creating partnerships with communities, NGOs, the private sector, and other government departments. The DFED was created by the post-Apartheid regime as a part of new provisional governmental structures intended to introduce greater democracy to South Africa. Until that time, traditional authorities (TAs), based on chieftanship, were the effective representatives of communities. The legal competences of the TAs are ambiguous because, while they are recognized in the constitution as legitimate centers of authority, their actual rights and responsibilities are not clearly spelled out (Ntsebeza & Hendricks, 1998).

## **1.2 The people of Limpopo Province**

### **1.2.1 Tsonga**

4 Research in Biodiversity – Models and Applications

We report here on research that included a focus on conflicts and synergies between local Tsonga people and conservation authorities in and around the boundaries of Kruger National Park (KNP) in Limpopo Province, South Africa. We begin below with a brief description of official conservation in Limpopo Province, followed by an introduction to *ntumboloko*, the Tsonga worldview that shapes the perspectives on the environment and conservation that was the focus of the research we report on here. We subsequently outline our theoretical approach, and our methods. Our results present perspectives on conservation and its context in the *ntumbuloko* worldview of the Tsonga. In our discussion we identify potential conservation synergies and conflicts between the two worldviews: that of 'western science', which predominantly influences official conservation management and practices in Limpopo Province, and that of Tsonga communities bordering the KNP. We conclude with some thoughts about the importance for conservation agencies to philosophically and practically understand and integrate local worldviews into their

The Kruger National Park (KNP), situated in the Republic of South Africa (Figure 1), is approximately 350 km from north to south and covers nearly 2 million ha (Mabunda et al., 2003). Established in 1926, KNP is home to an unparalleled diversity of wildlife and maintained by a very sophisticated management system (Braack, 2000). Internationally, KNP functions as a major tourism destination with up to one million visitors annually, and serves as an important socioeconomic and ecological component of the Great Limpopo Transfrontier Park. Traditionally, KNP management has strictly followed 'western' scientific principles, often dismissing other forms of knowledge systems (see Wolpert, 1993), with the aim of single species management and manipulating ecosystems to meet prescribed goals (Carruthers, 1995). Following the dynamic economic and political transformations within South Africa since 1994, South African National Parks, including KNP, has witnessed a transformation in its policies which seek to integrate conservation and socioeconomic objectives of neighboring communities, including community resource use policies and the

Outside protected areas in Limpopo Province, environmental management is primarily the responsibility of the Department of Finance and Economic Development – Environmental Affairs (DFED) (Limpopo Provincial Government, 2005). DFED is operationally sub-divided into municipal districts which provide conservation extension services, control damagecausing animals outside KNP, and monitor and regulate the use of natural resources. DFED activities are largely governed by the *Limpopo Environmental Management Act* (LEMA) No. 7 of 2003*,* which is analogous with national legislation*.* DFED is also mandated to promote sustainable development by creating partnerships with communities, NGOs, the private sector, and other government departments. The DFED was created by the post-Apartheid regime as a part of new provisional governmental structures intended to introduce greater democracy to South Africa. Until that time, traditional authorities (TAs), based on chieftanship, were the effective representatives of communities. The legal competences of the TAs are ambiguous because, while they are recognized in the constitution as legitimate

biodiversity conservation and socio-economic objectives.

establishment of community fora (Mabunda et al., 2003).

**1.1.2 Outside protected areas** 

**1.1 Official conservation in Limpopo Province 1.1.1 Within protected areas: Kruger National Park**  Communities in the study area comprise almost exclusively (96.2 – 99.1%) people from the Tsonga people group (Statistics South Africa 2003). Tsonga are a diverse population, and in the mid-1990s numbered about 1.5 million in South Africa, and at least 4.5 million in southern Mozambique and Zimbabwe (1UpInfo, 1996). In the 18th century, ancestors of the Tsonga lived in small, independent chiefdoms. Most Tsonga relied on fishing for subsistence, although goats, chickens, and crop cultivation were also important. Because their coastal lowland habitat was tsetse-fly infested, cattle were rare in their economies (1UpInfo, 1996).

During the *mfecane*1 and subsequent turmoil of the 19th century the history of the Tsonga was dominated by invasions of Zulu conquerors who left Chaka and enslaved the Ama-Thonga of the coast (Junod, 1912). Many Tsonga emigrated inland to the Transvaal from 1835 to 1840. Some successfully maintained their independence from the Zulu, while others were conquered by Zulu warriors even after they had fled. One Zulu military leader, Soshangane, established his authority over a large Tsonga population in the northern Transvaal in the mid-19th century (1UpInfo, 1996). The descendants of some of the conquered populations are known as the Shangaan, or Tsonga-Shangaan.

Tsonga who migrated inland brought new sources of food into the Transvaal, including cassava, certain kinds of groundnuts, potatoes and sorghum. Particularly important were the maize and fowls introduced in their new settlement areas. Agricultural work was performed almost exclusively by women, except for initial land clearing which was the men's responsibility (Magubane, 1998). Even today, labour division along gender lines still exists: men are traditionally hunters, herdsmen, fishermen, housing constructors, as well as traders; women are agriculturalists, gatherers, and collect water and fuelwood (Ombe, 2003). Crop harvesting was usually cooperative, done on a rotational basis, with area communities gathering to harvest each family's crop in turn.

By the early 1920s, the Tsonga-speakers constituted about 4% of the total South African population (Magubane, 1998). In the north, large chiefdoms, including Xikunda, Mhinga, Xigalo, and Makuleke occupied distinct reserves adjacent to the KNP. The Tsonga-Shangaan homeland, Gazankulu, was carved out of northern Transvaal Province during the 1960s and granted self-governing status in 1973. In the 1980s, the government of Gazankulu established a legislative assembly made up mostly of traditional chiefs. The chiefs opposed homeland independence but favored a federal arrangement with South Africa (1UpInfo, 1996).

Communities were torn apart as families were moved to the Tsonga homeland, and the resulting taxation and overpopulation made people increasingly dependent on migrant labour. This caused men to leave their families for long periods, and today even women in rural areas seek seasonal work on nearby citrus farms (Mathebula, pers. comm.).

 1 'The Crushing' - a series of Zulu and other Nguni wars and forced migrations in the early 19th century that changed the demographic, social, and political configuration of southern and central Africa.

Towards Bridging Worldviews in Biodiversity Conservation:

**2. Data collection methods and analyses** 

**2.1 Face-to-face questionnaire** 

explanation of its intended purpose(s).

show little stability and vary widely to external changes.

**2.2 Interviews** 

November 2004.

(Firey, 1978).

Exploring the Tsonga Concept of *Ntumbuloko* in South Africa 7

Resource use theory recognizes that ecological, economic, and ethnological/cultural frames of reference all interact with each other in a form of negotiation and trade-offs to optimize each of these frames and, thus, play a role in shaping perceptions of the use and fate of a resource system. This system is socially constructed and viewed differently by different social groups from their own frame of reference, based on personal needs, perceptions, and attitudes regarding a natural resource system (see Gergen, 1994; Hannigan, 1995). According to Firey, *any resource process, to be adopted, must first be accorded some worth by people in terms of their own system of activities* (emphasis ours). Thus, there are some resource complexes2 which are not valued by a given people and which, consequently, will not be adopted (no matter how superior they may be by other criteria). Resource conservation or sustainability depends on maintenance of a particular social order, because social order provides common expectations and values that make it possible for a group of people to set limits on environmental change by limiting destructive economic opportunism or 'gain-seeking'

Household face-to-face questionnaires were administered to 240 randomly selected households of 38 villages (C.I.=6.28; C.L.=95%) within 15 km of KNP in Limpopo Province (Figure 1). Data on socio-demographic variables including age, sex, household income, household size, education level and years family has resided in village were collected by trained local field assistants. Following the Firey framework, a series of questions concerning (i) community needs, (ii) the components and value of *ntumbuloko*, and (iii) costs and benefits of the KNP to local communities were also incorporated. The questionnaires integrated both closed- and open-ended questions, and were manifest (content) coded using a contextual

method based on positive/negative or topical classifications (c.f. Weisberg et al., 1996).

Questionnaires were first written in English, translated into Tsonga-Shangaan by a linguist, and then translated back into English by a field assistant. Inconsistencies and/or clarifications in the text were then modified based on pre-testing and discussions with the field assistants (c.f. Sudman, 1983). Whenever possible, household heads were surveyed and were defined as being the individual who assumed responsibility for the household (Budlender, 2003). Before administering the questionnaire, cultural norms were followed, i.e. an introduction of the administrators, the form and rationale of the questionnaire and an

Semi-structured interviews were conducted with KNP, DFED, TAs and other community representatives. Where necessary, follow-up interviews were carried out to clarify issues and explore further avenues of interest. Fieldwork was conducted from January-

<sup>2</sup> 'Resource complexes' are man-mind-land structures which show stability and resilience to external influences, and that impose constraints upon humans such that they willingly conform their behavior to the practices which comprise that resource system. These contrast with 'resource congeries', which

However, traditional Tsonga homesteads (*muti*) still exist: a typical settlement consists of a man, his wife or wives, their children and the families of their married sons (Magubane, 1998). Cylindrical houses with earth walls and conical thatched or reed roofs constitute the generally circular homestead, bordered with a perimeter wall or fence, made from branches and tree stumps. At the homestead center is the cattle kraal (*xivaya* or *tshanga*). A special meeting area (*huvo*), usually enclosed by branches and situated under a tree, exists within the community, as does the *gandzelo* for sacrificial purposes, which may be anywhere in the *muti*. The *vandal*, which may be inside or outside the *muti*, is where the men meet to discuss the administration and the affairs of the *muti*. No woman or child is allowed in this area.

Family authority rests with the father, who is traditionally treated with great respect by the wife and children. Within an extended family, the ranking and status of wives and their children is determined by the order in which they were married (Magubane, 1998). A typical Tsonga-Shangaan Traditional Authority is composed of a chief (*hosi*), under which a hierarchy exists to serve the community at large (Hartman et al.*,* 1993). Junod (1912, pp. 367) states that the role of the chief is tantamount to tribal life as 'the [chief] forms the center of national life. It is through him that the clan becomes conscious of its own unity. Without him, it loses its bearings and it has lost its head'. Chieftainship is hereditary and falls to the most senior member of the oldest lineage in the strongest clan in the group. The new chief must be approved by the council and formally inducted into office.

In the past the *hosi* yielded supreme power. He allocated land and sanctioned the start of initiation rites, harvest ceremonies and rain dances; he mediated between members of the group and ancestral spirits; he made all decisions relating to war and the army; he was also responsible for the administration of the group, and tried serious cases and those on appeal from headmen (*tindhuna*) (Magubane, 1998).

#### **1.2.2 Tsonga and** *ntumbuloko*

The concept of *ntumbuloko* dominates the Tsonga worldview and has been defined by Chitlango & Balcomb (2004:183) as 'cultural and social norms, customs, traditions, and institutions that constitute the basis for existence, self-understanding and identity in Tsonga society.' Traditional Tsonga cosmology includes that man has a physical body (*miri*) and a spiritual body with two attributes, *moya* and *ndzuti*. The *moya* (associated with the spirit) enters the body at birth, and on death is released to join the ancestors. According to Magubane (1998) the *ndzuti* is linked to a person's shadow, reflects human characteristics, and on death, leaves the body for the spirit world. The spirit of the dead (*swikwembu*) is imbued with the individual and human characteristics of the person and can hold much power with respect to causing rain to fall or trees to bear fruit (Junod, 1912). Not only is there life after death, but on entering the world of the dead the individual retains links with the living. Thus, for many Tsonga today, 'society'implies a concept including both the living and the dead.

#### **1.3 Theoretical approach: a space for consideration of Tsonga and official views of environment and conservation**

Our research primarily draws from Firey's (1960) resource use theory as it provides a comprehensive approach to understanding human dimensions of resource management.

However, traditional Tsonga homesteads (*muti*) still exist: a typical settlement consists of a man, his wife or wives, their children and the families of their married sons (Magubane, 1998). Cylindrical houses with earth walls and conical thatched or reed roofs constitute the generally circular homestead, bordered with a perimeter wall or fence, made from branches and tree stumps. At the homestead center is the cattle kraal (*xivaya* or *tshanga*). A special meeting area (*huvo*), usually enclosed by branches and situated under a tree, exists within the community, as does the *gandzelo* for sacrificial purposes, which may be anywhere in the *muti*. The *vandal*, which may be inside or outside the *muti*, is where the men meet to discuss the administration and the affairs of the *muti*. No woman or child is

Family authority rests with the father, who is traditionally treated with great respect by the wife and children. Within an extended family, the ranking and status of wives and their children is determined by the order in which they were married (Magubane, 1998). A typical Tsonga-Shangaan Traditional Authority is composed of a chief (*hosi*), under which a hierarchy exists to serve the community at large (Hartman et al.*,* 1993). Junod (1912, pp. 367) states that the role of the chief is tantamount to tribal life as 'the [chief] forms the center of national life. It is through him that the clan becomes conscious of its own unity. Without him, it loses its bearings and it has lost its head'. Chieftainship is hereditary and falls to the most senior member of the oldest lineage in the strongest clan in the group. The new chief

In the past the *hosi* yielded supreme power. He allocated land and sanctioned the start of initiation rites, harvest ceremonies and rain dances; he mediated between members of the group and ancestral spirits; he made all decisions relating to war and the army; he was also responsible for the administration of the group, and tried serious cases and those on appeal

The concept of *ntumbuloko* dominates the Tsonga worldview and has been defined by Chitlango & Balcomb (2004:183) as 'cultural and social norms, customs, traditions, and institutions that constitute the basis for existence, self-understanding and identity in Tsonga society.' Traditional Tsonga cosmology includes that man has a physical body (*miri*) and a spiritual body with two attributes, *moya* and *ndzuti*. The *moya* (associated with the spirit) enters the body at birth, and on death is released to join the ancestors. According to Magubane (1998) the *ndzuti* is linked to a person's shadow, reflects human characteristics, and on death, leaves the body for the spirit world. The spirit of the dead (*swikwembu*) is imbued with the individual and human characteristics of the person and can hold much power with respect to causing rain to fall or trees to bear fruit (Junod, 1912). Not only is there life after death, but on entering the world of the dead the individual retains links with the living. Thus, for many Tsonga today, 'society'implies a concept including both the living

**1.3 Theoretical approach: a space for consideration of Tsonga and official views of** 

Our research primarily draws from Firey's (1960) resource use theory as it provides a comprehensive approach to understanding human dimensions of resource management.

must be approved by the council and formally inducted into office.

from headmen (*tindhuna*) (Magubane, 1998).

**1.2.2 Tsonga and** *ntumbuloko*

**environment and conservation** 

and the dead.

allowed in this area.

Resource use theory recognizes that ecological, economic, and ethnological/cultural frames of reference all interact with each other in a form of negotiation and trade-offs to optimize each of these frames and, thus, play a role in shaping perceptions of the use and fate of a resource system. This system is socially constructed and viewed differently by different social groups from their own frame of reference, based on personal needs, perceptions, and attitudes regarding a natural resource system (see Gergen, 1994; Hannigan, 1995). According to Firey, *any resource process, to be adopted, must first be accorded some worth by people in terms of their own system of activities* (emphasis ours). Thus, there are some resource complexes2 which are not valued by a given people and which, consequently, will not be adopted (no matter how superior they may be by other criteria). Resource conservation or sustainability depends on maintenance of a particular social order, because social order provides common expectations and values that make it possible for a group of people to set limits on environmental change by limiting destructive economic opportunism or 'gain-seeking' (Firey, 1978).

## **2. Data collection methods and analyses**

## **2.1 Face-to-face questionnaire**

Household face-to-face questionnaires were administered to 240 randomly selected households of 38 villages (C.I.=6.28; C.L.=95%) within 15 km of KNP in Limpopo Province (Figure 1). Data on socio-demographic variables including age, sex, household income, household size, education level and years family has resided in village were collected by trained local field assistants. Following the Firey framework, a series of questions concerning (i) community needs, (ii) the components and value of *ntumbuloko*, and (iii) costs and benefits of the KNP to local communities were also incorporated. The questionnaires integrated both closed- and open-ended questions, and were manifest (content) coded using a contextual method based on positive/negative or topical classifications (c.f. Weisberg et al., 1996).

Questionnaires were first written in English, translated into Tsonga-Shangaan by a linguist, and then translated back into English by a field assistant. Inconsistencies and/or clarifications in the text were then modified based on pre-testing and discussions with the field assistants (c.f. Sudman, 1983). Whenever possible, household heads were surveyed and were defined as being the individual who assumed responsibility for the household (Budlender, 2003). Before administering the questionnaire, cultural norms were followed, i.e. an introduction of the administrators, the form and rationale of the questionnaire and an explanation of its intended purpose(s).

#### **2.2 Interviews**

Semi-structured interviews were conducted with KNP, DFED, TAs and other community representatives. Where necessary, follow-up interviews were carried out to clarify issues and explore further avenues of interest. Fieldwork was conducted from January-November 2004.

 <sup>2</sup> 'Resource complexes' are man-mind-land structures which show stability and resilience to external influences, and that impose constraints upon humans such that they willingly conform their behavior to the practices which comprise that resource system. These contrast with 'resource congeries', which show little stability and vary widely to external changes.

Towards Bridging Worldviews in Biodiversity Conservation:

**3.1 Demographic and socio-economic factors** 

incorporated Firey's concepts.

**3. Results** 

(mean=23.2±12.60).

**3.2 Community needs** 

concern for conservation agencies.

Overall

Exploring the Tsonga Concept of *Ntumbuloko* in South Africa 9

primary liaison between KNP and neighbouring communities in the northern part of the park since 1994. During analyses, we initially utilized a grounded coding process to identify themes in interview and archival data, followed by a more explicit coding process that

The questionnaire sample consisted of 83 males (34.6%) and 157 females (65.4%), ranging in age from 18 to 102 (mean=39.3±17.63). Household sizes ranged from 1 to 18 persons (mean=5.8±2.65), and families had resided in their village from 1 to 52 years

Respondents were also asked to list the ages and sex of all household members. Men (N=662, mean age=22.1±17.102) represented 47.52% of the sampled households, while women (N=731, mean age=26.5±19.716) constituted 52.48%. The population structure is broad-based with over half of the population <20 yrs of age, and comprises a higher

Survey respondents were asked to rank the five most important community needs from a predefined list, based on interviews with community members and municipal government staff (Table 1). A weighted score was calculated for each need and used as an indicator of its importance. Employment was ranked as the most important community need overall, followed by health, school, electricity and drinking water facilities. Of least importance to respondents were protecting forests and wild animals which, in contrast, are of primary

> Rank Community need *n* mean score 1 Employment 185 3.10 2 Health facilities 164 2.37 3 School facilities 182 2.34 4 Electricity facilities 144 1.95 5 Drinking water facilities 111 1.26 6 Road improvement 81 0.80 7 Training opportunities 86 0.74 8 Protection of crops/livestock 61 0.73 9 Housing 52 0.60 10 Preserving traditional culture 36 0.33 11 Tourism development 27 0.29 12 Protection of forest 29 0.26 13 Protection of wild animals 32 0.26

Table 1. Overall ranking of community needs by community survey respondents (N=238).

Mean scores range from 0 (no importance) to 5 (most important).

proportion of women compared to men, especially in age classes above 29 yrs.

Fig. 1. (a) The location of Kruger National Park in Southern Africa. (b) Expanded view illustrates study area with location of 38 villages (listed below with associated *de jure* Traditional Authorities). **Mhinga TA:** Matiyani (1), Josepha (2), Mhinga (3), Botsoleni (4), Maphophe (5), Maviligwe (6), Makuleke (7), Makahlule (8); **Shikundu TA:** Ximixoni (9), Saselemani (10), Nkovani (11); **Bevhula TA**: Ntlhaveni D (12), Nkavela (13), Makhubele (14), Bevhula (15); **Magona TA**: Nghomunghomu (16), Mashobye (17), Magona (18); **Madonsi TA**: Gijamhandzeni (19), Matsakali (20), Halahala (21), Peninghotsa (22), Govhu (23), Merwe A (24), Shisasi (25), Jilongo (26); **Mtititi TA**: Lombaard (27), Plange (28), Altein (29); **Xiviti TA**: Mininginisi Block 3 (30), Mininginisi Block 2 (31), Muyexe (32), Shitshamayoshe (33), Khakhala (34), Gawula (35), Mahlathi (36), Ndindani (37), Hlomela (38). Source: Anthony (2007); reproduced with permission from Cambridge University Press.

#### **2.3 Archival research**

Limpopo Province DFED records were compiled from both Mopani District, which extends from the Shingwedzi River south through the study area, and Vhembe District, which includes the northern section of the study area. Moreover, relevant records of the KNP and monthly meeting minutes were reviewed of the Hlanganani Forum, which has been the primary liaison between KNP and neighbouring communities in the northern part of the park since 1994. During analyses, we initially utilized a grounded coding process to identify themes in interview and archival data, followed by a more explicit coding process that incorporated Firey's concepts.

## **3. Results**

8 Research in Biodiversity – Models and Applications

Fig. 1. (a) The location of Kruger National Park in Southern Africa. (b) Expanded view illustrates study area with location of 38 villages (listed below with associated *de jure* Traditional Authorities). **Mhinga TA:** Matiyani (1), Josepha (2), Mhinga (3), Botsoleni (4), Maphophe (5), Maviligwe (6), Makuleke (7), Makahlule (8); **Shikundu TA:** Ximixoni (9), Saselemani (10), Nkovani (11); **Bevhula TA**: Ntlhaveni D (12), Nkavela (13), Makhubele (14), Bevhula (15); **Magona TA**: Nghomunghomu (16), Mashobye (17), Magona (18); **Madonsi TA**: Gijamhandzeni (19), Matsakali (20), Halahala (21), Peninghotsa (22), Govhu (23), Merwe A (24), Shisasi (25), Jilongo (26); **Mtititi TA**: Lombaard (27), Plange (28), Altein (29); **Xiviti TA**: Mininginisi Block 3 (30), Mininginisi Block 2 (31), Muyexe (32), Shitshamayoshe (33),

Source: Anthony (2007); reproduced with permission from Cambridge University Press.

Limpopo Province DFED records were compiled from both Mopani District, which extends from the Shingwedzi River south through the study area, and Vhembe District, which includes the northern section of the study area. Moreover, relevant records of the KNP and monthly meeting minutes were reviewed of the Hlanganani Forum, which has been the

Khakhala (34), Gawula (35), Mahlathi (36), Ndindani (37), Hlomela (38).

**2.3 Archival research** 

## **3.1 Demographic and socio-economic factors**

The questionnaire sample consisted of 83 males (34.6%) and 157 females (65.4%), ranging in age from 18 to 102 (mean=39.3±17.63). Household sizes ranged from 1 to 18 persons (mean=5.8±2.65), and families had resided in their village from 1 to 52 years (mean=23.2±12.60).

Respondents were also asked to list the ages and sex of all household members. Men (N=662, mean age=22.1±17.102) represented 47.52% of the sampled households, while women (N=731, mean age=26.5±19.716) constituted 52.48%. The population structure is broad-based with over half of the population <20 yrs of age, and comprises a higher proportion of women compared to men, especially in age classes above 29 yrs.

### **3.2 Community needs**

Survey respondents were asked to rank the five most important community needs from a predefined list, based on interviews with community members and municipal government staff (Table 1). A weighted score was calculated for each need and used as an indicator of its importance. Employment was ranked as the most important community need overall, followed by health, school, electricity and drinking water facilities. Of least importance to respondents were protecting forests and wild animals which, in contrast, are of primary concern for conservation agencies.


Table 1. Overall ranking of community needs by community survey respondents (N=238). Mean scores range from 0 (no importance) to 5 (most important).

Towards Bridging Worldviews in Biodiversity Conservation:

Direct value Indirect value

*Consumptive/non-market (27.5%)*:

fodder for animals

 traditional medicine construction materials traditional clothing

*Productive/commercial (4.7%):*  fodder for animals traditional medicine drawing tourists

**3.4 Traditional Authorities** 

food

fuelwood

Exploring the Tsonga Concept of *Ntumbuloko* in South Africa 11

 storm protection cleaning air soil protection sustains environment

God's creation')

*Option (6.7%)*:

to keep snake bites down at the college where no one has been bitten in 3 years'.

percentages of responses are included for each sub-category.

Table 2. Categorized responses as to why community members 'need' *ntumbuloko*. Relative

In an informal conversation, one high school teacher in the area stated that he believes the ancestors' spirits can control rain and consequently crop production, therefore those who are still living must continue to honor them through dances, drums, and meetings. However, this cosmology is neither universal nor static amongst the Tsonga. According to one college teacher from the area, the Tsonga primarily define 'beauty' of plants and animals according to their use or utility. He reported that since he began teaching in 1989, his personal perception on nature has changed because 'they [campus management] made it wrong for us to kill any animals on the campus'. He usually would kill a snake on site as is the Tsonga custom, but now he 'tries to chase it away'. He now believes that this 'has helped

Respondents were asked to evaluate both their respective TA and the municipal government, in terms of how well it was doing in its role with respect to land-use, whatever they conceived that to be. More than half of the respondents (51.6%) couldn't comment on the effectiveness of the municipal government, stating that they didn't know of its activities. For those that did evaluate the institution, 23.8% assessed it positively and 24.6% negatively. Negative opinions of the effectiveness of the municipal government were largely governed by housing and water shortages, poor road maintenance, and the belief that it 'does nothing in our area' and 'shows favoritism in its activities'. These data collectively suggest that the performance of municipal government is highly varied in the study area, with specific *de jure* TAs experiencing greater activity than others. In contrast, the roles and responsibilities of Traditional Authorities are much better recognized, with respondents stating that their functions are extensive, ranging from provision of residential and agricultural sites, to protecting forests/wild animals and

*Non-consumptive - ecological functions (19.5%)*:

*Non-consumptive – non-ecological functions (41.6%)*: part of creation ('I belong to it'; 'makes us aware of

for future generations, 'to build the future'.

learn from it as they grow up')

education ('we can learn much from it'; 'children

 historical heritage ('it serves as a reminder of the past') aesthetic ('brings and brightens life for people') cultural ('it is our culture to love *ntumbuloko*')

#### **3.3 Beliefs and attitudes**

Respondents were asked what they believed to be components of *ntumbuloko;* responses are summarized in Figure 2. Chi-square and correlation tests were conducted for gender, age, household income and education level but no significant associations were found, suggesting that beliefs in the sampled households regarding the different parts of *ntumbuloko* are independent of these variables.

Based on their concept of *ntumbuloko*, almost all (98.7%) respondents believed that they 'need' *ntumbuloko*, for a variety of reasons which we classified according to McNeely et al. (1990) (Table 2). In addition to more direct utilitarian values, respondents indicated that *ntumbuloko* is highly valued for its socio-cultural, educational, spiritual and historical attributes. When respondents were asked whether they believed they needed to protect *ntumbuloko*, a majority (85.4%) agreed. The need to maintain and enhance utilitarian use values ranked highest for those responding positively to this question, although socio-cultural and spiritual aspects were also noted, including the following: '*it is life'; 'to lose* ntumbuloko *is to lose ourselves'; '*ntumbuloko *dictates that we should continue initiation school 3*".

Fig. 2. Frequency of belief about components of *ntumbuloko* (N=240)

Ten percent of the respondents stated that they didn't know whether they should protect *ntumbuloko*, claiming that they didn't know how they could protect it. In contrast, 4.6% indicated that they did not believe they needed to protect *ntumbuloko*, citing that "*it was created long ago*".

<sup>3</sup> In traditional Tsonga culture, puberty marks the end of childhood and the beginning of adolescence. During this time young men and women enter initiation schools. Schools vary, but in principle they perform a similar social function, that of a 'rite of passage' marking the transition from adolescence to adulthood. This is much more than a physical change; it also represents a change in social status.

Respondents were asked what they believed to be components of *ntumbuloko;* responses are summarized in Figure 2. Chi-square and correlation tests were conducted for gender, age, household income and education level but no significant associations were found, suggesting that beliefs in the sampled households regarding the different parts of

Based on their concept of *ntumbuloko*, almost all (98.7%) respondents believed that they 'need' *ntumbuloko*, for a variety of reasons which we classified according to McNeely et al. (1990) (Table 2). In addition to more direct utilitarian values, respondents indicated that *ntumbuloko* is highly valued for its socio-cultural, educational, spiritual and historical attributes. When respondents were asked whether they believed they needed to protect *ntumbuloko*, a majority (85.4%) agreed. The need to maintain and enhance utilitarian use values ranked highest for those responding positively to this question, although socio-cultural and spiritual aspects were also noted, including the following: '*it is life'; 'to lose* ntumbuloko *is to lose ourselves';* 

**3.3 Beliefs and attitudes** 

*ntumbuloko* are independent of these variables.

*'*ntumbuloko *dictates that we should continue initiation school 3*".

Fig. 2. Frequency of belief about components of *ntumbuloko* (N=240)

*created long ago*".

Ten percent of the respondents stated that they didn't know whether they should protect *ntumbuloko*, claiming that they didn't know how they could protect it. In contrast, 4.6% indicated that they did not believe they needed to protect *ntumbuloko*, citing that "*it was* 

3 In traditional Tsonga culture, puberty marks the end of childhood and the beginning of adolescence. During this time young men and women enter initiation schools. Schools vary, but in principle they perform a similar social function, that of a 'rite of passage' marking the transition from adolescence to adulthood. This is much more than a physical change; it also represents a change in social status.


Table 2. Categorized responses as to why community members 'need' *ntumbuloko*. Relative percentages of responses are included for each sub-category.

In an informal conversation, one high school teacher in the area stated that he believes the ancestors' spirits can control rain and consequently crop production, therefore those who are still living must continue to honor them through dances, drums, and meetings. However, this cosmology is neither universal nor static amongst the Tsonga. According to one college teacher from the area, the Tsonga primarily define 'beauty' of plants and animals according to their use or utility. He reported that since he began teaching in 1989, his personal perception on nature has changed because 'they [campus management] made it wrong for us to kill any animals on the campus'. He usually would kill a snake on site as is the Tsonga custom, but now he 'tries to chase it away'. He now believes that this 'has helped to keep snake bites down at the college where no one has been bitten in 3 years'.

#### **3.4 Traditional Authorities**

Respondents were asked to evaluate both their respective TA and the municipal government, in terms of how well it was doing in its role with respect to land-use, whatever they conceived that to be. More than half of the respondents (51.6%) couldn't comment on the effectiveness of the municipal government, stating that they didn't know of its activities. For those that did evaluate the institution, 23.8% assessed it positively and 24.6% negatively. Negative opinions of the effectiveness of the municipal government were largely governed by housing and water shortages, poor road maintenance, and the belief that it 'does nothing in our area' and 'shows favoritism in its activities'. These data collectively suggest that the performance of municipal government is highly varied in the study area, with specific *de jure* TAs experiencing greater activity than others. In contrast, the roles and responsibilities of Traditional Authorities are much better recognized, with respondents stating that their functions are extensive, ranging from provision of residential and agricultural sites, to protecting forests/wild animals and

Towards Bridging Worldviews in Biodiversity Conservation:

of nature held by the Tsonga were also noted.

with permission from Anthony (2006)

**4.3 Community needs** 

Exploring the Tsonga Concept of *Ntumbuloko* in South Africa 13

educational, spiritual and historical qualities (Anthony & Bellinger 2007). The need to maintain and enhance utilitarian use values ranked highest for those responding positively to the question whether they need to protect *ntumbuloko*, although socio-cultural and spiritual aspects were also noted. In addition to holding a broader view of nature, Tsonga also believe in a plethora of practices which they see as being essential for its protection. In addition to reduced consumption of resources, environmental education, and altering practices to protect flora and fauna (which one might expect in western societies), the need to maintain cultural and spiritual traditions which are embedded in the broader definition

Opinions expressed on nature conservation, i.e. protecting trees and wild animals, lag far below more immediate development needs such as employment, health, education, and improving infrastructure. Employment needs were apparent as we noted male absence in the study area, which is likely attributable to outmigration (or cyclic migration) to larger urban centers or mines where employment opportunities are greater (Bryceson, 1999). Male absence in rural areas can create labor vacuums, especially in cases where domestic responsibilities are sharply divided amongst household members, and may disproportionately increase pressures on households with only women and children. In this research about one in eight households was comprised only of women and children. This constraint is exacerbated by time required for women and children carrying out domestic chores, including almost 20 hours per week for collecting fuelwood and drinking water alone in the study area (Figure 3) (Anthony, 2006). With water scarcity perceived to be widespread in the study area, and fuelwood becoming scarcer in some areas north of the

Fig. 3. Tsonga woman on route to collect drinking water from community tap. Reproduced

overseeing people's concerns. Considering that access to land for cultivation was secure for over 70% of respondents, and more than 85% felt their land was 'good', this suggests that *TAs are perceived as largely competent by local communities in securing access to good quality land for agriculture*. Moreover*, TAs have a much higher approval rating compared to local government by respondents,* with less than 12% of respondents reporting negatively overall.

In order to identify what variables might be influencing this evaluation, correlation analysis was used to compare responses with selected demographic and socio-economic variables. Although age (r=0.14, p<0.05, N=240) and level of education (r=-0.13, p<0.05, N=240) were significantly correlated with responses towards TA effectiveness, linear regression analysis revealed that they are very weak predictors of responses (R2=0.02), suggesting that the selected variables do not play a decisive role in influencing opinions.

These functional distinctions were also confirmed during interviews with various community members and representatives of TAs. According to one *hosi* (chief), although all communal lands are owned by the state, TAs have authority to grant lands for garden plots and homesteads to their *muganga* (village(s)) members. Mtititi TA representatives stated that they are responsible for access to and control over a number of resources, including allocation of grazing and residential sites, and granting permission to collect fuelwood. They play a judicial role in fining any persons caught illegally collecting any resource that requires a permit, especially those persons who do not reside within the TA area, in which case guilty parties receive a stiffer penalty. They also play an important role in resource monitoring stating, 'In the event that the tribal police see that the amounts of resources are dwindling, they inform the *hosi* who would then inform the community to cease collecting that resource.'

#### **4. Discussion**

#### **4.1 Components of** *ntumbuloko*

South Africa has undergone dramatic socio-political changes in the last decade, with enhanced opportunities for formal education in the rural areas. However, the extent to which formal education and exposure to alternative views has affected perceptions and attitudes of rural people towards nature and its conservation is still uncertain (see Els, 1994; Mabunda, 2004). *Ntumbuloko* permeates the Tsonga worldview, and our research supports previous work (Chitlango & Balcomb, 2004; Els, 2002; Junod, 1913; Terblanche, 1994) in that the Tsonga perceive *ntumbuloko* as more than just the biophysical environment: there is still strong belief that it also embraces people (*vanhu*), God (*Xikwembu*), ancestors' spirits (*swikembu*), and tradition, and this belief is independent of sex, age, household income and education level. These results are congruent with a study on perceptions regarding causes and treatment of diseases in Northern [now Limpopo] Province (Mabunda, 2001), which found that the notion of supernatural causality associated with many diseases predominated among all groups, but was highest among university students. In our study, supernatural causality still prevails and is manifested in the belief of many respondents, even amongst the young and more highly educated, that rain and associated harvests are strongly linked with appeasing ancestors' spirits, and not solely the product of environmental factors, which western science principles would prescribe.

#### **4.2 Value of** *ntumbuloko*

In addition to more direct utilitarian values, *ntumbuloko* is highly valued for its indirect nonconsumptive attributes, including non-ecological functions embracing socio-cultural, educational, spiritual and historical qualities (Anthony & Bellinger 2007). The need to maintain and enhance utilitarian use values ranked highest for those responding positively to the question whether they need to protect *ntumbuloko*, although socio-cultural and spiritual aspects were also noted. In addition to holding a broader view of nature, Tsonga also believe in a plethora of practices which they see as being essential for its protection. In addition to reduced consumption of resources, environmental education, and altering practices to protect flora and fauna (which one might expect in western societies), the need to maintain cultural and spiritual traditions which are embedded in the broader definition of nature held by the Tsonga were also noted.

#### **4.3 Community needs**

12 Research in Biodiversity – Models and Applications

overseeing people's concerns. Considering that access to land for cultivation was secure for over 70% of respondents, and more than 85% felt their land was 'good', this suggests that *TAs are perceived as largely competent by local communities in securing access to good quality land for agriculture*. Moreover*, TAs have a much higher approval rating compared to local government by* 

In order to identify what variables might be influencing this evaluation, correlation analysis was used to compare responses with selected demographic and socio-economic variables. Although age (r=0.14, p<0.05, N=240) and level of education (r=-0.13, p<0.05, N=240) were significantly correlated with responses towards TA effectiveness, linear regression analysis revealed that they are very weak predictors of responses (R2=0.02), suggesting that the

These functional distinctions were also confirmed during interviews with various community members and representatives of TAs. According to one *hosi* (chief), although all communal lands are owned by the state, TAs have authority to grant lands for garden plots and homesteads to their *muganga* (village(s)) members. Mtititi TA representatives stated that they are responsible for access to and control over a number of resources, including allocation of grazing and residential sites, and granting permission to collect fuelwood. They play a judicial role in fining any persons caught illegally collecting any resource that requires a permit, especially those persons who do not reside within the TA area, in which case guilty parties receive a stiffer penalty. They also play an important role in resource monitoring stating, 'In the event that the tribal police see that the amounts of resources are dwindling, they inform the

South Africa has undergone dramatic socio-political changes in the last decade, with enhanced opportunities for formal education in the rural areas. However, the extent to which formal education and exposure to alternative views has affected perceptions and attitudes of rural people towards nature and its conservation is still uncertain (see Els, 1994; Mabunda, 2004). *Ntumbuloko* permeates the Tsonga worldview, and our research supports previous work (Chitlango & Balcomb, 2004; Els, 2002; Junod, 1913; Terblanche, 1994) in that the Tsonga perceive *ntumbuloko* as more than just the biophysical environment: there is still strong belief that it also embraces people (*vanhu*), God (*Xikwembu*), ancestors' spirits (*swikembu*), and tradition, and this belief is independent of sex, age, household income and education level. These results are congruent with a study on perceptions regarding causes and treatment of diseases in Northern [now Limpopo] Province (Mabunda, 2001), which found that the notion of supernatural causality associated with many diseases predominated among all groups, but was highest among university students. In our study, supernatural causality still prevails and is manifested in the belief of many respondents, even amongst the young and more highly educated, that rain and associated harvests are strongly linked with appeasing ancestors' spirits, and not solely the product of

*respondents,* with less than 12% of respondents reporting negatively overall.

selected variables do not play a decisive role in influencing opinions.

*hosi* who would then inform the community to cease collecting that resource.'

environmental factors, which western science principles would prescribe.

In addition to more direct utilitarian values, *ntumbuloko* is highly valued for its indirect nonconsumptive attributes, including non-ecological functions embracing socio-cultural,

**4. Discussion** 

**4.1 Components of** *ntumbuloko*

**4.2 Value of** *ntumbuloko*

Opinions expressed on nature conservation, i.e. protecting trees and wild animals, lag far below more immediate development needs such as employment, health, education, and improving infrastructure. Employment needs were apparent as we noted male absence in the study area, which is likely attributable to outmigration (or cyclic migration) to larger urban centers or mines where employment opportunities are greater (Bryceson, 1999). Male absence in rural areas can create labor vacuums, especially in cases where domestic responsibilities are sharply divided amongst household members, and may disproportionately increase pressures on households with only women and children. In this research about one in eight households was comprised only of women and children. This constraint is exacerbated by time required for women and children carrying out domestic chores, including almost 20 hours per week for collecting fuelwood and drinking water alone in the study area (Figure 3) (Anthony, 2006). With water scarcity perceived to be widespread in the study area, and fuelwood becoming scarcer in some areas north of the

Fig. 3. Tsonga woman on route to collect drinking water from community tap. Reproduced with permission from Anthony (2006)

Towards Bridging Worldviews in Biodiversity Conservation:

**4.4 Traditional Authorities** 

Africa's communal areas.

caused by these animals (Anthony, 2007).

community-based conservation initiatives at all.

**4.5 'Gain-seekers' and resource exploitation** 

Exploring the Tsonga Concept of *Ntumbuloko* in South Africa 15

Embedded cultural and spiritual beliefs and practices hold value for the Tsonga and should be acknowledged when establishing partnerships in environmental protection. This includes the role that *ntumbuloko* has for the Tsonga in education, spiritual identity and as historical heritage. These beliefs, strongly held by many Tsonga, are thus very resistant to change and are likely to persist. It is these beliefs which have the greatest potential to conflict with western approaches to conservation, as they claim inherent differences with respect to *who* is responsible for protecting flora and fauna, and *how* they are to be used. Practically for conservation initiatives, the two concepts regarding Tsonga beliefs explained above translate into the recognition that *conservation programs are unlikely to be accepted in these contexts if they are based primarily on aesthetic values of nature, or if they do not acknowledge the* 

The strong role that TAs play in land allocation and resource access and use has a number of far reaching implications. Chiefly authority is ascribed by lineage rather than achieved through elections, and its patriarchal principles ensure that major decisions on land allocation are almost invariably taken by men. However, this research shows that many people, irrespective of gender, still look to their chiefs for land allocation and are satisfied with it. Indeed, only 10.2% of women respondents felt that their TAs are not doing a good job, compared with 14.5% of men. These results concur with Campbell & Shackleton (2001) and Ntsebeza (1999), who showed that TAs still maintain strong positive influence in South

The role of DFED in the study area is uncertain and ambiguous. Although the primary body responsible for implementing and enforcing LEMA 2003 regulations, its activities are limited. Indeed, TAs are *de facto* principally controlling access to natural resources and enforcing LEMA 2003 stipulations, with tribal courts functioning in part to fine transgressors. Perceptions of the DFED by local TAs are generally negative, as this agency is seen only within its role in enforcement. It is also criticized for its weakness in delivering much-needed environmental education and awareness to communities on the role of the provincial government. In addition, there is widespread criticism of the poor control of damage-causing animals by DFED and the withholding of compensation for damages

Similar to criticisms launched at the ineffectiveness of local government, weaknesses in cooperative governance between DFED and TAs are inhibiting resource conservation, leading to situations in which opportunities are established for 'gain-seekers' to exploit resources at unsustainable rates. DFED managerial staff acknowledge that discussion and co-operation regarding land use, including biodiversity conservation, between provincial and municipal governments and TAs is practically non-existent, and needs to be strengthened (Anthony, 2006). In light of the increasing pressures on natural resources and the aspirations of some communities to engage in conservation agreements with the KNP, it would be wise for these institutions to heed these trends and seek co-operative ways to halt resource overexploitation before conditions render it practically impossible to effectively pursue any

The Firey model contends that resource conservation is possible only when people share expectations that others will forego opportunistic practices threatening sustainability.

*belief by local communities of the role that God and ancestors' spirits play in nature*.

Shingwedzi River, the extent of these constraints appears to be worsening (Anthony, 2006). These constraints suggest that opportunities for women and children desiring to secure formal employment, training, and/or education are severely limited. For conservation agencies, recognizing these limitations is an important step in articulating any conservation and/or development programs that seek local relevance. *Time is a precious commodity that should be understood in its local context, and household members are unlikely to engage in activities making extensive demands on their time unless these are directly related to improving livelihoods.*

As we think further about the needs of communities, the question arises, *If local communities are so dependent on local wild resources, why is their protection ranked so low?* The answer may be found in two related concepts of Tsonga beliefs, i.e. values associated with *ntumbuloko*, and the role of humans in the environment.

First, the Tsonga value *ntumbuloko* more for its utilitarian rather than aesthetic qualities, believing that local resources were given by God, and it is their right to use them to maintain human survival (Eckert et al., 2001; Els, 2002). However, 'meaningful and judicious use is not always implied by this inherent right, and this difference in conceptual approach often leads to conflict with nature conservation authorities' (Els, 2002, p.655); thus resource use conflicts are often rooted deeply in culture. Hence, the negotiations of resource users as conceptualized in Firey's theory then become operational: the perceived *aesthetic values* of nature are 'traded off' for more imperative needs of human survival and development. Here, however, distinctions within and between Firey's three frames become blurred, limiting its application in these contexts. Western concepts of the 'ecological frame', developed mainly by ecologists and geographers, are based on the interactions between organisms and their biophysical environments. Conversely, the 'ethnological frame' to resource phenomena has principally been developed by anthropologists and sociologists and focuses on a people's culture. Firey's definition and explanation of these frames treats them as separate entities. However, the Tsonga concept of *ntumbuloko* embodies both ecological and cultural frames; decoupling it into two separate frames, at present, is irrational for most Tsonga. Therefore, *developing nature conservation activities in these contexts have a greater chance of being rejected if they do not incorporate the wider concept of* ntumbuloko *constructed by the Tsonga*. This also has implications for current stakeholders and future researchers in similar contexts: research findings may have lower relevance and/or be more difficult to communicate locally if these distinctions in conceptual definitions are not recognized.

Second, it is inconceivable and irrational for the Tsonga to believe that protection of forests and wild animals is man's responsibility (Els, 2002). On one hand, our research supports Els' view, as most respondents believe that it is God's (*Xikwembu*) responsibility to ultimately ensure the sustainability of resources. On the other hand, although God (*Xikwembu*) and ancestors' spirits (*swikwembu*) are still believed to be components of *ntumbuloko*, such beliefs may not be as widespread as they were in the past. For example, in a study of Tsonga communities in a more densely populated region adjacent to KNP to the south, Hunter et al. (2010) found that environmental concern was strongly related to material needs and livelihoods, and this was gendered and varied substantially by village. This transition may be the result of increasing exposure to Christianity, alternative views of nature in educational institutions (Millar, 2004), economic development opportunities or cultural taboos (Kuriyan, 2002), and/or restrictions on resource use imposed by government and TAs, although such causal relationships were beyond the scope of our research.

Embedded cultural and spiritual beliefs and practices hold value for the Tsonga and should be acknowledged when establishing partnerships in environmental protection. This includes the role that *ntumbuloko* has for the Tsonga in education, spiritual identity and as historical heritage. These beliefs, strongly held by many Tsonga, are thus very resistant to change and are likely to persist. It is these beliefs which have the greatest potential to conflict with western approaches to conservation, as they claim inherent differences with respect to *who* is responsible for protecting flora and fauna, and *how* they are to be used. Practically for conservation initiatives, the two concepts regarding Tsonga beliefs explained above translate into the recognition that *conservation programs are unlikely to be accepted in these contexts if they are based primarily on aesthetic values of nature, or if they do not acknowledge the belief by local communities of the role that God and ancestors' spirits play in nature*.

## **4.4 Traditional Authorities**

14 Research in Biodiversity – Models and Applications

Shingwedzi River, the extent of these constraints appears to be worsening (Anthony, 2006). These constraints suggest that opportunities for women and children desiring to secure formal employment, training, and/or education are severely limited. For conservation agencies, recognizing these limitations is an important step in articulating any conservation and/or development programs that seek local relevance. *Time is a precious commodity that should be understood in its local context, and household members are unlikely to engage in activities making extensive demands on their time unless these are directly related to improving livelihoods.* As we think further about the needs of communities, the question arises, *If local communities are so dependent on local wild resources, why is their protection ranked so low?* The answer may be found in two related concepts of Tsonga beliefs, i.e. values associated with *ntumbuloko*, and

First, the Tsonga value *ntumbuloko* more for its utilitarian rather than aesthetic qualities, believing that local resources were given by God, and it is their right to use them to maintain human survival (Eckert et al., 2001; Els, 2002). However, 'meaningful and judicious use is not always implied by this inherent right, and this difference in conceptual approach often leads to conflict with nature conservation authorities' (Els, 2002, p.655); thus resource use conflicts are often rooted deeply in culture. Hence, the negotiations of resource users as conceptualized in Firey's theory then become operational: the perceived *aesthetic values* of nature are 'traded off' for more imperative needs of human survival and development. Here, however, distinctions within and between Firey's three frames become blurred, limiting its application in these contexts. Western concepts of the 'ecological frame', developed mainly by ecologists and geographers, are based on the interactions between organisms and their biophysical environments. Conversely, the 'ethnological frame' to resource phenomena has principally been developed by anthropologists and sociologists and focuses on a people's culture. Firey's definition and explanation of these frames treats them as separate entities. However, the Tsonga concept of *ntumbuloko* embodies both ecological and cultural frames; decoupling it into two separate frames, at present, is irrational for most Tsonga. Therefore, *developing nature conservation activities in these contexts have a greater chance of being rejected if they do not incorporate the wider concept of* ntumbuloko *constructed by the Tsonga*. This also has implications for current stakeholders and future researchers in similar contexts: research findings may have lower relevance and/or be more difficult to communicate locally if these distinctions in conceptual

Second, it is inconceivable and irrational for the Tsonga to believe that protection of forests and wild animals is man's responsibility (Els, 2002). On one hand, our research supports Els' view, as most respondents believe that it is God's (*Xikwembu*) responsibility to ultimately ensure the sustainability of resources. On the other hand, although God (*Xikwembu*) and ancestors' spirits (*swikwembu*) are still believed to be components of *ntumbuloko*, such beliefs may not be as widespread as they were in the past. For example, in a study of Tsonga communities in a more densely populated region adjacent to KNP to the south, Hunter et al. (2010) found that environmental concern was strongly related to material needs and livelihoods, and this was gendered and varied substantially by village. This transition may be the result of increasing exposure to Christianity, alternative views of nature in educational institutions (Millar, 2004), economic development opportunities or cultural taboos (Kuriyan, 2002), and/or restrictions on resource use imposed by government and

TAs, although such causal relationships were beyond the scope of our research.

the role of humans in the environment.

definitions are not recognized.

The strong role that TAs play in land allocation and resource access and use has a number of far reaching implications. Chiefly authority is ascribed by lineage rather than achieved through elections, and its patriarchal principles ensure that major decisions on land allocation are almost invariably taken by men. However, this research shows that many people, irrespective of gender, still look to their chiefs for land allocation and are satisfied with it. Indeed, only 10.2% of women respondents felt that their TAs are not doing a good job, compared with 14.5% of men. These results concur with Campbell & Shackleton (2001) and Ntsebeza (1999), who showed that TAs still maintain strong positive influence in South Africa's communal areas.

The role of DFED in the study area is uncertain and ambiguous. Although the primary body responsible for implementing and enforcing LEMA 2003 regulations, its activities are limited. Indeed, TAs are *de facto* principally controlling access to natural resources and enforcing LEMA 2003 stipulations, with tribal courts functioning in part to fine transgressors. Perceptions of the DFED by local TAs are generally negative, as this agency is seen only within its role in enforcement. It is also criticized for its weakness in delivering much-needed environmental education and awareness to communities on the role of the provincial government. In addition, there is widespread criticism of the poor control of damage-causing animals by DFED and the withholding of compensation for damages caused by these animals (Anthony, 2007).

Similar to criticisms launched at the ineffectiveness of local government, weaknesses in cooperative governance between DFED and TAs are inhibiting resource conservation, leading to situations in which opportunities are established for 'gain-seekers' to exploit resources at unsustainable rates. DFED managerial staff acknowledge that discussion and co-operation regarding land use, including biodiversity conservation, between provincial and municipal governments and TAs is practically non-existent, and needs to be strengthened (Anthony, 2006). In light of the increasing pressures on natural resources and the aspirations of some communities to engage in conservation agreements with the KNP, it would be wise for these institutions to heed these trends and seek co-operative ways to halt resource overexploitation before conditions render it practically impossible to effectively pursue any community-based conservation initiatives at all.

#### **4.5 'Gain-seekers' and resource exploitation**

The Firey model contends that resource conservation is possible only when people share expectations that others will forego opportunistic practices threatening sustainability.

Towards Bridging Worldviews in Biodiversity Conservation:

Exploring the Tsonga Concept of *Ntumbuloko* in South Africa 17

effective in limiting the impact of external harvesters. With national political changes, however, TAs no longer have the resources to control land as they previously did and, at best, can only work in co-operation with provincial departments. Juxtaposed with the decreasing power and ability of TAs to control resource use, local and provincial government is, at present, unable to fill this institutional vacuum, especially given other

The outcome is a situation where, at least in some parts of the study area, external gainseekers have seized the opportunity to either hire locals or harvest resources themselves at convenient times so as to maximize profit and minimize risks of being caught in illegal activities. This includes sand removal, illegal commercial harvesting of trees and poaching game (Anthony, 2006). Firey posits that, in conditions where the social order begins to disintegrate, incentives to inhibit one's propensity for gainful resource processes may be removed, security will be exchanged for economic efficiency, and resource congeries in the form of calculating opportunism will become the norm. Of further concern is that this new agency, having no determinate structure, can offer little resistance to further change. Therefore, *if left unabated and where sanctions are relatively ineffective, unsustainable resource extraction will continue in these areas and may severely limit future opportunities and environments in which community-based conservation can be implemented or, in a worse scenario, will deplete natural resources from which local communities currently derive much of their livelihoods*. Moreover, this will likely have potential implications for ecological integrity, creating an 'edge effect' along the KNP boundary (Woodroffe & Ginsberg, 1998). The situation calls for returning social stability to the rural areas and the institutions that *de facto* govern resources within them. As Firey (1960, p. 238) reminds us, development that involves cultural stabilization brings about non-gainful-but-likely practices that 'insinuate themselves into people's thinking and, abetted by a stable environment, enter into behavior as elements of a resource complex…and become supports for social order, contributing to its maintenance and resisting its change.' Consequently, the solution we outline below involves working to

pressing priorities such as provision of water, sanitation and electricity.

improve management and helping it to meet the new challenges it faces.

especially if there are weak mechanisms for accountability (Ribot, 2002).

The problem of opportunistic exploitation can be resolved in our context through a number of means. Firstly, increasing capacity of provincial conservation structures to effectively enforce environmental legislation will likely lead to decreased opportunism, but will not adequately address the cultural conundrum. Resource conservation depends on the ability to obscure resource users' perception of private gain, to gratify their incentives for security in personal relationships, and to enlist the willing conformity of all resource users. Plans, including excessive coercion or rule enforcement, which do not win consent on these fronts will usually fail as they are often expensive and considered illegitimate. Indeed, by increasing powers only to municipal and provincial governments and ignoring local customs and traditions in these contexts, a reverse effect may result in which TAs and their devotees may see this as a return to the 'fences and fines' approach to conservation under Apartheid (this time outside the KNP), and further polarize themselves from government objectives (Gibson & Marks, 1995; Michaelidou et al., 2002). A second alternative, which may lead to cultural stabilization, involves devolving natural resource access and use powers to local TAs. The drawbacks here, however, are that not all TAs are considered legitimate, and may not have the required capacity to effectively handle these responsibilities (Anthony, 2006). Moreover, current and potential possibilities of corruption, misrepresentation and elitism are left unabated in devolving powers to this lower level,

Firey's predictions may indeed be materializing in South Africa. Political transformation processes have led in many cases to *de facto* open access systems with new forms of opportunism, manifested by perverse incentives for unsustainable resource extraction, especially by 'gain-seeking' outsiders (Figure 4). These are exacerbated by low capacities in the provincial government structures and fueled by the stripping of powers of legitimate TAs (Anthony et al., 2010). According to a KNP internal report, increasing rates and magnitude of *inter alia* deforestation has been observed in areas adjacent to KNP claiming that 'trucks transporting newly cut poles and wood are often observed along the roads in adjacent areas'. In its summary, this report emphasized that 'the rate at which the destruction and degeneration is taking place will render the area useless for future community-based conservation projects.'

Fig. 4. Illegally collected fuelwood (mostly *Colophospermum mopane*) confiscated by Magona Traditional Authority in August 2004. Reproduced with permission from Anthony (2006)

Concerns about increased extraction and use of fuelwood, sand and medicinal plants by 'outsiders' have been observed elsewhere in Limpopo Province (Kirkland et al., 2007; Twine et al., 2003). Similarly, there is widespread belief in our study area that new political freedoms and democracy, coupled with the disintegration of powers of TAs, imply an uncontrolled liberty in which people are allowed to access and use resources as they wish. As early as 1994, DFED staff had noted that with respect to hunting game in rural areas, '…with the current constitutional changes, many people think the old laws are no longer valid and that this is creating problems' (cited in Anthony, 2006). In addition to these misconceptions, one of the key issues in the increased exploitation of resources by external harvesters is the control of access to resources by TAs. Although believed to be imperfect by some government staff, and involving corruption by some current TA personnel, the previous permit and enforcement system under TAs was generally recognised as being

Firey's predictions may indeed be materializing in South Africa. Political transformation processes have led in many cases to *de facto* open access systems with new forms of opportunism, manifested by perverse incentives for unsustainable resource extraction, especially by 'gain-seeking' outsiders (Figure 4). These are exacerbated by low capacities in the provincial government structures and fueled by the stripping of powers of legitimate TAs (Anthony et al., 2010). According to a KNP internal report, increasing rates and magnitude of *inter alia* deforestation has been observed in areas adjacent to KNP claiming that 'trucks transporting newly cut poles and wood are often observed along the roads in adjacent areas'. In its summary, this report emphasized that 'the rate at which the destruction and degeneration is taking place will render the area useless for future

Fig. 4. Illegally collected fuelwood (mostly *Colophospermum mopane*) confiscated by Magona Traditional Authority in August 2004. Reproduced with permission from Anthony (2006) Concerns about increased extraction and use of fuelwood, sand and medicinal plants by 'outsiders' have been observed elsewhere in Limpopo Province (Kirkland et al., 2007; Twine et al., 2003). Similarly, there is widespread belief in our study area that new political freedoms and democracy, coupled with the disintegration of powers of TAs, imply an uncontrolled liberty in which people are allowed to access and use resources as they wish. As early as 1994, DFED staff had noted that with respect to hunting game in rural areas, '…with the current constitutional changes, many people think the old laws are no longer valid and that this is creating problems' (cited in Anthony, 2006). In addition to these misconceptions, one of the key issues in the increased exploitation of resources by external harvesters is the control of access to resources by TAs. Although believed to be imperfect by some government staff, and involving corruption by some current TA personnel, the previous permit and enforcement system under TAs was generally recognised as being

community-based conservation projects.'

effective in limiting the impact of external harvesters. With national political changes, however, TAs no longer have the resources to control land as they previously did and, at best, can only work in co-operation with provincial departments. Juxtaposed with the decreasing power and ability of TAs to control resource use, local and provincial government is, at present, unable to fill this institutional vacuum, especially given other pressing priorities such as provision of water, sanitation and electricity.

The outcome is a situation where, at least in some parts of the study area, external gainseekers have seized the opportunity to either hire locals or harvest resources themselves at convenient times so as to maximize profit and minimize risks of being caught in illegal activities. This includes sand removal, illegal commercial harvesting of trees and poaching game (Anthony, 2006). Firey posits that, in conditions where the social order begins to disintegrate, incentives to inhibit one's propensity for gainful resource processes may be removed, security will be exchanged for economic efficiency, and resource congeries in the form of calculating opportunism will become the norm. Of further concern is that this new agency, having no determinate structure, can offer little resistance to further change. Therefore, *if left unabated and where sanctions are relatively ineffective, unsustainable resource extraction will continue in these areas and may severely limit future opportunities and environments in which community-based conservation can be implemented or, in a worse scenario, will deplete natural resources from which local communities currently derive much of their livelihoods*. Moreover, this will likely have potential implications for ecological integrity, creating an 'edge effect' along the KNP boundary (Woodroffe & Ginsberg, 1998). The situation calls for returning social stability to the rural areas and the institutions that *de facto* govern resources within them. As Firey (1960, p. 238) reminds us, development that involves cultural stabilization brings about non-gainful-but-likely practices that 'insinuate themselves into people's thinking and, abetted by a stable environment, enter into behavior as elements of a resource complex…and become supports for social order, contributing to its maintenance and resisting its change.' Consequently, the solution we outline below involves working to improve management and helping it to meet the new challenges it faces.

The problem of opportunistic exploitation can be resolved in our context through a number of means. Firstly, increasing capacity of provincial conservation structures to effectively enforce environmental legislation will likely lead to decreased opportunism, but will not adequately address the cultural conundrum. Resource conservation depends on the ability to obscure resource users' perception of private gain, to gratify their incentives for security in personal relationships, and to enlist the willing conformity of all resource users. Plans, including excessive coercion or rule enforcement, which do not win consent on these fronts will usually fail as they are often expensive and considered illegitimate. Indeed, by increasing powers only to municipal and provincial governments and ignoring local customs and traditions in these contexts, a reverse effect may result in which TAs and their devotees may see this as a return to the 'fences and fines' approach to conservation under Apartheid (this time outside the KNP), and further polarize themselves from government objectives (Gibson & Marks, 1995; Michaelidou et al., 2002). A second alternative, which may lead to cultural stabilization, involves devolving natural resource access and use powers to local TAs. The drawbacks here, however, are that not all TAs are considered legitimate, and may not have the required capacity to effectively handle these responsibilities (Anthony, 2006). Moreover, current and potential possibilities of corruption, misrepresentation and elitism are left unabated in devolving powers to this lower level, especially if there are weak mechanisms for accountability (Ribot, 2002).

Towards Bridging Worldviews in Biodiversity Conservation:

neighbors, and for those in similar contexts elsewhere.

comments on an earlier draft of this manuscript.

constraints.

**6. Acknowledgement** 

**7. References** 

Exploring the Tsonga Concept of *Ntumbuloko* in South Africa 19

incentives for illegal exploitation of resources, especially by external forces. If left unabated, these conditions will have increasingly adverse effects on local livelihoods and are likely to jeopardize future conservation initiatives. Where this is occurring, or is imminent, improving social cohesion and circumventing unsustainable resource practices through a more co-operative and adaptive approach to resource management by relevant institutions is needed. This principle applies not only to our study area, but also to conservation agencies elsewhere which face similar challenges, especially in cases characterized by dramatic transformations in institutional responsibilities and increasing financial

Conservation agencies have a formidable task, both philosophically and practically, in attempting to understand and integrate local worldviews into their biodiversity conservation and socio-economic objectives. Interactions with local people are complex, dynamic, and driven by economic as well as socio-political forces. We offer no single remedy or solution to address conflicts in the study area, but rather a suite of possibilities that should be explored. The question remains as to whether strategies developed by KNP to effectively involve local communities will gain normative weight so that local institutions will be able to meet their biodiversity conservation and socio-economic objectives, or whether these institutions will further lose control to pressures originating from within and from external sources. This research has shed light on these complexities and it is hoped that its findings will contribute to a more stable and sustainable future for both the KNP and its

We thank the CEU Doctoral Research Support Grant for funding assistance; all Traditional Authorities in the study area for support; and the questionnaire respondents for opening their homes to us. We also thank Akua Addo-Boadu and Zebedee-Feka Njisuh for their

1UpInfo. (1996). *Country Study & Guide: South Africa* [on-line. Retrieved from

Alpert, P. (1996). Integrated Conservation and Development Projects. *BioScience* Vol.46,

Anthony, B.P. (2006). *A View from the Other Side of the Fence: Tsonga Communities and the* 

Anthony, B. (2007). The Dual Nature of Parks: Attitudes of Neighbouring Communities

Anthony, B.P., & Bellinger, E.G. (2007). Importance Value of Landscapes, Flora and Fauna to

*African Journal of Science* Vol.103, No.3-4, pp. 148-154, ISSN 0038-2353

*Kruger National Park, South Africa*. Unpublished doctoral dissertation, Central

towards Kruger National Park, South Africa. *Environmental Conservation* Vol. 34,

Tsonga Communities in the Rural Areas of Limpopo Province, South Africa. *South* 

http://www.1upinfo.com/country-guide-study/south-africa/

No.11, (December 1996), pp. 845-855, ISSN 0006-3568

No. 3, (September 2007), pp. 236-245, ISSN 0376-8929

European University, Budapest, Hungary

Instead of these more extreme alternatives, we advocate a more co-operative approach which sees provincial structures striving to work more hand-in-hand with local TAs in both communicating, and enforcing, natural resource legislation. Similarly, *defining what resources should be conserved, and how and for whom they should be managed should be based on interactive dialogue between the DFED and local communities*. This has promise for at least three reasons. First, it would promote citizen involvement, through traditional structures, in government affairs and redistributing power and resources to enable local people to participate in decisions that directly affect their lives (Luckham et al., 2000). Second, by maintaining and utilizing traditional structures, which are largely believed to be 'good' and 'preferable' by local communities, anxiety may be minimized regarding proposed changes in natural resource management (Anthony, 2006). Finally, it would be one tangible avenue through which government could effectively harmonize the institution of traditional leadership within the new system of democratic governance as laid out in the *Traditional Leadership and Governance Framework Act* No. 41 of 2003. Provincial structures in this arrangement would continue to play an overseer role especially in managing external threats (Michaelidou et al., 2002), but would allow TAs (where considered legitimate by local communities) to continue to exercise traditional resource management powers and, where feasible, decentralize enforcement to TAs coupled with corresponding capacity-building. Areas of conflict (e.g. use of specific protected species) would ideally be mutually agreed upon through interactive dialogue, based on research investigating sustainable harvesting of resources, and supported by flexible policies.

#### **5. Conclusion**

At an international level it has been recognized that natural resources cannot be managed effectively without the co-operation and participation of resource users to make laws and regulations work (Baland & Platteau, 1996). This makes managing protected areas an even more complex and dynamic undertaking than the traditional 'fences and fines' approach. This is exacerbated in contexts where socio-economic and political forces are also experiencing dramatic transformation. The core of natural resource management in South Africa's communal areas, including the use and value of resources, often lies in deeply rooted and relatively stable concepts which are unlikely to change in the near future, and are often not obvious in their alignment with western conservation principles. For any degree of long-term resource sustainability, compatibility must be sought between western concepts of nature conservation and local worldviews of the intended beneficiaries of any conservation and/or development projects. Moreover, the knowledge system of any culture, including that of western science, is not static, but '[a]ssimilation of "outside" knowledge, and synthesis and hybridisation with existing knowledge, are continuing processes' (Howes & Chambers, 1979, p. 12). For PAs wishing to engage in extending management options to neighboring communities, it is critical to both develop an ongoing understanding of, and recognize, how communities conceptualize humankind's relationship to the environment, rights to resource access and use, and resource management principles.

Another feature indicative of South Africa's emerging democracy is the disintegration of TAs in the rural areas, exacerbated by institutional non-uniformity, and minimal capacity of provincial government in enforcing environmental legislation. This has created *de facto* open access systems exemplified by escalating opportunities for gain-seeking and perverse incentives for illegal exploitation of resources, especially by external forces. If left unabated, these conditions will have increasingly adverse effects on local livelihoods and are likely to jeopardize future conservation initiatives. Where this is occurring, or is imminent, improving social cohesion and circumventing unsustainable resource practices through a more co-operative and adaptive approach to resource management by relevant institutions is needed. This principle applies not only to our study area, but also to conservation agencies elsewhere which face similar challenges, especially in cases characterized by dramatic transformations in institutional responsibilities and increasing financial constraints.

Conservation agencies have a formidable task, both philosophically and practically, in attempting to understand and integrate local worldviews into their biodiversity conservation and socio-economic objectives. Interactions with local people are complex, dynamic, and driven by economic as well as socio-political forces. We offer no single remedy or solution to address conflicts in the study area, but rather a suite of possibilities that should be explored. The question remains as to whether strategies developed by KNP to effectively involve local communities will gain normative weight so that local institutions will be able to meet their biodiversity conservation and socio-economic objectives, or whether these institutions will further lose control to pressures originating from within and from external sources. This research has shed light on these complexities and it is hoped that its findings will contribute to a more stable and sustainable future for both the KNP and its neighbors, and for those in similar contexts elsewhere.

## **6. Acknowledgement**

We thank the CEU Doctoral Research Support Grant for funding assistance; all Traditional Authorities in the study area for support; and the questionnaire respondents for opening their homes to us. We also thank Akua Addo-Boadu and Zebedee-Feka Njisuh for their comments on an earlier draft of this manuscript.

## **7. References**

18 Research in Biodiversity – Models and Applications

Instead of these more extreme alternatives, we advocate a more co-operative approach which sees provincial structures striving to work more hand-in-hand with local TAs in both communicating, and enforcing, natural resource legislation. Similarly, *defining what resources should be conserved, and how and for whom they should be managed should be based on interactive dialogue between the DFED and local communities*. This has promise for at least three reasons. First, it would promote citizen involvement, through traditional structures, in government affairs and redistributing power and resources to enable local people to participate in decisions that directly affect their lives (Luckham et al., 2000). Second, by maintaining and utilizing traditional structures, which are largely believed to be 'good' and 'preferable' by local communities, anxiety may be minimized regarding proposed changes in natural resource management (Anthony, 2006). Finally, it would be one tangible avenue through which government could effectively harmonize the institution of traditional leadership within the new system of democratic governance as laid out in the *Traditional Leadership and Governance Framework Act* No. 41 of 2003. Provincial structures in this arrangement would continue to play an overseer role especially in managing external threats (Michaelidou et al., 2002), but would allow TAs (where considered legitimate by local communities) to continue to exercise traditional resource management powers and, where feasible, decentralize enforcement to TAs coupled with corresponding capacity-building. Areas of conflict (e.g. use of specific protected species) would ideally be mutually agreed upon through interactive dialogue, based on research investigating sustainable harvesting of resources,

At an international level it has been recognized that natural resources cannot be managed effectively without the co-operation and participation of resource users to make laws and regulations work (Baland & Platteau, 1996). This makes managing protected areas an even more complex and dynamic undertaking than the traditional 'fences and fines' approach. This is exacerbated in contexts where socio-economic and political forces are also experiencing dramatic transformation. The core of natural resource management in South Africa's communal areas, including the use and value of resources, often lies in deeply rooted and relatively stable concepts which are unlikely to change in the near future, and are often not obvious in their alignment with western conservation principles. For any degree of long-term resource sustainability, compatibility must be sought between western concepts of nature conservation and local worldviews of the intended beneficiaries of any conservation and/or development projects. Moreover, the knowledge system of any culture, including that of western science, is not static, but '[a]ssimilation of "outside" knowledge, and synthesis and hybridisation with existing knowledge, are continuing processes' (Howes & Chambers, 1979, p. 12). For PAs wishing to engage in extending management options to neighboring communities, it is critical to both develop an ongoing understanding of, and recognize, how communities conceptualize humankind's relationship to the environment,

Another feature indicative of South Africa's emerging democracy is the disintegration of TAs in the rural areas, exacerbated by institutional non-uniformity, and minimal capacity of provincial government in enforcing environmental legislation. This has created *de facto* open access systems exemplified by escalating opportunities for gain-seeking and perverse

rights to resource access and use, and resource management principles.

and supported by flexible policies.

**5. Conclusion** 


Towards Bridging Worldviews in Biodiversity Conservation:

Tucson

Brighton, UK

Washington

ISBN 0415112559, London

1995), pp. 395-414, ISSN 0960-3115

Archaeology, University of Pretoria, Pretoria

Development, ISBN 1-899825-77-0, London

(June 1995), pp. 111-123, ISSN 0300-7839

Neuchatel, Switzerland, Retrieved from

Neuchatel, Switzerland, Retrieved from

*Development* Vol.16, No. 1, pp. 38-39, ISSN 0733-6608

Exploring the Tsonga Concept of *Ntumbuloko* in South Africa 21

Gibson, C.C., & Marks, S. (1995). Transforming Rural Hunters into Conservationists: An

Goodland, R. (1991). Prerequisites for Ethnic Identity and Survival. In: *Resident Peoples and* 

Hannigan, J.A. (1995). *Environmental Sociology: A Social Constructionist Perspective*. Routledge,

Happold, D.C.D. (1995). The Interaction between Humans and Mammals in Africa in

Hartman, J.B., Kriel, J.D., Boonzaaier, C.C., Els, H., & Wasserman, I. (1993). *Report on the* 

Heinen, J.T. (1996). Human Behaviour, Incentives and Protected Area Management. *Conservation Biology* Vol.10, No.2, (April 1996), pp. 681-684, ISSN 1523-1739 Howes, M., & Chambers, R. (1979). *Indigenous Technical Knowledge: Analysis, Implications and* 

Hughes, R., & Flintan, F. (2001). *Integrating Conservation and Development Experience: A Review* 

Hunter, L.M., Strife, S., & Twine, W. (2010). Environmental Perceptions of Rural South

Junod, H.A. (1912). *The Life of a South African Tribe* (Volume 1). Imprimerie Attinger Freres,

 http://www.archive.org/stream/lifeofsouthafric01junouoft#page/n7/mode/2up Junod, H.A. (1913). *The Life of a South African Tribe* (Volume 2). Imprimerie Attinger Freres,

 http://www.archive.org/stream/lifeofsouthafric02junouoft#page/n5/mode/2up Kirkland, T., Hunter, L.M., & Twine, W. (2007). 'The Bush is No More': Insights on

Kleymeyer, C.D. (1992). Cultural Energy and Grassroots Development. *Grassroots* 

*Society and Natural Resources* Vol.20, No.4, pp. 337-350, ISSN 0894-1920 Kiss, A. (ed.). (1990). *Living with Wildlife: Wildlife Resource Management with Local Participation* 

Institutional Change and Natural Resource Availability in Rural South Africa.

*in Africa*. World Bank Technical Paper 130. World Bank, ISBN 0-8213-1697-4,

*Natural Resources* Vol. 23, No.6, (June 2010), pp. 525-541, ISSN 0894-1920 Johnston, B.R. (1995). Human Rights and the Environment. *Human Ecology* Vol.23, No.2,

*Development* Vol.23, No. 6, (June 1995), pp. 941-957, ISSN 0305-750X

Assessment of Community-Based Wildlife Management Programs in Africa. *World* 

*National Parks: Social Dilemmas and Strategies in International Conservation*, eds. P.C. West & S.R. Brechin, pp. 301-311, University of Arizona Press, ISBN 0-8165-1128-4,

Relation to Conservation: A Review. *Biodiversity and Conservation* Vol.4, No. 4, (June

*Development of Tribal Authorities in Gazankulu*. Department of Anthropology and

*Issues*. IDS Bulletin 10.2. Institute of Development Bulletin, University of Sussex,

*and Bibliography of the ICDP Literature*. International Institute for Environment and

African Residents: The Complex Nature of Environmental Concern. *Society &* 


Baland, J.M., & Platteau, J.P. (1996). *Halting Degradation of Natural Resources: Is There a Role for* 

Barrett, C.B., Lee, D.R., & McPeak, J.G. (2005). Institutional Arrangements for Rural Poverty

Braack, L.E.O. (2000). *Kruger National Park* (2nd ed.). New Holland, ISBN 1859744710,

Brechin, S.R., Wilshusen, P.R., Fortwangler, C.L., & West, P.C. (2002). Beyond the Square

Bryceson, D.F. (1999). *Sub-Saharan Africa Betwixt and Between: Rural Livelihood Practices and* 

Budlender, D. (2003). The Debate about Household Headship. *Social Dynamics: A Journal of* 

Campbell, B., & Shackleton, S. (2001). The Organizational Structures for Community-Based

Carruthers, J. (1995). *The Kruger National Park: A Social and Political History*. University of

Chitlango, A.J., & Balcomb, A. (2004). Ntumbuluko, Theology and Issues of Gender in the

Eckert, B., de Beer, F.C., & Vorster, L.P. (2001). Worldviews and Decision Making: Natural

Els, H. (2002). The Role of African Rural People within their Environment. In: *Game Ranch* 

Firey, W. (1960). *Man, Mind, and Land: A Theory of Resource Use,* Free Press, ISBN

Firey, W. (1978). Environmentalism and the Rediscovery of Limits. In: *Sociocultural Change* 

Gergen, K.J. (1994). *Realities and Relationships: Soundings in Social Construction*. Harvard

*South Africa Journal of Ethnology* Vol.24, No. 3, pp. 88-98, ISSN 0379-8860 Els, H. (1994). Werknermers van die Nasionale Krugerwildtuin se waarde-oordele oor

Natal Press, ISBN 0869809156, Pietermaritzburg, South Africa

*Journal of Ethnology* Vol.17, No. 4, pp. 109-120, ISSN 0379-8860

Mozambique. *Scriptura* Vol.86, No. 2, pp. 180-189, ISSN 0254-1807

Reduction and Resource Conservation. *World Development* Vol.33, No.2, (February

Wheel: Toward a More Comprehensive Understanding of Biodiversity Conservation as Social and Political Processes. *Society and Natural Resources* Vol.15,

*Policies*. ASC Working Paper No. 43. African Studies Centre, Leiden, The

Natural Resource Management in Southern Africa. *African Studies Quarterly* Vol. 5,

Tsonga Worldview: Towards an African Woman's Theology of Liberation in

Resource Management of the Laka of Mapela in an Anthropological Perspective.

aspekte van die natuur en natuurbewaring [Value Judgements of Personnel of the Kruger National Park on Aspects of Nature and Nature Conservation]. *South Africa* 

*Management* (4th ed.), ed. J. du P. Bothma, pp. 645-661, Van Schaik, ISBN

*Since 1950,* eds*.* T.L. Smith & M. Singh, pp. 273-278. Vikas, ISBN 0706906594, New

*Rural Communities?* FAO, ISBN 0198289219, Rome, Italy

*African Studies* Vol.29, No.2, pp. 48-72, ISSN 0253-3952

2005), pp. 193-197, ISSN 0305-750X

No.1, pp. 41-64, ISSN 0894-1920

No.3., ISSN 215-2448, Retrieved from http://web.africa.ufl.edu/asq/v5/v5i3a6.htm

9780627024719, Pretoria, South Africa

University Press, ISBN 0674749316, Cambridge

9780837198347, Glencoe, Illinois

Delhi

London

Netherlands


http://www.archive.org/stream/lifeofsouthafric02junouoft#page/n5/mode/2up


Towards Bridging Worldviews in Biodiversity Conservation:

Cape Town, South Africa, 30 May – 2 June, 1998

Press, ISBN 1559633468, Washington, D.C.

Bank, Washington, D.C., 2-3 April 1992

Statistics South Africa (2003). *Census 2001*. Retrieved from

http://www.statssa.gov.za/census01/html/default.asp

D.C.

Washington, D.C.

Diego, CA

Africa.

*Knowledge Systems* Vol.2, No.1, pp. 1-9, ISSN 1683-0296

Exploring the Tsonga Concept of *Ntumbuloko* in South Africa 23

Ntsebeza, L. (1999). *Land Tenure Reform in South Africa: An Example from the Eastern Cape* 

Ombe, Z.A. (2003). Indigenous Land Use Management in Lower Changane Chibuto: Sacred

Ribot, J.C. (2002). *Democratic Decentralization of Natural Resources: Institutionalizing Popular Participation*. World Resources Institute, ISBN 1-56973-522-0, Washington, D.C. Robinson, J.G., & Redford, K.H. (1994). Community-Based Approaches to Wildlife

Seragaldin, I., & Taboroff, J. (1992). *Culture and Development in Africa*. Proceedings of Culture

Smith, E.A. (2001). On the Coevolution of Cultural, Linguistic, and Biological Diversity. In:

Stevens, S. (1997). Introduction. In: *Conservation Through Cultural Survival: Indigenous Peoples* 

Sudman, S. (1983). Applied sampling. In: *Handbook of Survey Research,* eds. P.H. Rossi, J.D.

Terblanche, H.P. (1994). Geselekteerde Tegniese Skeppinge van die Tsongavrou met

Twine, W., Siphugu, V., & Moshe, D. (2003). Harvesting of Communal Resources by

Veech, J.A. (2003). Incorporating Socioeconomic Factors into the Analysis of Biodiversity

Vol.10, No. 3, (September 2003), pp. 263-274, ISSN 1350-4509

Development in sub-Saharan Africa, Berkshire, UK, 16-19 February 1999 Ntsebeza, L., & Hendricks, F.T. (1998). *Democracy and Despotism in Post-Apartheid South* 

*Province*. Paper presented at Workshop on Land Rights and Sustainable

*Africa: The Role of Chiefs/Traditional Authorities in Rural Local Government.* Paper presented at Conference on Comparing Experiences on Democratization between South Africa and Nigeria, Centre for African Studies, University of Cape Town,

and Profane Desacralisation and Recovery. *Indilinga African Journal of Indigenous* 

Conservation in Neotropical Forests. In: *Natural Connections: Perspectives in Community Based Conservation,* eds. D. Western & R.M. Wright, pp. 300-322, Island

and Development in Africa - International Conference, ISBN 0-8213-2780-1, World

*On Biocultural Diversity: Linking Language, Knowledge, and the Environment,* ed. L. Maffi, pp. 95-117, Smithsonian Institution Press, ISBN 1560989300, Washington,

*and Protected Areas,* ed. S. Stevens, pp. 1-8, Island Press, ISBN 9781559634496,

Wright, & A.B. Anderson, pp. 145-194, Academic Press, ISBN 0125982275, San

spesifieke verwysing ma die Tsongakraal opelug Museum [Selected Technical Creations of Tsonga Women with Specific Reference to the Tsonga Village Open Air Museum]. Unpublished Masters thesis, University of Pretoria, Pretoria, South

'Outsiders' in Rural South Africa: A Case of Xenophobia or a Real Threat to Sustainability? *International Journal of Sustainable Development and World Ecology* 

Hotspots. *Applied Geography* Vol.23, No.1, (January 2003), pp. 73-88, ISSN 0143-6228


Kuriyan, R. (2002). Linking Local Perceptions of Elephants and Conservation: Samburu

Limpopo Provincial Government. (2005). Department of Economic Development,

Luckham, R., Goetz, A.M., & Kaldor, M. (2000). *Democratic Institutions and Politics in Contexts* 

Mabunda, D., Pienaar, D.J., & Verhoef, J. (2003). The Kruger National Park: A Century of

Mabunda, M.M. (2001). Perceptions of Disease, Illness, and Healing among Selected Black

Maffi, L. (2004). Conservation and the Two 'Cultures' - Bridging the Gap. *Policy Matters*

Magubane, P. (1998). *Vanishing Cultures of South Africa*, Struik, ISBN 1868259676, London Manfredo, M., Teel, T., & Bright, A.D. (2004). Application of the Concepts of Values and

D. Field, & P. Brown, pp. 271-282. Modern Litho, Jefferson, Missouri Mathebula, M.R. Tribal Secretary, Homu Tribal Authority. Informal interviews. Homo

McNeely, J.A., Miller, K.R., Reid, W.V., Mittermeier, R.A., & Werner, T.B. (1990). *Conserving* 

& World Bank, ISBN 0915825422, Gland, Switzerland and Washington, D.C. Michaelidou, M., Decker, D.J., & Lassoie, J.P. (2002). The Interdependence of Ecosystem

Millar, D. (2004). *Interfacing Two Knowledge Systems: Local knowledge and Science in Africa*.

Newmark, W.D., & Hough, J.L. (2000). Conserving Wildlife in Africa: Integrated

Nietschmann, B. (1992). *The Interdependence of Biological and Cultural Diversity*. Occasional Paper No. 21. Centre for World Indigenous Studies, Olympia, Washington

Institute of Development Studies, University of Sussex, Brighton, UK Mabunda, D. (2004). *An Integrated Tourism Management Framework for the Kruger National* 

Environment and Tourism. Retrieved August 24, 2005, from http://www.limpopo.gov.za/prov\_depts/econ\_dev\_environ\_tourism.htm

957, ISSN 0894-1920

Pretoria, South Africa.

Press, ISBN 1-55963-982-2, Washington, D.C.

Vol.13, (November 2004), pp. 256-266

North, Limpopo Province, South Africa.

1920

17-20 March 2004

2000), pp. 585-592, ISSN 0006-3568

*Ethnology* Vol.24, No. 1, pp. 11-16, ISSN 0379-8860

Pastoralists in Northern Kenya. *Society and Natural Resources* Vol.15, No. 10, pp. 949-

*of Inequality, Poverty, and Conflict: A Conceptual Framework*. IDS Working Paper 104.

*Park, South Africa, 2003*. Unpublished doctoral dissertation, University of Pretoria,

Management and Research. In: *The Kruger Experience: Ecology and Management of Savanna Heterogeneity*, eds. J.T. du Toit, K.H. Rogers, & H.C. Biggs, pp. 3-21. Island

Communities in the Northern Province, South Africa. *South Africa Journal of* 

Attitudes in Human Dimensions of Natural Resources Research. In: *Society and Natural Resources: A Summary of Knowledge,* eds. M. Manfredo, J. Vaske, B. Bruyere,

*the World's Biological Diversity*. IUCN, WRI, Conservation International, WWF-US,

and Community Viability: A Theoretical Framework to Guide Research and Application. *Society and Natural Resources* Vol.15, No.7, pp. 599-616, ISSN 0894-

Proceedings of Bridging Scales and Epistemologies: Linking Local Knowledge with Global Science in Multi-Scale Assessments, ISBN 1-59726-037-1, Alexandra, Egypt,

Conservation and Development Projects and Beyond. *BioScience* Vol. 50, No.7, (July


**2** 

*China* 

**A Study on Biodiversity Mechanism** 

Yong Liu, Guolei Li, Haiqun Yu and Ruiheng Lv

*Beijing Forestry University, Beijing* 

**by the Creativity Theory of Ecosystem** 

*Key Laboratory for Silviculture and Conservation of Ministry of Education,* 

The mechanisms of biodiversity have been intensively studied in recent decades. Significant attention has been given to finding those mechanisms that explain the patterns of species richness found changing with latitudinal gradients (Hubbell, 1979; Jablonski, 2006; Lyons, 1999; Root, 1988). A large number of these species richness hypotheses have been proposed, and new ones continue to appear, with the total now exceeding thirty (Hawkins et al., 2003; Huston, 1979; Ritchie & Olff 1999). Yet there remains considerable controversy about the hypotheses that

The theories of local determinism generally try to find a few key environmental factors and establish their simple relationships with species richness in that distinct environment (Ricklefs, 2006). By doing so, the species richness could hopefully be predicted by measuring these environmental factors and their results could then become the principles of biodiversity conservation. Initially a single prominent factor is regressed against species richness, for example, the *species-energy hypothesis*, *species-area hypothesis*, or *species-productivity hypothesis* treats the single factor of energy, area, and productivity, respectively, as the most important factor to influence species richness (Allen et al., 2002; Mittelbach et al., 2001; Turner et al., 1988). Later on multiple factors are used to explain the causes of biodiversity, such as the *hypothesis of water-energy dynamics* that suggests the link between water-energy and species richness is widespread and generally strong (Hawkins et al., 2003). Ironically, more and more environmental factors are found to be important, and the relationships between these factors

The theories of community explain the forces that maintain species diversity from the aspect of community ecology, for example the *niche-assembly theory* asserts that species co-occur in a community only when they differ from one another in resource use. But this theory has some difficulties to explaining the diversity often observed in species-rich communities such as tropical forests (Zhou & Zhang, 2006). The *neutral theory*, on the other hand, assumes that all individuals of all species in a trophically similar community are ecologically equivalent. The number of species in a community is controlled by species extinction and immigration, and speciation of new species (Hubbell, 2001). Based on the fundamental processes of birth, death, dispersal and speciation, neutral theory presented a mechanism that generates species abundance distributions remarkably similar to those observed in nature, however controversy persists (McGill, 2003). Some ecologist believe that the most important task

underlie the observed patterns of biodiversity (Kerswell, 2006; Willing et al., 2003).

and species richness are variants according to different locations and scales.

**1. Introduction** 


## **A Study on Biodiversity Mechanism by the Creativity Theory of Ecosystem**

Yong Liu, Guolei Li, Haiqun Yu and Ruiheng Lv *Key Laboratory for Silviculture and Conservation of Ministry of Education, Beijing Forestry University, Beijing China* 

#### **1. Introduction**

24 Research in Biodiversity – Models and Applications

Weisberg, H.F., Krosnick, J.A., & Bowen, B.D. (1996). *An Introduction to Survey Research, Polling, and Data Analysis* (3rd ed.), Sage, ISBN 0803974027, Thousand Oaks, CA Wolpert, L. (1993). *The Unnatural Nature of Science*, Harvard University Press, ISBN

Woodroffe, R., & Ginsberg, J.R. (1998). Edge Effects and the Extinction of Populations Inside

Protected Areas. *Science* Vol.280, No.5372, (June 1998), pp. 2126-2128, ISSN 0036-

0571169724, Cambridge, MA

8075

The mechanisms of biodiversity have been intensively studied in recent decades. Significant attention has been given to finding those mechanisms that explain the patterns of species richness found changing with latitudinal gradients (Hubbell, 1979; Jablonski, 2006; Lyons, 1999; Root, 1988). A large number of these species richness hypotheses have been proposed, and new ones continue to appear, with the total now exceeding thirty (Hawkins et al., 2003; Huston, 1979; Ritchie & Olff 1999). Yet there remains considerable controversy about the hypotheses that underlie the observed patterns of biodiversity (Kerswell, 2006; Willing et al., 2003).

The theories of local determinism generally try to find a few key environmental factors and establish their simple relationships with species richness in that distinct environment (Ricklefs, 2006). By doing so, the species richness could hopefully be predicted by measuring these environmental factors and their results could then become the principles of biodiversity conservation. Initially a single prominent factor is regressed against species richness, for example, the *species-energy hypothesis*, *species-area hypothesis*, or *species-productivity hypothesis* treats the single factor of energy, area, and productivity, respectively, as the most important factor to influence species richness (Allen et al., 2002; Mittelbach et al., 2001; Turner et al., 1988). Later on multiple factors are used to explain the causes of biodiversity, such as the *hypothesis of water-energy dynamics* that suggests the link between water-energy and species richness is widespread and generally strong (Hawkins et al., 2003). Ironically, more and more environmental factors are found to be important, and the relationships between these factors and species richness are variants according to different locations and scales.

The theories of community explain the forces that maintain species diversity from the aspect of community ecology, for example the *niche-assembly theory* asserts that species co-occur in a community only when they differ from one another in resource use. But this theory has some difficulties to explaining the diversity often observed in species-rich communities such as tropical forests (Zhou & Zhang, 2006). The *neutral theory*, on the other hand, assumes that all individuals of all species in a trophically similar community are ecologically equivalent. The number of species in a community is controlled by species extinction and immigration, and speciation of new species (Hubbell, 2001). Based on the fundamental processes of birth, death, dispersal and speciation, neutral theory presented a mechanism that generates species abundance distributions remarkably similar to those observed in nature, however controversy persists (McGill, 2003). Some ecologist believe that the most important task

A Study on Biodiversity Mechanism by the Creativity Theory of Ecosystem 27

that biodiversity patterns actually are the self-creations of the global ecosystems after a long history of interactions among organisms and physical environments in different scales. It is clear that biodiversity is one of the most prominent emergent properties for the ecosystems, and creativity can naturally be employed as an indicator for the potential of yielding

1. Ecosystem creativity is a measure of an ecosystem's potential to yield emergent properties. There are three indispensable factors influencing ecosystem creativity. These

3. Environmental diversity is defined as environmental complexity of the ecosystems, which at least will include spatial heterogeneity and climatic variability. Theoretically the more heterogeneous and complex the physical environments, the more complex the plant and animal communities will be, and the higher the species diversity (Krebs, 2001). Ecological studies have also shown positive relationships between environmental complexity and species diversity for many groups of organisms, including mammals, lizards, plankton, marine gastropods, reef fish, algae and plants (Manuel, 2002). Analysis using 85 data sets ranging from plants to vertebrates and invertebrates on publications, Hawkins et al (2003) found that climatic variables were the strongest predictors of richness in 83 of the 85 cases. This finding offers widespread support for the hypothesis that climate in general has a major influence on diversity gradients across large spatial extents. It is obvious that environmental diversity has a positive

4. Adaptation is an ordinary phenomenon within the biotic kingdom and has been considered as a primary force in evolution (Zhang, 1998). Global biodiversity today can also be interpreted as the result of organism adaptation, because organisms change their material environment as well as adapt to it (Lovelock, 2003). Naturally, using adaptability as a measure of adaptation for the ecosystems should be a reasonable indicator to understand biodiversity mechanisms from the aspect of an organism's contribution. Adaptability can be defined as the quality or state of being adaptable, where adaptable means capable of being adapted or suitable without change (Gove, 1976). With this concept, it is clear that adaptability includes a continuum of states from adaptable (without change) to not adaptable (maximum change). Because creation must involve changes, and great creation means great changes, adaptability therefore is negatively related to creativity. This relationship has been demonstrated by Buckling et al. (2003) with the bacterium *Pseudomonas fluoresceus,* that adaptation itself is likely to limit a population's ability to diversify. In general, local adaptation to source habitats can limit local adaptation in sinks and restrict the use of alternate niches (Urban, 2006). Based on the above definitions we find that ecosystem creativity is positively correlated with energy (*e*) and environmental diversity (*d*), but negatively correlated with adaptability (*a*) of the ecosystems. We then introduce creativity index (*CI*) as an indicator to reflect the

include energy, environmental diversity, and adaptability within the ecosystem. 2. Energy is the most fundamental of the ecosystems and can be regarded as the capacity for doing work being associated with material bodies or for motion with systems, thus there cannot be creation without energy involved. The higher the energy input, the more emergent properties would be produced within a system. This quality has been proved by species richness distribution of birds (Hawkins, 2003; Root, 1988; Turner, 1988), Lizard (Scheibe, 1987), vascular plants (Mittelbach et al., 2001), benthic marine algae (Kerswell,

2006), etc. So energy is generally positively related to ecosystem creativity.

emergent features. Thus, the CTE is based on four main concepts:

relationship with ecosystem creativity.

creativity of ecosystems. We can write *CI* as

ahead is to integrate niche and neutral theories, that is to add more processes in neutral theories and more stochasticity in niche theories (Alonso & McKane, 2004; Chave, 2004). This demonstrates that there is a strong wish for the ecologists to search for a general principle about the mechanisms of biodiversity (Tilman, 1999). Nevertheless such integration only within the community ecology may not solve the existing problems and identify the general principle embraced. Ricklefs (2006) pointed out that to assess the relative roles of local ecological constraint compared to regional and historical unfolding of diversity-environment relationships, we must abandon localized concepts of the community and adopt historical (particularly phylogenetic) and geographic methods to evaluate the evolution of diversity within large regions and its influence on diversity at local scales.

Although many hypotheses have been proposed and ecologists have amassed a wealth of detail to explain global patterns in species richness, there is no integrated hypothesis of how the ecosystems work as a whole evolved entity. For example, local determinism explains the biodiversity gradients from a physical environment perspective, and the theories of community consider more about ecological processes or population dynamics. Moreover, it is inappropriate to run a regression of species richness against environmental factors by treating the species richness as a dependent variable and the environmental factors as independent variables. This is because from the aspect of ecosystems, both species richness and physical environments are independent variables and their interactions contribute to the properties of the ecosystems. Therefore species richness itself through interactions among species also contributes to further species richness. In addition, we all agree that biodiversity is the emergent property of ecosystems through interactions of physical environments and organisms after a long evolutionary history. In contrast, there have been few studies on such indicators to reflect the ecosystem's potential of emerging properties, which undoubtedly influences our effort to understand the mechanisms of biodiversity thoroughly. Obviously, it is necessary to establish an integrated theory to study the causes of biodiversity by embracing *system thinking*, and regarding the ecosystem as an entity and treating biodiversity as the emergent property of the ecosystems.

The objective of this paper is to study the mechanisms of biodiversity by establishing a new theory from the aspect of ecosystem creativity. Here we develop an integrated theory, which we call Creativity Theory of Ecosystems (CTE), to study the mechanisms of biodiversity with a different perspective compared to more conventional approaches. Basing our theory on system thinking, the CTE establishes a model according to four concepts of creativity, energy, environment diversity and adaptability and the relationships among them. Chiefly by introducing the adaptability as one of the independent variables, the CTE model not only integrates biotic and abiotic factors but also combines spatial and temporal scales needed to predict plant species richness. This new approach is a very general theory and can be applied to any ecosystem because it is premised on system thinking, and is not tied to any specific scale or particular experimental design. A quantitative test of CTE was also conducted with statistical methods according to data obtained for species richness and environmental factors from 27 provinces of China.

#### **2. Creativity Theory of Ecosystems (CTE)**

An ecosystem is one of the complex systems whose properties are not fully explained by an understanding of its component parts (Gallagher & Appenzeller, 1999). It is an essential approach for an ecologist to view the ecosystem as an evolved entity, by doing so we find

ahead is to integrate niche and neutral theories, that is to add more processes in neutral theories and more stochasticity in niche theories (Alonso & McKane, 2004; Chave, 2004). This demonstrates that there is a strong wish for the ecologists to search for a general principle about the mechanisms of biodiversity (Tilman, 1999). Nevertheless such integration only within the community ecology may not solve the existing problems and identify the general principle embraced. Ricklefs (2006) pointed out that to assess the relative roles of local ecological constraint compared to regional and historical unfolding of diversity-environment relationships, we must abandon localized concepts of the community and adopt historical (particularly phylogenetic) and geographic methods to evaluate the evolution of diversity within large regions and its influence on diversity at local scales. Although many hypotheses have been proposed and ecologists have amassed a wealth of detail to explain global patterns in species richness, there is no integrated hypothesis of how the ecosystems work as a whole evolved entity. For example, local determinism explains the biodiversity gradients from a physical environment perspective, and the theories of community consider more about ecological processes or population dynamics. Moreover, it is inappropriate to run a regression of species richness against environmental factors by treating the species richness as a dependent variable and the environmental factors as independent variables. This is because from the aspect of ecosystems, both species richness and physical environments are independent variables and their interactions contribute to the properties of the ecosystems. Therefore species richness itself through interactions among species also contributes to further species richness. In addition, we all agree that biodiversity is the emergent property of ecosystems through interactions of physical environments and organisms after a long evolutionary history. In contrast, there have been few studies on such indicators to reflect the ecosystem's potential of emerging properties, which undoubtedly influences our effort to understand the mechanisms of biodiversity thoroughly. Obviously, it is necessary to establish an integrated theory to study the causes of biodiversity by embracing *system thinking*, and regarding the ecosystem as an entity and

treating biodiversity as the emergent property of the ecosystems.

environmental factors from 27 provinces of China.

**2. Creativity Theory of Ecosystems (CTE)** 

The objective of this paper is to study the mechanisms of biodiversity by establishing a new theory from the aspect of ecosystem creativity. Here we develop an integrated theory, which we call Creativity Theory of Ecosystems (CTE), to study the mechanisms of biodiversity with a different perspective compared to more conventional approaches. Basing our theory on system thinking, the CTE establishes a model according to four concepts of creativity, energy, environment diversity and adaptability and the relationships among them. Chiefly by introducing the adaptability as one of the independent variables, the CTE model not only integrates biotic and abiotic factors but also combines spatial and temporal scales needed to predict plant species richness. This new approach is a very general theory and can be applied to any ecosystem because it is premised on system thinking, and is not tied to any specific scale or particular experimental design. A quantitative test of CTE was also conducted with statistical methods according to data obtained for species richness and

An ecosystem is one of the complex systems whose properties are not fully explained by an understanding of its component parts (Gallagher & Appenzeller, 1999). It is an essential approach for an ecologist to view the ecosystem as an evolved entity, by doing so we find that biodiversity patterns actually are the self-creations of the global ecosystems after a long history of interactions among organisms and physical environments in different scales. It is clear that biodiversity is one of the most prominent emergent properties for the ecosystems, and creativity can naturally be employed as an indicator for the potential of yielding emergent features. Thus, the CTE is based on four main concepts:


can limit local adaptation in sinks and restrict the use of alternate niches (Urban, 2006). Based on the above definitions we find that ecosystem creativity is positively correlated with energy (*e*) and environmental diversity (*d*), but negatively correlated with adaptability (*a*) of the ecosystems. We then introduce creativity index (*CI*) as an indicator to reflect the creativity of ecosystems. We can write *CI* as

A Study on Biodiversity Mechanism by the Creativity Theory of Ecosystem 29

variables because the variables of *d* and *a* are spatial and temporal dependent, respectively. Thus to calculate the *CI* value quantitatively we need to specify ecosystems in spatial and temporal scales and find the relationship among variables for certain scale ecosystems. We then test the CTE with the methods of Principal Component Analysis (PCA) and regression

China, with a vast area of 9.6 million km2, is an ideal region to test CTE. Its territorial distance from south to north is 5500km, including tropic, subtropic, temperate warm and temperate cool zones. From China's east to west is 5200km, including a great expanse of land from the Pacific Ocean to Mt. Everest with the terrain rising gradually. Due to the various geographic and climatic conditions, China is one of the regions with the most

We collected the data from 27 provinces (22 provinces and 5 autonomous regions) in China (Table 1 in Appendix 1) Three municipalities directly under the Central Government (Beijing, Tianjing, Shanghai) were excluded because their relative small areas have been strongly influenced by urbanization. The data from the Chongqing municipality was included in Sichuan province. The areas of Hong Kong, Macao, and Taiwan were not

1. Animal species richness. The number of terrestrial vertebrate species (including birds, reptiles, amphibians, and mammals) was used to indicate the animal species richness. Data were taken from Editorial board for series of natural resources in China (1995) and

2. Plant species richness. The number of vascular plant species (including pteridophyte, gymnosperm, and angiosperm) was used to indicate the plant species richness. Data were taken from publications related to the flora and vegetation for every selected

3. Energy (*e*). Energy input was estimated by annual mean temperature. Data were available online from the Scientific Database of the Chinese Academy of Science

4. Environmental diversity (*d*). We used annual precipitation, altitude difference (maximum minus minimum altitudes), and land area to estimate environmental diversity. Water availability is a critical factor to constrain species distribution and altitude difference is the most prominent geographic feature in China. Area is also employed as a variable to influence environmental diversity because the areas were greatly different among provinces. The combination of these three factors can predict the *d* variable in regional scale very well. Data were taken from the publication Editorial board for series of natural resources in China (1995) and Editorial board for the complete series of Chinese agriculture (1998) and online from

5. Reciprocal of adaptability (*a-1*). According to Eq. (2): *a-1 = 1/a = 1/ n/(N-n) = (N–n) / n*, this actually is the changed rate of ecosystem properties during evolutionary time period △t. Because we treated every province as a consolidated entity, and it is impossible to

Editorial board for the complete series of Chinese agriculture (1998).

for ecosystems on a regional scale.

abundant displays of biodiversity in the world (Shi, 1991).

included because we were unable to get relevant data. The variables were collected and determined as follows:

province (see Appendix 1).

(http://www.sdb.ac.cn/).

(http://www.sdb.ac.cn/).

**3. Test of the CTE** 

**3.1 Material and methods** 

$$\text{CI} \equiv f \text{ ( $e$ , d, a)}\tag{1}$$

Generally speaking, in this equation energy can be easily understood and measured, however environmental diversity and adaptability are problematic. Because there is no single parameter to completely express environmental diversity, it is probably wise to consider it according to spatial scales. Freestone & Inouye (2006) found that the mechanisms driving species coexistence and diversity in serpentine seeps appear scale-dependent. Willis & Whittaker (2002) classified the spatial scales in five categories as local, landscape, regional, continental, and global. It is becoming increasingly apparent that the factor best accounting for patterns of biodiversity seems to be delimited by scale. Consequently, *d* is a variable of scale-dependent in Eq. (1).

According to the definitions, an ecosystem is completely adaptable to its environmental conditions if there are no changes occurring, but the ecosystem displays some nonadaptable quality if there are emergent properties occurring after a certain period of evolutionary time. Thus the more emergent properties that appear within the ecosystem the bigger the changes become and the lower the level of adaptability. We then can calculate the *a* value by the reciprocal of the ecosystem's changed rate with the following equation:

$$a = 1/(\text{N-n})/\text{n} = \text{n/(N-n)}\tag{2}$$

Where, *n* is the number of original properties at time t1, *N* is the number of properties at time t2, △t (△t = t2- t1) is the evolutionary time of an ecosystem. Thus, *N–n* is the number of emergent properties and *(N–n)/n* is the changed rate of ecosystem properties during evolutionary time period △t.

Apparently we measure the adaptability of the ecosystem from historical and evolutionary aspects, because in the biological sense current adaptations are the result of selection that was in progress at some time in the past (Ridley, 2004). We actually judge the adaptability from differences between current properties of the ecosystem and its properties in the past. This then implies that our *a* value is also a variable that is time-scale-dependent. In addition, though the ecological processes of competition and predation (Bush, 2003; Fine et al., 2006; Schmitz, 2006; Straub, 2006) are greatly different within evolutionary time periods, the Bible teaches us to judge a tree by its fruit. Correspondingly, we judge the perfection of an organism by its power to survive and multiply (Egbert Giles Leigb, 1971) or in our hypotheses by the final emergent properties of ecosystems.

If the *a* variable in Eq. (1) is replaced by Eq. (2), the model yields: *CI = f(e, d, a) = f[e, d, n/(N– n)]*, Since *a* is negatively correlated with creativity, the above equation should be:

$$\text{CI} \equiv f[\text{e, d, (N-n)/n}] \tag{3}$$

Hence, CTE treats the ecosystems as a consolidated entity, and biodiversity is its emergent property through interactions among organisms and environments after evolving through a certain spatial and temporal scale. For an ecosystem the higher the energy input the more diversified environments become and the less adaptability. Subsequently, the greater the creativity also means the higher the biodiversity.

However, due to the combinations among the variables of *e, d, and a* could be various forms such as plus, multiplication, power, etc., and the *CI* model (3) is only a functional equation that cannot be calculated directly. This is a reflection of complexity of the ecosystems, i.e., one cannot predict the creativity of every ecosystem with a single combination of these variables because the variables of *d* and *a* are spatial and temporal dependent, respectively. Thus to calculate the *CI* value quantitatively we need to specify ecosystems in spatial and temporal scales and find the relationship among variables for certain scale ecosystems. We then test the CTE with the methods of Principal Component Analysis (PCA) and regression for ecosystems on a regional scale.

## **3. Test of the CTE**

28 Research in Biodiversity – Models and Applications

Generally speaking, in this equation energy can be easily understood and measured, however environmental diversity and adaptability are problematic. Because there is no single parameter to completely express environmental diversity, it is probably wise to consider it according to spatial scales. Freestone & Inouye (2006) found that the mechanisms driving species coexistence and diversity in serpentine seeps appear scale-dependent. Willis & Whittaker (2002) classified the spatial scales in five categories as local, landscape, regional, continental, and global. It is becoming increasingly apparent that the factor best accounting for patterns of biodiversity seems to be delimited by scale. Consequently, *d* is a variable of

According to the definitions, an ecosystem is completely adaptable to its environmental conditions if there are no changes occurring, but the ecosystem displays some nonadaptable quality if there are emergent properties occurring after a certain period of evolutionary time. Thus the more emergent properties that appear within the ecosystem the bigger the changes become and the lower the level of adaptability. We then can calculate the *a* value by the reciprocal of the ecosystem's changed rate with the following equation:

Where, *n* is the number of original properties at time t1, *N* is the number of properties at time t2, △t (△t = t2- t1) is the evolutionary time of an ecosystem. Thus, *N–n* is the number of emergent properties and *(N–n)/n* is the changed rate of ecosystem properties during

Apparently we measure the adaptability of the ecosystem from historical and evolutionary aspects, because in the biological sense current adaptations are the result of selection that was in progress at some time in the past (Ridley, 2004). We actually judge the adaptability from differences between current properties of the ecosystem and its properties in the past. This then implies that our *a* value is also a variable that is time-scale-dependent. In addition, though the ecological processes of competition and predation (Bush, 2003; Fine et al., 2006; Schmitz, 2006; Straub, 2006) are greatly different within evolutionary time periods, the Bible teaches us to judge a tree by its fruit. Correspondingly, we judge the perfection of an organism by its power to survive and multiply (Egbert Giles Leigb, 1971) or in our

If the *a* variable in Eq. (1) is replaced by Eq. (2), the model yields: *CI = f(e, d, a) = f[e, d, n/(N–*

Hence, CTE treats the ecosystems as a consolidated entity, and biodiversity is its emergent property through interactions among organisms and environments after evolving through a certain spatial and temporal scale. For an ecosystem the higher the energy input the more diversified environments become and the less adaptability. Subsequently, the greater the

However, due to the combinations among the variables of *e, d, and a* could be various forms such as plus, multiplication, power, etc., and the *CI* model (3) is only a functional equation that cannot be calculated directly. This is a reflection of complexity of the ecosystems, i.e., one cannot predict the creativity of every ecosystem with a single combination of these

*n)]*, Since *a* is negatively correlated with creativity, the above equation should be:

scale-dependent in Eq. (1).

evolutionary time period △t.

hypotheses by the final emergent properties of ecosystems.

creativity also means the higher the biodiversity.

*CI = f (e, d, a)* (1)

*a = 1/(N–n)/n = n/(N-n)* (2)

*CI = f[e, d, (N–n)/n]* (3)

## **3.1 Material and methods**

China, with a vast area of 9.6 million km2, is an ideal region to test CTE. Its territorial distance from south to north is 5500km, including tropic, subtropic, temperate warm and temperate cool zones. From China's east to west is 5200km, including a great expanse of land from the Pacific Ocean to Mt. Everest with the terrain rising gradually. Due to the various geographic and climatic conditions, China is one of the regions with the most abundant displays of biodiversity in the world (Shi, 1991).

We collected the data from 27 provinces (22 provinces and 5 autonomous regions) in China (Table 1 in Appendix 1) Three municipalities directly under the Central Government (Beijing, Tianjing, Shanghai) were excluded because their relative small areas have been strongly influenced by urbanization. The data from the Chongqing municipality was included in Sichuan province. The areas of Hong Kong, Macao, and Taiwan were not included because we were unable to get relevant data.

The variables were collected and determined as follows:


A Study on Biodiversity Mechanism by the Creativity Theory of Ecosystem 31

with a single parameter of temperature, altitude difference, or land area are not ideal and account for 12.3%,17.6%, and 28.0% of the variation in plant species richness respectively, with less F value and P<0.05. The model with multiple parameters of temperature, precipitation, and altitude difference shows much better than that of models with a single parameter, and account for 59.3% of the variation in plant species richness (F=19.915,

The CTE successfully predicts the plant species richness distribution in provinces of China. This success comes mainly from our new system thinking approach of study. First, we treat every province as a consolidated ecosystem (though their land areas may differ greatly), and that plant species richness is one of its emergent properties through the interactions of biotic and abiotic factors. Conventional approaches using equal and small area quadrat as the basic studying unit actually divide the ecosystems into many small component parts. This approach is probably reasonable for a small scale evaluation within a community or landscape, but is inappropriate for a large spatial scale like China with its 9.6 million km2 land area, and distances exceeding 5000 km from south to north and east to west. Adding up all of the component parts is not equal to a whole ecosystem. This also explains why Willis and Whittaker (2002) concluded that variables that best account for species richness on a local spatial scale may not be the same as those accounting for

Secondly, we use an integrated factor from the results of PCA to calculate *CI*, which represents the contribution weights of every independent variable to the whole system. This approach is much better than the single variable model of temperature, precipitation, or area, because species richness is the emergent property of ecosystems through interactions of multiple factors. So a single factor model cannot explain the ecosystem property, especially on a large scale. On the other hand, though model established by multiple variables with stepwise regression fits better than a single variable model, it has a major drawback from the aspect of system thinking. This is because it selects variables only considering the correlations between dependent and independent variables, but neglects the interactions among variables. For instance, the last model in Table 1 only selects altitude difference and precipitation and rejects temperature. But we well know that nothing will happen in the real world without energy input. Thus, Hawkins et al. (2003) concluded that the interaction between water and energy provides a strong explanation for globally extensive plant and animal diversity gradients. Those analyses that do not include waterenergy variables are missing a key component for explaining the broad-scale patterns of diversity. In our theory we not only consider correlation, but also pay great attention to the

Finally, and most importantly, we introduce adaptability into the independent variables. This is estimated by the changed rate of animal species richness in an evolutionary time scale. In this way, we not only integrate biotic and abiotic factors but also combine spatial and temporal scales to predict plant species richness. This is a brand new approach when compared to the conventional hypotheses. Our results show that the changed rate of animal species richness has a great influence on plant species richness. Not only do theoretical works support this, but also empirical studies at the population and community levels have

indispensable components of the emergent properties of ecosystems.

*P*<0.001), and is far behind the CI model.

richness at regional spatial scales.

**3.3 Discussion** 

measure the complete properties of the ecosystems, we used the animal species richness at present-day (*N*) to estimate the ecosystem properties. We then assumed that the number of original properties (animal species richness) at the processes occurring was 1, that is *n* = 1. The evolutionary time period (△t) might be over the last 10000 years, i.e., since the end of the last glacial period (Willis 2002).

*CI* values for every province were calculated from Eq. (3) by the following steps and methods: First, the data (from Table 1 in Appendix 1) were normalized using the standard deviation method (Xu, 2002) in order to eliminate influences caused by different units and dimensions (Table 2 in Appendix 1). Second, a Principal Component Analysis (PCA) was performed on the *e*, *d* and *a-1* in an effort to reduce the dimensionality of the data sets. The varimax rotation was used to simplify the interpretation of the results. Two components accounted for 87.63% of variance in *e*, *d* and *a-1* depending on the eigenvalues, percent of variance, and cumulative percent (Table 3 in Appendix 1). The major components of the first factor are temperature, precipitation, and the reciprocal of adaptability (Table 4 in Appendix 1). This is consistent with the natural situation in China where there are two most prominent climatic features of temperature increasing from north to south and precipitation decreasing from east to west. Factor 2 is composed of altitude difference and area (Table 4 in Appendix 1), each of which reflects the geographic characteristics in China. From the rotated component matrix (Table 4 in Appendix 1) and component score coefficient matrix (Table 5 in Appendix 1), we find the component score coefficient rotated matrix for all the provinces, which were F1 and F2 (Table 6 in Appendix 1). Finally, the *CI* value (Table 6 in Appendix 1) was estimated by the integrated factor (∑F) according to the percent of variance for F1 and F2:

#### *CI =*∑F = % of variance for component 1×F1 + % of variance for component 2×F2 = 0.55569×F1 + 0.32065×F2

Therefore, *CI* represents integrated levels of energy, reciprocal of adaptability, and environmental diversity in terms of annual precipitation, altitude difference, and area. *CI* value is above average if it is positive, and below average if negative.

In order to test the CTE we took the plant species richness as the emergent property of the ecosystem for each province and as the dependent variable, where CI was the independent variable. We ran regressions on the normalized data using the equation: plant species richness = a + b(*CI*) + c(*CI*)2, in linear, quadratic and cubic models. Only the best model was shown in Table 1. The results were compared with that of conventional methods, i.e., *energyhypothesis*, *spatial heterogeneity hypothesis, area hypothesis*, and *water-energy dynamic hypothesis,* with a multiple regression equation of: plant species richness = a + b (energy or area…) + c (energy or area…)2 also in linear, quadratic and cubic models. In order to be as liberal as possible in discovering patterns, relationships were considered significant if *P* < 0.05. The majority of relationships considered significant had *P* < 0.01.

#### **3.2 Results**

Regression analysis shows (Table 1) relationships that are almost exactly those predicted by the CTE hypotheses. We find the CI value is the best predictor with the cubic model and accounted for 94.0% of the variation in plant species richness (F=137.516, *P*<0.0001) , whereas model with an integrated environmental factor excluding adaptability can only account for 42.3% of the variation in plant species richness (F=20.054, *P*<0.0001). The models with a single parameter of temperature, altitude difference, or land area are not ideal and account for 12.3%,17.6%, and 28.0% of the variation in plant species richness respectively, with less F value and P<0.05. The model with multiple parameters of temperature, precipitation, and altitude difference shows much better than that of models with a single parameter, and account for 59.3% of the variation in plant species richness (F=19.915, *P*<0.001), and is far behind the CI model.

#### **3.3 Discussion**

30 Research in Biodiversity – Models and Applications

*CI* values for every province were calculated from Eq. (3) by the following steps and methods: First, the data (from Table 1 in Appendix 1) were normalized using the standard deviation method (Xu, 2002) in order to eliminate influences caused by different units and dimensions (Table 2 in Appendix 1). Second, a Principal Component Analysis (PCA) was performed on the *e*, *d* and *a-1* in an effort to reduce the dimensionality of the data sets. The varimax rotation was used to simplify the interpretation of the results. Two components accounted for 87.63% of variance in *e*, *d* and *a-1* depending on the eigenvalues, percent of variance, and cumulative percent (Table 3 in Appendix 1). The major components of the first factor are temperature, precipitation, and the reciprocal of adaptability (Table 4 in Appendix 1). This is consistent with the natural situation in China where there are two most prominent climatic features of temperature increasing from north to south and precipitation decreasing from east to west. Factor 2 is composed of altitude difference and area (Table 4 in Appendix 1), each of which reflects the geographic characteristics in China. From the rotated component matrix (Table 4 in Appendix 1) and component score coefficient matrix (Table 5 in Appendix 1), we find the component score coefficient rotated matrix for all the provinces, which were F1 and F2 (Table 6 in Appendix 1). Finally, the *CI* value (Table 6 in Appendix 1) was estimated by the integrated factor (∑F) according to the percent of variance for F1 and

*CI =*∑F = % of variance for component 1×F1 + % of variance for component 2×F2 = 0.55569×F1 + 0.32065×F2 Therefore, *CI* represents integrated levels of energy, reciprocal of adaptability, and environmental diversity in terms of annual precipitation, altitude difference, and area. *CI*

In order to test the CTE we took the plant species richness as the emergent property of the ecosystem for each province and as the dependent variable, where CI was the independent variable. We ran regressions on the normalized data using the equation: plant species richness = a + b(*CI*) + c(*CI*)2, in linear, quadratic and cubic models. Only the best model was shown in Table 1. The results were compared with that of conventional methods, i.e., *energyhypothesis*, *spatial heterogeneity hypothesis, area hypothesis*, and *water-energy dynamic hypothesis,* with a multiple regression equation of: plant species richness = a + b (energy or area…) + c (energy or area…)2 also in linear, quadratic and cubic models. In order to be as liberal as possible in discovering patterns, relationships were considered significant if *P* < 0.05. The

Regression analysis shows (Table 1) relationships that are almost exactly those predicted by the CTE hypotheses. We find the CI value is the best predictor with the cubic model and accounted for 94.0% of the variation in plant species richness (F=137.516, *P*<0.0001) , whereas model with an integrated environmental factor excluding adaptability can only account for 42.3% of the variation in plant species richness (F=20.054, *P*<0.0001). The models

value is above average if it is positive, and below average if negative.

majority of relationships considered significant had *P* < 0.01.

since the end of the last glacial period (Willis 2002).

F2:

**3.2 Results** 

measure the complete properties of the ecosystems, we used the animal species richness at present-day (*N*) to estimate the ecosystem properties. We then assumed that the number of original properties (animal species richness) at the processes occurring was 1, that is *n* = 1. The evolutionary time period (△t) might be over the last 10000 years, i.e.,

> The CTE successfully predicts the plant species richness distribution in provinces of China. This success comes mainly from our new system thinking approach of study. First, we treat every province as a consolidated ecosystem (though their land areas may differ greatly), and that plant species richness is one of its emergent properties through the interactions of biotic and abiotic factors. Conventional approaches using equal and small area quadrat as the basic studying unit actually divide the ecosystems into many small component parts. This approach is probably reasonable for a small scale evaluation within a community or landscape, but is inappropriate for a large spatial scale like China with its 9.6 million km2 land area, and distances exceeding 5000 km from south to north and east to west. Adding up all of the component parts is not equal to a whole ecosystem. This also explains why Willis and Whittaker (2002) concluded that variables that best account for species richness on a local spatial scale may not be the same as those accounting for richness at regional spatial scales.

> Secondly, we use an integrated factor from the results of PCA to calculate *CI*, which represents the contribution weights of every independent variable to the whole system. This approach is much better than the single variable model of temperature, precipitation, or area, because species richness is the emergent property of ecosystems through interactions of multiple factors. So a single factor model cannot explain the ecosystem property, especially on a large scale. On the other hand, though model established by multiple variables with stepwise regression fits better than a single variable model, it has a major drawback from the aspect of system thinking. This is because it selects variables only considering the correlations between dependent and independent variables, but neglects the interactions among variables. For instance, the last model in Table 1 only selects altitude difference and precipitation and rejects temperature. But we well know that nothing will happen in the real world without energy input. Thus, Hawkins et al. (2003) concluded that the interaction between water and energy provides a strong explanation for globally extensive plant and animal diversity gradients. Those analyses that do not include waterenergy variables are missing a key component for explaining the broad-scale patterns of diversity. In our theory we not only consider correlation, but also pay great attention to the indispensable components of the emergent properties of ecosystems.

> Finally, and most importantly, we introduce adaptability into the independent variables. This is estimated by the changed rate of animal species richness in an evolutionary time scale. In this way, we not only integrate biotic and abiotic factors but also combine spatial and temporal scales to predict plant species richness. This is a brand new approach when compared to the conventional hypotheses. Our results show that the changed rate of animal species richness has a great influence on plant species richness. Not only do theoretical works support this, but also empirical studies at the population and community levels have

A Study on Biodiversity Mechanism by the Creativity Theory of Ecosystem 33

richness 0.0648 0.4557 0.3886 0.3962 0.9239

Table 2. Partial Correlation Coefficients between plant species richness and area, altitude

In addition, we find an interesting phenomenon that the geographical area has a very weak partial correlation coefficient with plant species richness, and the best model is an inverse one. This finding is counter to the *area hypothesis* (Table 2, Table 1). We postulate that the *area hypothesis* mainly considers a small scale, and that the pattern of species richness increasing with area will not exist if the area exceeds a critical size. Lyons & Willing (1999) also found that area effects on species richness for bats and marsupials are a minor importance at the area scales of 1000-25000km2. We believe that the smallest area in our study (33900km2 Hainan Province) may have been sufficiently large to have sampled most taxa as a regional

The mechanism of biodiversity is a complex issue that needs additional study from both system-specific models, and a more general theoretical framework that subsumes systemspecific models as special cases (Fox, 2006). By embracing system thinking and regarding an ecosystem as an entity, and by treating biodiversity as the emergent property of the ecosystem, the Creativity Theory of Ecosystems integrates biotic and abiotic factors and combines spatial and temporal scales into a single model. Among the three variables of the model, the introduction of an adaptability variable is a unique and most important innovation. This enables our model to embrace both biotic and temporal factors. Thus we believe that the CTE provides a new approach to the study of the mechanisms of biodiversity from the aspect of a general theoretical framework. In addition, using the methods of PCA, the CI can be quantitatively calculated and will successfully predict plant species richness distribution on a regional scale within China. This demonstrates that the

We thank Z. H. Li for providing climatic data assistance Q. Q. Zhang for data collecting and analyzing assistance, and Mr. Richard R. Faltonson for editing in English. This work was supported by the National Natural Science Foundation of China (NSFC) grant 30972353 and by the Doctoral Discipline Special Foundation of High Educational Universities in China

*P* 0.748 0.017 0.045 0.041 0.000

Annual precipitation Annual Mean air temperature

Animal species richness

difference

difference, precipitation, temperature, and animal species richness

Creativity Theory of Ecosystems is feasible and promising.

Area Altitude

Plant species

species pool within China.

**5. Acknowledgements** 

grant 20090014110011.

**4. Conclusions** 

documented that herbivores can reduce a plants' potential distribution, restricting them to a subset of habitats that they might physiologically tolerate (Harley, 2003). More studies demonstrated that higher trophic-levels can have important effects on plant diversity and ecosystem properties (Fine, 2006; Schmitz, 2006).


Table 1. Regression analysis of plant species richness against *CI* value and other environmental variables


Table 2. Partial Correlation Coefficients between plant species richness and area, altitude difference, precipitation, temperature, and animal species richness

In addition, we find an interesting phenomenon that the geographical area has a very weak partial correlation coefficient with plant species richness, and the best model is an inverse one. This finding is counter to the *area hypothesis* (Table 2, Table 1). We postulate that the *area hypothesis* mainly considers a small scale, and that the pattern of species richness increasing with area will not exist if the area exceeds a critical size. Lyons & Willing (1999) also found that area effects on species richness for bats and marsupials are a minor importance at the area scales of 1000-25000km2. We believe that the smallest area in our study (33900km2 Hainan Province) may have been sufficiently large to have sampled most taxa as a regional species pool within China.

## **4. Conclusions**

32 Research in Biodiversity – Models and Applications

documented that herbivores can reduce a plants' potential distribution, restricting them to a subset of habitats that they might physiologically tolerate (Harley, 2003). More studies demonstrated that higher trophic-levels can have important effects on plant diversity and

Table 1. Regression analysis of plant species richness against *CI* value and other

environmental variables

ecosystem properties (Fine, 2006; Schmitz, 2006).

The mechanism of biodiversity is a complex issue that needs additional study from both system-specific models, and a more general theoretical framework that subsumes systemspecific models as special cases (Fox, 2006). By embracing system thinking and regarding an ecosystem as an entity, and by treating biodiversity as the emergent property of the ecosystem, the Creativity Theory of Ecosystems integrates biotic and abiotic factors and combines spatial and temporal scales into a single model. Among the three variables of the model, the introduction of an adaptability variable is a unique and most important innovation. This enables our model to embrace both biotic and temporal factors. Thus we believe that the CTE provides a new approach to the study of the mechanisms of biodiversity from the aspect of a general theoretical framework. In addition, using the methods of PCA, the CI can be quantitatively calculated and will successfully predict plant species richness distribution on a regional scale within China. This demonstrates that the Creativity Theory of Ecosystems is feasible and promising.

## **5. Acknowledgements**

We thank Z. H. Li for providing climatic data assistance Q. Q. Zhang for data collecting and analyzing assistance, and Mr. Richard R. Faltonson for editing in English. This work was supported by the National Natural Science Foundation of China (NSFC) grant 30972353 and by the Doctoral Discipline Special Foundation of High Educational Universities in China grant 20090014110011.

A Study on Biodiversity Mechanism by the Creativity Theory of Ecosystem 35

Anhui -0.547 -0.668 0.571 0.472 -0.398 -0.267 -0.398\* Fujian -0.594 -0.530 1.367 0.991 0.860 0.109 0.860 Gansu 0.256 0.964 -1.236 -0.877 0.915 -0.084 0.915 Guangdong -0.449 -0.638 1.717 1.539 0.952 0.783 0.952 Guangxi -0.300 -0.629 1.384 1.473 1.177 0.969 1.177 Guizhou -0.454 -0.234 0.437 0.494 1.323 0.799 1.323 Hainan -0.818 -0.667 1.532 2.095 -0.278 -0.288 -0.278 Hebei -0.425 -0.179 -0.783 -0.311 -0.375 -0.534 -0.375 Henan -0.478 -0.542 -0.339 0.215 -0.889 -0.219 -0.889 Heilongjiang 0.257 -0.905 -0.782 -1.686 -0.577 -0.902 -0.577 Hubei -0.429 -0.070 0.620 0.536 -0.347 -0.037 -0.347 Hunan -0.363 -0.564 1.066 0.762 -0.200 0.107 -0.200 Jilin -0.425 -0.271 -0.530 -1.297 -0.971 -0.665 -0.971 Jiangsu -0.642 -1.261 0.237 0.446 -0.650 -0.674 -0.650 Jiangxi -0.478 -0.535 1.523 0.881 -0.416 0.053 -0.416 Liaoning -0.531 -0.914 -0.485 -0.596 -0.664 -1.074 -0.664

Mongolia 2.119 0.107 -1.253 -1.213 -0.531 -0.572 -0.531 Ningxia -0.772 -0.238 -1.262 -0.656 -1.091 -0.972 -1.091 Qinghai 0.938 0.941 -1.142 -1.818 -1.026 -0.599 -1.026 Shandong -0.503 -0.827 -0.465 0.025 -0.788 -0.983 -0.788 Shanxi -0.506 -0.179 -0.837 -0.587 -0.994 -0.582 -0.994 Shanxi(Xi'an) -0.379 0.190 -0.556 -0.115 -0.264 -0.208 -0.264 Sichuan 0.544 2.029 0.056 0.080 1.764 1.710 1.764 Tibet 2.164 1.967 -0.806 -1.366 0.497 0.487 0.497 Xinjiang 3.349 2.648 -1.557 -0.814 -0.283 -0.318 -0.283 Yunnan 0.102 1.639 0.452 0.703 3.178 3.896 3.178 Zhejiang -0.635 -0.633 1.072 0.625 0.075 0.063 0.075 *Notes:* § *a-1* = *(N–n) / n*, where N is the animal species richness, n is 1, first calculated it from table 1, then

\* The data are normalized by standard deviation method with the following equations:

*j x x x im <sup>s</sup>*

1

*ij*

1 *m ij i <sup>x</sup> <sup>m</sup> <sup>j</sup> <sup>s</sup>*<sup>=</sup>

*xij* are the original data from table 1, *xij'* are the normalized data

( 1,2,... ; j 1,2,...,n) *ij j*

1 <sup>1</sup> ( ) *m*

*i x x <sup>m</sup>* 2

*ij j*

Annual Mean temperature

Animal species richness

Plant species richness

a-1§

Annual Precipitation

Provinces (or Autonomous region)

Inner

normalized it.

where, *<sup>j</sup> x* =

Table 2. The normalized data from Table 1

Area Altitude

difference

## **6. Appendix 1**


Table 1. The data on climate, environmental diversity, and species richness for the 27 provinces in China.



*Notes:* § *a-1* = *(N–n) / n*, where N is the animal species richness, n is 1, first calculated it from table 1, then normalized it.

\* The data are normalized by standard deviation method with the following equations:

$$\mathbf{x}'\_{ij} = \frac{\mathbf{x}\_{ij} - \overline{\mathbf{x}\_j}}{s\_j} \text{(\$i = 1, 2, \dots m\$; j = 1, 2, \dots, n\$)}$$

34 Research in Biodiversity – Models and Applications

Annual Precipitation (mm)

Anhui 13.98 1860 1192.1 15.06 535 3644 Fujian 12.14 2148 1588.3 18.21 809 4709 Gansu 45.4 5258 292.48 6.88 821 4164 Guangdong 17.8 1922 1762.5 21.53 829 6621 Guangxi 23.67 1941 1596.7 21.13 878 7148 Guizhou 17.61 2763 1125.5 15.19 910 6665 Hainan 3.39 1863.1 1670.5 24.9 561 3585 Hebei 18.77 2879 518.31 10.31 540 2888 Henan 16.7 2123 739.38 13.5 428 3779 Heilongjiang 45.46 1366 518.62 1.97 496 1846 Hubei 18.59 3105.4 1216.3 15.45 546 4295 Hunan 21.17 2076 1438.4 16.82 578 4705 Jilin 18.74 2686 644.19 4.33 410 2516 Jiangsu 10.26 624.7 1025.8 14.9 480 2492 Jiangxi 16.69 2138 1665.8 17.54 531 4552 Liaoning 14.59 1348 666.54 8.58 477 1358

Mongolia 118.34 3474.4 284.13 4.84 506 2781 Ningxia 5.18 2756 279.56 8.22 384 1647 Qinghai 72.12 5210 339.37 1.17 398 2703 Shandong 15.7 1530 676.2 12.35 450 1616 Shanxi 15.6 2878 491.29 8.64 405 2751 Shanxi(Xi'an) 20.56 3647 631.29 11.5 564 3813 Sichuan 56.71 7476 935.72 12.68 1006 9249 Tibet 120.1 7348.13 506.78 3.91 730 5780 Xinjiang 166.49 8765 132.81 7.26 560 3500 Yunnan 39.4 6663.6 1133 16.46 1314 15444 Zhejiang 10.53 1933 1441.7 15.99 638 4579

Table 1. The data on climate, environmental diversity, and species richness for the 27

Annual Mean temperature (℃)

Animal species richness (No. of species)

Plant species richness (No. of species)

**6.1 Tables of basic data and steps for CI value calculation by PCA** 

Altitude difference (m)

**6. Appendix 1** 

Provinces (or Autonomous region)

Inner

provinces in China.

Area (×104Km2)

$$\text{where,}\\
\begin{aligned}
\overline{\boldsymbol{\alpha}\_{j}} &= \frac{1}{m} \sum\_{i=1}^{m} \boldsymbol{\alpha}\_{ij} \quad \boldsymbol{s}\_{j} = \sqrt{\frac{1}{m} \sum\_{i=1}^{m} (\boldsymbol{\alpha}\_{ij} - \overline{\boldsymbol{\alpha}\_{j}})^2} \end{aligned}$$

*xij* are the original data from table 1, *xij'* are the normalized data

Table 2. The normalized data from Table 1

A Study on Biodiversity Mechanism by the Creativity Theory of Ecosystem 37

Ningxia -1.10848 -0.70104 0.55569 0.32065 -0.84075973 Heilongjiang -1.23733 -0.43886 0.55569 0.32065 -0.82829237 Jilin -1.06291 -0.56709 0.55569 0.32065 -0.77248587 Shanxi -0.90525 -0.57085 0.55569 0.32065 -0.68608143 Qinghai -1.52923 0.51429 0.55569 0.32065 -0.68487073 Liaoning -0.69384 -0.81544 0.55569 0.32065 -0.64703079 Shandong -0.50111 -0.83787 0.55569 0.32065 -0.54712483 Inner Mongolia -1.29225 0.71707 0.55569 0.32065 -0.48816191 Henan -0.4076 -0.75018 0.55569 0.32065 -0.46704446 Hebei -0.53405 -0.31812 0.55569 0.32065 -0.39877142 Jiangsu -0.05018 -1.09471 0.55569 0.32065 -0.37890329 Shanxi (Xi'an) -0.31057 -0.10938 0.55569 0.32065 -0.20765334 Anhui 0.2189 -0.70929 0.55569 0.32065 -0.1057933 Hubei 0.31361 -0.37556 0.55569 0.32065 0.053846627 Gansu -0.35612 0.99807 0.55569 0.32065 0.122138823 Hunan 0.58149 -0.56129 0.55569 0.32065 0.14315054 Jiangxi 0.71499 -0.70264 0.55569 0.32065 0.172011277 Xinjiang -1.07383 2.43101 0.55569 0.32065 0.182786764 Zhejiang 0.66299 -0.57856 0.55569 0.32065 0.182901649 Tibet -0.63809 1.97223 0.55569 0.32065 0.277815317 Hainan 1.24772 -0.87933 0.55569 0.32065 0.411388362 Guizhou 0.90862 0.19321 0.55569 0.32065 0.566863834 Fujian 1.23396 -0.24656 0.55569 0.32065 0.606639768 Guangdong 1.58504 -0.25305 0.55569 0.32065 0.799650395 Guangxi 1.52001 -0.08699 0.55569 0.32065 0.816761013 Sichuan 0.87298 1.80694 0.55569 0.32065 1.064501567 Yunnan 1.84052 1.96399 0.55569 0.32065 1.652511952

*Notes:* § F1=-0.0820×area+0.0709×altitude difference + 0.371 × precipitation + 0.382 × temperature +

※ F2=0.360×area+0.474×altitude difference -0.0609 × precipitation -0.0399 × temperature + 0.391 × *a-1*  ﹡*CI=*∑F = % of variance for component 1 × F1 + % of variance for component 2 × F2 = 0.55569 × F1 +

Table 6. Component score coefficient rotated matrix and *CI* values for all the provinces

% of Variance for component 1

% of Variance for component 2

*CI*﹡

F1§ F2※

Provinces (or Autonomous region)

0.419 × *a-1* 

0.32065 × F2


*Note:* Component 1 and 2 accounted for 87.63% of variance in *e*, *d* and *a-1* depending on the initial eigenvalues in percent of variance and cumulative percent, which fulfill the requirements of cumulative percent for the two components larger than 80.00%, and of total initial eigenvalues larger than 1. So, we need only calculate the two components, F1 and F2, and they can represent the integrated level of the total variables.

Table 3. Total Variance Explained


*Note:* The major components of the first factor (F1) include temperature, precipitation, and the reciprocal of adaptability (*a-1*); F2 is composed of altitude difference and area.

#### Table 4. Rotated Component Matrix


Table 5. Component Score Coefficient Matrix


Initial Eigenvalues Extraction Sums of Squared Loadings

1 2.778 55.569 55.569 2.778 55.569 55.569 2 1.603 32.065 87.634 1.603 32.065 87.634

%

*Note:* Component 1 and 2 accounted for 87.63% of variance in *e*, *d* and *a-1* depending on the initial eigenvalues in percent of variance and cumulative percent, which fulfill the requirements of cumulative percent for the two components larger than 80.00%, and of total initial eigenvalues larger than 1. So, we need only calculate the two components, F1 and F2, and they can represent the integrated level of the

> Annual Mean Temperature 0.895 -0.305 Annual Precipitation 0.884 -0.343

Altitude difference -0.113 0.953

 Component 1 2

*a-1* 0.419 0.391

Area -0.0820 0.360

*Note:* The major components of the first factor (F1) include temperature, precipitation, and the reciprocal

Annual Mean Temperature 0.371 -0.0609 Annual Precipitation 0.382 -0.0399

Altitude difference 0.0709 0.474

*a-1* 0.730 0.578

Area -0.396 0.803

% Total % of Variance Cumulative

Component 1 2 %

Comp

Comp

total variables.

Table 3. Total Variance Explained

Table 4. Rotated Component Matrix

Table 5. Component Score Coefficient Matrix

onent Total % of Variance Cumulative

3 0.373 7.453 95.087 4 0.159 3.181 98.269 5 0.087 1.731 100.000 Rotation Sums of Squared Loadings

onent Total % of Variance Cumulative

1 2.284 45.681 45.681 2 2.098 41.953 87.634

of adaptability (*a-1*); F2 is composed of altitude difference and area.


*Notes:* § F1=-0.0820×area+0.0709×altitude difference + 0.371 × precipitation + 0.382 × temperature + 0.419 × *a-1* 

※ F2=0.360×area+0.474×altitude difference -0.0609 × precipitation -0.0399 × temperature + 0.391 × *a-1*  ﹡*CI=*∑F = % of variance for component 1 × F1 + % of variance for component 2 × F2 = 0.55569 × F1 + 0.32065 × F2

Table 6. Component score coefficient rotated matrix and *CI* values for all the provinces

A Study on Biodiversity Mechanism by the Creativity Theory of Ecosystem 39

Kunming Institute of Botany, Chinese Academy of Science. (1977, 1995, 2006). *Flora of* 

Li, S. X. (1988). *Flora of Liaoning (Vol.1)*. Liaoning Scientific and Technical Publisher, Dalian,

Ma, D. Z. & Liu, H. L. (1986, 1990). *Flora of Ningxia (Vol.1, Vol. 2)*. Ningxia People's

Ma, Z. Q. (2001). *Vegetation of Shanxi*. Scientific and Technical Publisher of China, Beijing,

Ni, H.W. & Tang, S.B. (1995). Studies on the Floristic Components and Characteristics of Heilongjiang Province, China. *Territory & Natural Resources Study* 64:57-61 Northwest Plateau Institute of Biology, Academia Sinica. (1997, 1999, 1996, 1999). *Flora of Qinghai (Vol.1-4)*. Qinghai People's Publishing House, Xining, China Qi, C. J. (1990). *Vegetation of Hunan*. Hunan Scientific and Technical Publisher, Changsha,

Scientific and Technical Committee of Fujian Province and Editorial Board for Flora of

South China Institute of Botany, Academia Sinica. (1964, 1965, 1974, 1977). *Flora of* 

South China Institute of Botany, Academia Sinica. (2000). *Flora of Guangdong (Vol.4)*.

Su, Z. Y. ; Liao, W. B. & Zhang, H. D. (1996). Studies on the Floristic Components and

The Comprehensive Scientific Expedition to the Qinghai-Xizang Plateau, Academia Sinica. (1983). *Flora of Tibet (Vol.1)*. Science Press, Beijing, China, pp. 1, 356, 406 Wang, P. S. & Wang, X. Y. (2001). *Ferns of Guizhou*. Guizhou Scientific and Technical

Wang, W.Y. & Pei, L. H. (1998). The Variety and Protection of Gymnospermae in Jiangxi,

Wuhan Institute of Botany, Academia Sinica. (2001, 2001, 2002, 2002). *Flora of Hubei (Vol.1-4)*.

Yang, C. Y. (1993). *Flora of Xinjiang (Vol.1)*. Xinjiang Science and Technology and Hygiene

Zhang, W. & Zhao, S. L. (2001). The Evolution and Sources of the Flora in Shandong

Zhang, Y. J. (1989). The phytogeographical distribution and floristic characteristics of pteridophytes of Gansu Province. *Journal of Lanzhou University* 25:87-92 Zheng, J. H. (1987). Research on the basic elements and main features of pteridophytes flora of Hubei province, China. *Journal of Wuhan Botanical Research* 5:227-233 Zhou, H.G. et al. (2004). Study on pteridophyte flora of Guangxi, China. *Guihaia* 24:311-

China. *Jiangxi Forestry Science and Technology (Suppl.)*:13-15,34

Hubei Scientific and Technical Publisher,Wuhan, China

Province. *Journal of Wuhan Botanical Research* 19:57-64

Publishing House, Wulumuqi, China, pp1-51

Guangzhou Scientific and Technical Publisher, Guangzhou, China

Fujian. (1982, 1995). *Flora of Fujian (Vol.1, Vol.6)*. Fujian Science and Technology

Characteristics of Guangxi Province, China: I. The Floristic Components. *The Journal* 

*Yunnan (Vol.1, 16, 20)*. Science Press, Beijing, China Lei, M. D. (1999). *Vegetation of Shanxi*. Science Press, Beijing, China, pp52

*Guangdong (Vol.1-4)*. Science Press, Beijing, China

*of Sun Yatsen University* 2 (Suppl.):81-86

Publishing House, Yinchuan, China

China

China, pp41

China, pp40

Publisher, Fuzhou, China

Publisher, Guiyang, China

316

#### **6.2 References to data of plant species richness for 27 Provinces of China**


Chen, H. B. (1992). *Flora of Shandong (Vol.1)*. Qingdao Publishing House, Qingdao, China,

Cooperating Group for Flora of Anhui. (1985, 1987, 1991, 1990, 1992). *Flora of Anhui (Vol.1-5)*.

Cooperating Group for Vegetation of Sichuan. (1980). *Vegetation of Sichuan*. Sichuan People's

Ding, B. Z.; Wang, S. Y.& Gao, Z. Y. (1981). *Flora of Henan (Vol.1)*. Henan People's Publishing

Editorial Board for Flora of Guizhou. (1982, 1986, 1986, 1989, 1988, 1989, 1989, 1988,

Editorial Board for Flora of Hebei. (1986, 1988, 1991). *Flora of Hebei (Vol.1-3)*. Hebei Scientific

Editorial Board for Flora of Hunan. (2004). *Flora of Hunan (Vol.1, Vol.2)*. Hunan Scientific and

Editorial Board for Flora of Inner Mongolia. (1998, 1991, 1990, 1992, 1994). *Flora of Inner Mongolia (Vol.1-5)*. Inner Mongolia People's Publishing House, Huhhot, China Editorial Board for Flora of Shanxi. (1992). *Flora of Shanxi (Vol.1)*. Scientific and Technical

Editorial Board for Flora of Zhejiang. (1993, 1992, 1993, 1993, 1989, 1993, 1993). *Flora of* 

Editorial Board for Series of Natural Resources in China. (1995). *The Series of Natural* 

Editorial Board for Series of Natural Resources in China. (1995)..*The Series of Natural* 

Editorial Board for the Complete Series of Chinese Agriculture. (1999). *The Complete Series of* 

Editorial Board for the Complete Series of Chinese Agriculture. (2001). *The Complete Series of* 

Fu. S. X. (2001). *Flora of Hubei (Vol.1)*. Hubei Scientific and Technical Publisher, Wuhan,

Guangdong Institute of Botany. (1976). *Vegetation of Guangdong*. Science Press, Beijing,

Huang, D.X. (1997). *Flora of Gansu*. Gansu Science and Technology Press, Lanzhou, China,

Institute of Botany in Jiangsu Province. (1982, 1997). *Flora of Jiangsu (vol.1-2)*. Jiangsu

People's Publishing House, Nanjing, China

*Zhejiang (Vol.1-7)*. Zhejiang Scientific and Technical Publisher, Hangzhou,

*Resources in China (Heilongjiang Volume)*. China Environmental Science Press,

*Resources in China (Xinjiang Volume)*. China Environmental Science Press, Beijing,

*Chinese Agriculture (Henan Volume)*. Chinese Agriculture Publisher, Beijing, China,

*Chinese Agriculture (Jiangxi Volume)*. Chinese Agriculture Publisher, Beijing, China,

1989). *Flora of Guizhou (Vol.1-10)*. Guizhou People's Publishing House, Guiyang,

**6.2 References to data of plant species richness for 27 Provinces of China** 

Chen, H. Y. (1964). *Flora of Hainan (Vol.1)*. Science Press, Beijing, China, pp1-222

Anhui Science and Technology Publisher, Hefei, China

Publishing House, Chengdu, China

House, Zhengzhou, China, pp1-123

and Technical Publisher, Shijiazhuang

Publisher of China, Beijing

Beijing, China, pp. 469

China, pp319

pp. 41

pp22-23

China, pp1

China, pp18

pp36-37

Technical Publisher, Changsha, China, pp1-50

pp13-151

China

China


A Study on Biodiversity Mechanism by the Creativity Theory of Ecosystem 41

Kerswell, A. P. (2006). Global biodiversity patterns of benthic marine algae. *Ecology* 87:2479-

Krebs, C.J. (2001) Ecology: *The Experimental Analysis of Distribution and Abundance (5 edition)*.

Lyons, S. K. & Willig, M. R. (1999). A hemispheric assessment of scale dependence in

Manuel, C. & Molles, J. r. (2002). *Ecology: Concepts and Applications (second edition)*, Beijing:

McGill, B.J, (2003). A test of the unified neutral theory of biodiversity. *Nature* 422:881-

Mittelbach, G. G.; Steiner, C.F. & Scheiner, S. M. et al. (2001). What is the observed relationship between species richness and productivity? *Ecology* 82:2381-2396 Ricklefs, R. E. (2006). Evolutionary diversification and the origin of the diversity-

Ridley, M. (2004). *Evolution (Second edition)*. Oxford University Press Inc., New York,

Ritchie, M. E. &Olff, H. (1999). Spatial scaling laws yield a synthetic theory of biodiversity.

Root, T. (1988). Energy constraints on avian distributions and abundances. *Ecology* 69:330-

Scheibe, J. S. (1987). Climate, competition, and the structure of temperate zone lizard

Schmitz, O. J. (2006). Predators have large effects on ecosystem properties by changing plant

Shi, K.; Deng, J. & Ding, Y. (1991). *Development of Forestry Science and Technology in China*.

Straub, C. S. & Snyder, W. E. (2006). Species identity dominates the relationship between predator biodiversity and herbivore suppression. *Ecology* 87:277-282 Tilman, D. (1999) The ecological consequences of changes in biodiversity: a search for

Turner, J. R. G.; Lennon, J. J. & Lawrenson, J. A. (1988) .British bird species distributions and

Urban, M. C. & Skelly, D. K. (2006). Evolving metacommunities: toward an evolutionary

Willig, M. R.; Kaufman, D. M. & Stevens, R. D. (2003). Latitudinal gradients of biodiversity:

Willis, K. J. & Whittaker, R. J. (2002). Species diversity – scale matters. *Science* 295:1245-

Xu, J. H. (2002). *Mathematical Methods in Contemporary Geography*. High Education Press,

pattern, process, scale, and synthesis. *Annual Review of Ecology, Evolution, and* 

latitudinal gradients of species richness. *Ecology* 80:2483-2491

environment relationship. *Ecology* 87 (Suppl.): S3-S13

diversity, not plant biomass. *Ecology* 87:1432-1437

perspective on metacommunities. *Ecology* 87:1616-1626

Zhang, Y. (1998). *Biological Evolution*. Beijing University Press, Beijing

China Science and Technology Press, Beijing

general principles. *Ecology* 80:1455-1474

the energy theory. *Nature* 335:539-541

*Systematics* 34:273-309

1248

Beijing

Higher Education Press and The McGraw-Hill Book Co. Singapore

2488

885

USA

339

*Nature* 400:527-557

communities. *Ecology* 68:1424-1436

Benjamin Cummings.

Lovelock, J. (2003). The living earth. *Nature* 426:769~770

#### **7. References**


Allen, A. P.; Brown, J. H. & Gillooly, J. F. (2002). Global biodiversity, biochemical kinetics,

Alonso, D. & McKane, A. J. (2004). Sampling Hubbell's neutral theory of biodiversity.

Buckling, A.; Wills, M. A. & Colegrave, N. et al. (2003). Adaptation limits diversification of

Bush, M. B. (2003). *Ecology of a Changing Planet (Third edition)*. Pearson Education Asia

Editorial Board for Series of Natural Resources in China. (1995). *The Series of Natural* 

Editorial Board for the Complete Series of Chinese Agriculture. (1998). *The Complete Series of* 

Egbert Giles Leigb J.r. (1971). *Adaptation and diversity: natural history and the mathematics of* 

Fine, P. V.; Miller Z.J. & Mesones, I. et al. (2006). The growth-defense trade-off and

Fox, J. W. (2006). Using the price equation to partition the effects of biodiversity loss on

Freestone, A. L. & Inouye B. D. (2006). Dispersal limitation and environmental

Gove, P. B. & the Merriam-Webster editorial staff. (1976). *Webster's Third New International* 

Harley, C. D. G. (2003). Abiotic stress and herbivory interact to set range limits across a two-

Hawkins, B. A.; Porter, E. E. & Diniz-Filho, J. A. F. (2003). Productivity and history as

Hawkins, B. A.; Field, R. & Cornell, H. V. et al. (2003) .Energy, water, and broad-scale

Hubbell, S. P. (1979). Tree dispersion, abundance, and diversity in a tropical dry forest.

Hubbell, S. P. (2001). *A Unified Neutral Theory of Biodiversity and Biogeography*. Princeton

Huston, M. A. (1979) .A general hypothesis of species diversity. *American Naturalist* 113:81-

Jablonski, D.; Roy, K. & Valentine J. W. (2006). Out of the tropics: evolutionary dynamics of

*evolution*. Freeman, Cooper & Company, San Francisco, California

Gallagher, R. & Appenzeller, T. (1999). Beyond reductionism. *Science* 284:79

geographic patterns of species richness. *Ecology* 84:3105-3117

the latitudinal diversity gradient. *Science* 314:102-106

dimensional stress gradient. *Ecology* 84:1477-1488

*Resources in China (in Chinese)*. China Environmental Science Press, Beijing,

habitat speciation by plants in Amazonian forests. *Ecology* 87 (Suppl.): S150-

heterogeneity shape scale-dependent diversity patterns in plant communities.

*Dictionary of the English Language Unabridged*. G. & C. MERRIAM COMPANY,

predictors of the latitudinal diversity gradient of terrestrial birds. *Ecology* 84:1608-

and the energetic-equivalence rule. *Science* 297:1545-1548

experimental bacterial populations. *Science* 302:2107-2109

Chave, J. (2004). Neutral theory and community ecology. *Ecology letters* 7:241-253

Limited and Tsinghua University Press, Beijing

*Chinese Agriculture*. China Agriculture Press, Beijing

ecosystem function. *Ecology* 87:2687-2696

**7. References** 

China

S162

*Ecology* 87:2452-2432

*Science* 203:1299-1309

University Press, Princeton, NJ

publishers

1623

101

*Ecology letters* 7:901-910


Why do some places have much higher diversity than others? Explaining patterns of species diversity on the earth is a problem of long-standing for ecologists. Numerous diversity patterns have been documented [1], but the mechanisms behind these patterns remain poorly understood. If we knew the mechanisms, surer decisions in conservation and management would be possible. Understanding can be sought in many different ways. A dominant approach in community ecology is to search for explanations through the study of species interactions [2, 3]. This approach is motivated by the competitive exclusion principle, which suggests that there are limits to how similar species can be in their ecology while coexisting with one another in a stable way [2-6]. However, the competitive exclusion principle is challenged by the existence of many highly diverse communities of seemingly similar species. A response to this challenge is neutral theory, which postulates that many coexisting species are ecologically identical in respects that matter, and do not coexist stably. Instead, they undergo random walks to extinction, with diversity replenished by speciation, and immigration of species from other areas [7]. The neutral theory has been successful in predicting some diversity patterns in nature [8], but fails in other areas [9-12], and is limited

Other responses to the diversity challenge take the competitive exclusion principle seriously, although with a broad definition of stability [5]. These responses seek to understand the various ways that species differ from one another, how these differences structure species interactions and ultimately contribute to diversity maintenance in terms of species coexistence mechanisms [3, 5, 14, 15]. Traditional approaches focus on the differences between similar species in the ways they exploit resources, with some attention to the role of natural enemies in mediating or modifying interactions. New developments emphasize that the relationships that species have with their natural enemies are potentially just as important for diversity maintenance as their relationships with resources, and can have very similar outcomes [16]. Other directions focus on structuring and variation in the physical environment, emphasizing that the abilities of similar species to coexist with one

another may have much to do with physical environmental structure [15, 17-19].

Testing diversity maintenance hypotheses in nature is a vexing challenge [3]. Data are often limited and manipulating systems experimentally for testing diversity maintenance questions poses serious difficulties. Moreover, devising definitive tests of mechanisms has been problematic [20, 21]. However, work on coexistence mechanisms in variable

in the nature of the predictions that it can produce [13].

**1. Introduction** 

Peter Chesson

*USA* 

*Department of Ecology and Evolutionary Biology,* 

*The University of Arizona, Tucson,* 

Zhou, S. R. & Zhang, D. Y. (2006). Neutral theory in community ecology. *Journal of Plant Ecology*. 30:868-877 **3** 

#### Peter Chesson

*Department of Ecology and Evolutionary Biology, The University of Arizona, Tucson, USA* 

#### **1. Introduction**

42 Research in Biodiversity – Models and Applications

Zhou, S. R. & Zhang, D. Y. (2006). Neutral theory in community ecology. *Journal of Plant* 

Why do some places have much higher diversity than others? Explaining patterns of species diversity on the earth is a problem of long-standing for ecologists. Numerous diversity patterns have been documented [1], but the mechanisms behind these patterns remain poorly understood. If we knew the mechanisms, surer decisions in conservation and management would be possible. Understanding can be sought in many different ways. A dominant approach in community ecology is to search for explanations through the study of species interactions [2, 3]. This approach is motivated by the competitive exclusion principle, which suggests that there are limits to how similar species can be in their ecology while coexisting with one another in a stable way [2-6]. However, the competitive exclusion principle is challenged by the existence of many highly diverse communities of seemingly similar species. A response to this challenge is neutral theory, which postulates that many coexisting species are ecologically identical in respects that matter, and do not coexist stably. Instead, they undergo random walks to extinction, with diversity replenished by speciation, and immigration of species from other areas [7]. The neutral theory has been successful in predicting some diversity patterns in nature [8], but fails in other areas [9-12], and is limited in the nature of the predictions that it can produce [13].

Other responses to the diversity challenge take the competitive exclusion principle seriously, although with a broad definition of stability [5]. These responses seek to understand the various ways that species differ from one another, how these differences structure species interactions and ultimately contribute to diversity maintenance in terms of species coexistence mechanisms [3, 5, 14, 15]. Traditional approaches focus on the differences between similar species in the ways they exploit resources, with some attention to the role of natural enemies in mediating or modifying interactions. New developments emphasize that the relationships that species have with their natural enemies are potentially just as important for diversity maintenance as their relationships with resources, and can have very similar outcomes [16]. Other directions focus on structuring and variation in the physical environment, emphasizing that the abilities of similar species to coexist with one another may have much to do with physical environmental structure [15, 17-19].

Testing diversity maintenance hypotheses in nature is a vexing challenge [3]. Data are often limited and manipulating systems experimentally for testing diversity maintenance questions poses serious difficulties. Moreover, devising definitive tests of mechanisms has been problematic [20, 21]. However, work on coexistence mechanisms in variable

*c Rv*

1

*l R R*

Sensitivity

*R jl l l kl R j k*

2

*j jl l l l l*

*c vK c*

*r*

*s c vK r*

*l l*

1

<sup>1</sup> <sup>1</sup>

2

<sup>1</sup> <sup>1</sup>

Notation *Nj* : Density of (consumer) species *j*.

species *j* on species *i*

The fitness measure *κ* derived from this model involves three things (Table 1). The first is the ability of a species to harvest resources. This is the total resource intake of a species

*Rl* : Density of resource *l*.

*<sup>l</sup> v* : unit value of resource *l*.

Harvesting ability requirement

( )

*jl l l j*

*R*

*cKv*

/

*R R*

*s s*

*dN r NN*

*dN r NN*

 

> 

*jl c* : Consumption rate of resource *l* by species *j*.

*<sup>j</sup>* : resource maintenance requirement of species

*<sup>l</sup>* : Intraspecific competition for resource *l*

*ij* : Coefficient for competition for the effect of

1 11 1 12 2

2 21 1 22 2

*lr* : Maximum growth rate of resource l.

1 / *<sup>R</sup> Kl* , the carrying capacity of the resource.

<sup>1</sup> <sup>1</sup>

*dR r R Nc*

*l ll j jl <sup>j</sup> <sup>l</sup>*

*jl l l j l*

Maintenance

(consumer equation)

*j*

*dN*

*j*

*R dt*

1

*s*

1

*N dt*

2

*j*. *R*

> *R*

Table 1. MacArthur's consumer-resource equations

when its resources are at their carrying capacities.

*N dt*

*j*

*j l*

*N dt*

(resource equation)

MacArthur Consumerresource Equations

Average fitness measure

Overlap measure

Derived Lotka-Volterra Equations

environments has led to methods of quantifying the strength of coexistence mechanisms, and these quantifications have suggested definitive tests of mechanisms in nature [22]. Implementing these new methods comes with all the usual difficulties of ecological field manipulation, but the ability to focus on surer methods [20] may ultimately lead to the kind of focus in experimental technique that leads to breakthroughs.

In spite of the difficulty of understanding which coexistence mechanisms are active in a given system, enough commonalities exist between different mechanisms of stable coexistence to allow some general advice for conservation and management. Although we are a long way from truly understanding diversity maintenance, we have enough knowledge to suggest a number of areas where caution is needed. Species should not be managed in isolation. Factors improving the situation for one set of species may degrade it for others. Maintaining trophic structure, physical environmental structure, disturbance regimes, and spatial connectivity are all common sense ideas that receive support from existing understanding of diversity maintenance mechanisms.

### **2. Fitness differences, niches and coexistence**

When we consider how species interactions limit diversity, an important concept is the average fitness of a species as a whole, for this average fitness determines a species' ability to dominate other species with which it interacts. This concept is not to be confused with the fitness of a genotype or individual organism. The most common use of the idea of fitness is for genotypes or individual alleles of a gene, where its use is often relative, to predict the survival of a genotype or allele compared with others [23, 24]. Ecologists, however, often give fitness an absolute meaning as the performance of an individual organism, and this individual-level fitness is often measured as the total number of offspring that an individual leaves in its life time [25]. An alternative and more pragmatic meaning is its contribution to the population over a defined period of time through its own survival and reproduction [23]. Fitness at the species level applies in a similar relative context to fitness of a genotype or allele, but in this usage fitness specifies the relative degree of adaptedness of a species compared to others having a similar way of life and living in the same area, i.e. this fitness measure is to relative other species in the same guild, living in the same area [20].

When this species-level fitness measure is applied to problems of species coexistence, it assumes that the area of land in question is large enough for populations to be closed on an ecological timescale [5]. Then the fitness measure, which we denote *κ*, determines the degree of adaptedness of a species to the conditions applicable on that area of land. If the species in a guild cannot coexist with one another, then it is the species with the largest value of *κ* that persists, excluding others. More generally, the *κ* values rank the species in terms of their adaptation to the environment, and in essence rank the species in terms of how secure their persistence is when interacting with the others species in the guild.

#### **2.1 The MacArthur consumer-resource model**

A program to measure average fitness measures in nature has been proposed [26], but at the present time, these measures are easier to define in models. We consider the model of MacArthur [27] where a guild of animal species consume common biological resources in a lower trophic level, as reworked by Chesson [5, 28]. This model has had a key role in the development of ideas on resource partitioning for animal guilds (Table 1). Fig. 1 diagrams the foodweb being modeled, and Table 1 specifies the equations.

environments has led to methods of quantifying the strength of coexistence mechanisms, and these quantifications have suggested definitive tests of mechanisms in nature [22]. Implementing these new methods comes with all the usual difficulties of ecological field manipulation, but the ability to focus on surer methods [20] may ultimately lead to the kind

In spite of the difficulty of understanding which coexistence mechanisms are active in a given system, enough commonalities exist between different mechanisms of stable coexistence to allow some general advice for conservation and management. Although we are a long way from truly understanding diversity maintenance, we have enough knowledge to suggest a number of areas where caution is needed. Species should not be managed in isolation. Factors improving the situation for one set of species may degrade it for others. Maintaining trophic structure, physical environmental structure, disturbance regimes, and spatial connectivity are all common sense ideas that receive support from

When we consider how species interactions limit diversity, an important concept is the average fitness of a species as a whole, for this average fitness determines a species' ability to dominate other species with which it interacts. This concept is not to be confused with the fitness of a genotype or individual organism. The most common use of the idea of fitness is for genotypes or individual alleles of a gene, where its use is often relative, to predict the survival of a genotype or allele compared with others [23, 24]. Ecologists, however, often give fitness an absolute meaning as the performance of an individual organism, and this individual-level fitness is often measured as the total number of offspring that an individual leaves in its life time [25]. An alternative and more pragmatic meaning is its contribution to the population over a defined period of time through its own survival and reproduction [23]. Fitness at the species level applies in a similar relative context to fitness of a genotype or allele, but in this usage fitness specifies the relative degree of adaptedness of a species compared to others having a similar way of life and living in the same area, i.e. this fitness

measure is to relative other species in the same guild, living in the same area [20].

persistence is when interacting with the others species in the guild.

the foodweb being modeled, and Table 1 specifies the equations.

**2.1 The MacArthur consumer-resource model** 

When this species-level fitness measure is applied to problems of species coexistence, it assumes that the area of land in question is large enough for populations to be closed on an ecological timescale [5]. Then the fitness measure, which we denote *κ*, determines the degree of adaptedness of a species to the conditions applicable on that area of land. If the species in a guild cannot coexist with one another, then it is the species with the largest value of *κ* that persists, excluding others. More generally, the *κ* values rank the species in terms of their adaptation to the environment, and in essence rank the species in terms of how secure their

A program to measure average fitness measures in nature has been proposed [26], but at the present time, these measures are easier to define in models. We consider the model of MacArthur [27] where a guild of animal species consume common biological resources in a lower trophic level, as reworked by Chesson [5, 28]. This model has had a key role in the development of ideas on resource partitioning for animal guilds (Table 1). Fig. 1 diagrams

of focus in experimental technique that leads to breakthroughs.

existing understanding of diversity maintenance mechanisms.

**2. Fitness differences, niches and coexistence** 


Table 1. MacArthur's consumer-resource equations

The fitness measure *κ* derived from this model involves three things (Table 1). The first is the ability of a species to harvest resources. This is the total resource intake of a species when its resources are at their carrying capacities.

*j ij i jj*

where the quantity *ρ* measures overlap in resource use between and species *j* [5]. Fundamentally, we see that the ratio of the fitness measures determines how much one species affects another, compared with how much it affects itself. Understandably, however, relative fitness alone does not do this. It is most important to know how much the species interact with one another also. Thus, the fitness ratio is multiplied by the overlap measure, *ρ*, which is on a scale from 0 to 1, with 1 meaning complete overlap and 0 meaning

Fig. 2. Niche relationships of two species, in terms of the consumption rates on limiting resources. (A) Partial overlap in resource use, *ρ* = 0.33. (B) Complete overlap, *ρ* = 1.

affect the overlap, but is a factor explaining the difference between the fitnesses, *κ*.

Fig. 2 shows the niche relationships of two species as specified by their consumption rate curves for both partial overlap and complete overlap, while Table 1 shows how this overlap measure is calculated. The difference in the heights of the consumption rate curves does not

Interpreting this result is simplest in the case of two species competing with one another. If expression (1) is greater than 1, then species *j* excludes *i* from the system. In terms of competition coefficients, this means that species *j* places less limitation on its own growth as it increases in abundance than it places on the growth of the species *i*. Thus, species *j* continues increasing in abundance beyond the point that the growth of species *i* becomes negative. In terms of the fitness ratio and overlap measure, if the species overlap completely in resource use, then species *j* excludes *i* if it simply has a higher fitness than species *i*. However, with less than complete overlap, this fitness ratio needs to be discounted by the degree of overlap to determine if species *j* can exclude *i*. For example, if *ρ* = 2/3, then the fitness ratio needs to be more than 3/2, rather than just more than 1. Said another way, if one species is fitter than another, it only harms that other species to the extent that they

, (1)

no overlap.

This quantity, therefore, is a maximum resource harvesting rate. Subtracted from it is the resource maintenance requirement, which is the level of resource intake needed for a per capita growth rate of exactly zero, i.e. it is the resource intake that gives an individual organism a fitness of 1, which means that that one individual is replaced by exactly one individual on average in one unit of time. The maximum resource intake, less the maintenance requirement, measures the ability of a species to meet its needs, and is called the net maximum harvesting rate. Finally, the net maximum harvesting rate is divided by a quantity *s* which measures the sensitivity of the growth rate of the species to changes in resource availability, as discussed in detail by Chesson [20]. The resulting quantity is the average fitness measure *κ* (Table 1).

Fig. 1. Foodweb in the MacArthur model. Squares are the focal guild members and circles are resources. Arrows indicate the directions of effects of one species on another, with thickness differences indicating differences in strengths of these effects. The feedback loops for guild members come from links from a square to a circle back to a square. Circular loops for resources indicate resource self-limitation.

Although this fitness measure *κ* may seem intuitively reasonable, the real test is its ability to predict the outcome of the interactions of this species with others in the same guild. The measure *κ* has this desired property because it is related to competition coefficients derived from the MacArthur consumer-resource model. These competition coefficients are denoted *ij* giving the effect of interspecific competition from species *j* onto species *i*, and *jj* for intraspecific competition within species *j*. They measure the feedback loops illustrated in Fig. 1 from one species to another, and from one species back to itself. Moreover, when resource dynamics are fast relative to consumer dynamics, e.g. because resource species have much shorter generation times, the MacArthur consumer-resource model reduces to the Lotka-Volterra competition model defined by these competition coefficients [5] — see Table 1 for the two-species case. Of most importance, the ratio of interspecific competition to intraspecific competition is related to the ratio of the *κ* values:

This quantity, therefore, is a maximum resource harvesting rate. Subtracted from it is the resource maintenance requirement, which is the level of resource intake needed for a per capita growth rate of exactly zero, i.e. it is the resource intake that gives an individual organism a fitness of 1, which means that that one individual is replaced by exactly one individual on average in one unit of time. The maximum resource intake, less the maintenance requirement, measures the ability of a species to meet its needs, and is called the net maximum harvesting rate. Finally, the net maximum harvesting rate is divided by a quantity *s* which measures the sensitivity of the growth rate of the species to changes in resource availability, as discussed in detail by Chesson [20]. The resulting quantity is the

Fig. 1. Foodweb in the MacArthur model. Squares are the focal guild members and circles are resources. Arrows indicate the directions of effects of one species on another, with thickness differences indicating differences in strengths of these effects. The feedback loops for guild members come from links from a square to a circle back to a square. Circular loops

Although this fitness measure *κ* may seem intuitively reasonable, the real test is its ability to predict the outcome of the interactions of this species with others in the same guild. The measure *κ* has this desired property because it is related to competition coefficients derived from the MacArthur consumer-resource model. These competition coefficients are denoted

intraspecific competition within species *j*. They measure the feedback loops illustrated in Fig. 1 from one species to another, and from one species back to itself. Moreover, when resource dynamics are fast relative to consumer dynamics, e.g. because resource species have much shorter generation times, the MacArthur consumer-resource model reduces to the Lotka-Volterra competition model defined by these competition coefficients [5] — see Table 1 for the two-species case. Of most importance, the ratio of interspecific competition to

*jj* for

*ij* giving the effect of interspecific competition from species *j* onto species *i*, and

average fitness measure *κ* (Table 1).

for resources indicate resource self-limitation.

intraspecific competition is related to the ratio of the *κ* values:

$$\frac{\kappa\_j}{\kappa\_i} \, \rho = \frac{\alpha\_{ij}}{\alpha\_{ji}} \, \tag{1}$$

where the quantity *ρ* measures overlap in resource use between and species *j* [5]. Fundamentally, we see that the ratio of the fitness measures determines how much one species affects another, compared with how much it affects itself. Understandably, however, relative fitness alone does not do this. It is most important to know how much the species interact with one another also. Thus, the fitness ratio is multiplied by the overlap measure, *ρ*, which is on a scale from 0 to 1, with 1 meaning complete overlap and 0 meaning no overlap.

Fig. 2. Niche relationships of two species, in terms of the consumption rates on limiting resources. (A) Partial overlap in resource use, *ρ* = 0.33. (B) Complete overlap, *ρ* = 1.

Fig. 2 shows the niche relationships of two species as specified by their consumption rate curves for both partial overlap and complete overlap, while Table 1 shows how this overlap measure is calculated. The difference in the heights of the consumption rate curves does not affect the overlap, but is a factor explaining the difference between the fitnesses, *κ*.

Interpreting this result is simplest in the case of two species competing with one another. If expression (1) is greater than 1, then species *j* excludes *i* from the system. In terms of competition coefficients, this means that species *j* places less limitation on its own growth as it increases in abundance than it places on the growth of the species *i*. Thus, species *j* continues increasing in abundance beyond the point that the growth of species *i* becomes negative.

In terms of the fitness ratio and overlap measure, if the species overlap completely in resource use, then species *j* excludes *i* if it simply has a higher fitness than species *i*. However, with less than complete overlap, this fitness ratio needs to be discounted by the degree of overlap to determine if species *j* can exclude *i*. For example, if *ρ* = 2/3, then the fitness ratio needs to be more than 3/2, rather than just more than 1. Said another way, if one species is fitter than another, it only harms that other species to the extent that they

degree of separation of those loops for the relative degrees of resource shortage that the two species can tolerate due to their overall adaptation to the environment, to give the

Fig. 3. Simulation showing species coexisting stably according to the lottery competition in a temporally varying environment. Note that each species shows a strong tendency to recover

When the conditions (2) above are satisfied in the MacArthur model, the result is stable coexistence in the sense that if either species is perturbed to low density, it recovers from that low density state. In general, the MacArthur model leads to a stable equilibrium point. However, by stable coexistence, a stable equilibrium is not implied. Instead, the requirement is recovery of each species from low density, which is fully compatible with fluctuating coexistence, as illustrated in Fig. 3. Indeed, Fig. 3 illustrates stable coexistence in the lottery model [31], which requires temporal fluctuations for stable coexistence to occur. i.e. it is an example of fluctuation-dependent coexistence [32], as discussed below under variation in time and space. Neutral models, on the other hand, do not allow stable coexistence. Instead, they are characterized by having average fitnesses the same for all species. Thus, all species have the same *κ* values, and the niche overlap measures *ρ* between all pairs of species are all 1. In general, the key assumption of neutral models is stronger: it is that the species identity of an individual has no bearing on how it is affected by any other individual or how it affects other individuals [33]. However, this stronger assumption is no particular concern here. With equal average fitnesses and complete niche overlap, in the MacArthur model expression (1) is exactly equal to 1, and the inequalities (2) are not satisfied. Stable coexistence does not occur, but the model is neutrally stable. For example,

in the two-species case, population densities approach the line 11 1 22 2

remain at any point on this line, until perturbed. In particular, there is no tendency for either species to increase from low density if perturbed there, and so there is no stable

 *N N* 

1 , but can

competition coefficient ratio (1).

from fluctuations to low density.

coexistence.

**2.2 Stable versus neutral coexistence** 

overlap in resource use. As we shall see below, this statement generalizes to all forms of overlap associated with interactions between the species.

In the case where neither of two competing species can exclude the other, they coexist. For two species labeled 1 and 2, this means that expression (1) must be less than 1 for (i, j) = (1, 2) and (2, 1), discounting the unlikely boundary case of equality with 1, discussed below. This condition can be rearranged to

$$
\rho < \frac{\kappa\_1}{\kappa\_2} < \frac{1}{\rho} \tag{2}
$$

and shows that overlap, *ρ*, places constraints on how different the average fitness measures for the two species can be and still allow coexistence. Simply put, the more similar the species are in resource use (the closer *ρ* is to 1), the more similar they have to be in average fitness for coexistence to occur. In opposite terms, the more different the species are in resource use, the more different they must be in average fitness before one excludes the other.

This analysis of MacArthur's model yields some important general lessons. The competitive exclusion principle holds that species that are very similar in their ecology should have difficulty coexisting. Similarity in ecology within the MacArthur model can be interpreted as meaning a *ρ* value near 1, i.e. high overlap in resource use. Difficulty in coexisting can now be interpreted as meaning that the species have to be evenly matched in average fitness if they are to coexist, i.e. if one species gets a small edge in average fitness over the other, then exclusion will occur.

These findings for the MacArthur model generalize to others [5, 29], and they highlight two different sorts of similarity between species. The first is similarity in average fitness, i.e. having *κ* values near to each other. As discussed above, the *κ* values allow ranking of overall adaptation to the environment. Almost by definition, having similar *κ* values is favorable to species coexistence. However, similarity in *κ* values is a very different concept from similarity in way of life, or similarity in niches, when niches are defined in terms of the way of life of a species [5, 30]. In the MacArthur model, similarity in way of life reduces to similarity in resource use, which then determines how much the species interact with one another. Thus, similarity in niches constrains differences in average fitness compatible with coexistence, by conditons (2).

Looked at another way, we can think of these similarities and differences as relating to average performance compared with performance under specific conditions. Species 1 and 2 in Fig. 2(A) have very different performance under specific conditions; for example, species 1 derives very little benefit from resources 13 to 17, instead gaining most benefit from resources 5 to 12 while species 2 has an opposite pattern. The average performances of the species, which depend on the heights of the curves in the figure, are nowhere near as different as their performances for most specific resources.

Differences in performance under specific conditions lead to tolerance of average performance differences. For example, in the MacArthur model, specific-condition differences give a small value of *ρ*, and thus wide tolerance in the *κ* ratio. More generally, the nature of these specific differences is very important. They must relate directly or indirectly to separation of feedback loops because it is separation of these feedback loops that makes it possible for intraspecific competition to exceed interspecific competition, the key to coexistence. With the MacArthur model, the measure *ρ* is a specific measure of the degree of separation of those feedback loops. Multiplication by the fitness ratio adjusts the

overlap in resource use. As we shall see below, this statement generalizes to all forms of

In the case where neither of two competing species can exclude the other, they coexist. For two species labeled 1 and 2, this means that expression (1) must be less than 1 for (i, j) = (1, 2) and (2, 1), discounting the unlikely boundary case of equality with 1, discussed below.

> 1 2 1

and shows that overlap, *ρ*, places constraints on how different the average fitness measures for the two species can be and still allow coexistence. Simply put, the more similar the species are in resource use (the closer *ρ* is to 1), the more similar they have to be in average fitness for coexistence to occur. In opposite terms, the more different the species are in resource use, the

This analysis of MacArthur's model yields some important general lessons. The competitive exclusion principle holds that species that are very similar in their ecology should have difficulty coexisting. Similarity in ecology within the MacArthur model can be interpreted as meaning a *ρ* value near 1, i.e. high overlap in resource use. Difficulty in coexisting can now be interpreted as meaning that the species have to be evenly matched in average fitness if they are to coexist, i.e. if one species gets a small edge in average fitness over the other,

These findings for the MacArthur model generalize to others [5, 29], and they highlight two different sorts of similarity between species. The first is similarity in average fitness, i.e. having *κ* values near to each other. As discussed above, the *κ* values allow ranking of overall adaptation to the environment. Almost by definition, having similar *κ* values is favorable to species coexistence. However, similarity in *κ* values is a very different concept from similarity in way of life, or similarity in niches, when niches are defined in terms of the way of life of a species [5, 30]. In the MacArthur model, similarity in way of life reduces to similarity in resource use, which then determines how much the species interact with one another. Thus, similarity in niches constrains differences in average fitness compatible with

Looked at another way, we can think of these similarities and differences as relating to average performance compared with performance under specific conditions. Species 1 and 2 in Fig. 2(A) have very different performance under specific conditions; for example, species 1 derives very little benefit from resources 13 to 17, instead gaining most benefit from resources 5 to 12 while species 2 has an opposite pattern. The average performances of the species, which depend on the heights of the curves in the figure, are nowhere near as

Differences in performance under specific conditions lead to tolerance of average performance differences. For example, in the MacArthur model, specific-condition differences give a small value of *ρ*, and thus wide tolerance in the *κ* ratio. More generally, the nature of these specific differences is very important. They must relate directly or indirectly to separation of feedback loops because it is separation of these feedback loops that makes it possible for intraspecific competition to exceed interspecific competition, the key to coexistence. With the MacArthur model, the measure *ρ* is a specific measure of the degree of separation of those feedback loops. Multiplication by the fitness ratio adjusts the

, (2)

more different they must be in average fitness before one excludes the other.

overlap associated with interactions between the species.

This condition can be rearranged to

then exclusion will occur.

coexistence, by conditons (2).

different as their performances for most specific resources.

degree of separation of those loops for the relative degrees of resource shortage that the two species can tolerate due to their overall adaptation to the environment, to give the competition coefficient ratio (1).

Fig. 3. Simulation showing species coexisting stably according to the lottery competition in a temporally varying environment. Note that each species shows a strong tendency to recover from fluctuations to low density.

#### **2.2 Stable versus neutral coexistence**

When the conditions (2) above are satisfied in the MacArthur model, the result is stable coexistence in the sense that if either species is perturbed to low density, it recovers from that low density state. In general, the MacArthur model leads to a stable equilibrium point. However, by stable coexistence, a stable equilibrium is not implied. Instead, the requirement is recovery of each species from low density, which is fully compatible with fluctuating coexistence, as illustrated in Fig. 3. Indeed, Fig. 3 illustrates stable coexistence in the lottery model [31], which requires temporal fluctuations for stable coexistence to occur. i.e. it is an example of fluctuation-dependent coexistence [32], as discussed below under variation in time and space. Neutral models, on the other hand, do not allow stable coexistence. Instead, they are characterized by having average fitnesses the same for all species. Thus, all species have the same *κ* values, and the niche overlap measures *ρ* between all pairs of species are all 1. In general, the key assumption of neutral models is stronger: it is that the species identity of an individual has no bearing on how it is affected by any other individual or how it affects other individuals [33]. However, this stronger assumption is no particular concern here. With equal average fitnesses and complete niche overlap, in the MacArthur model expression (1) is exactly equal to 1, and the inequalities (2) are not satisfied. Stable coexistence does not occur, but the model is neutrally stable. For example, in the two-species case, population densities approach the line 11 1 22 2 *N N* 1 , but can remain at any point on this line, until perturbed. In particular, there is no tendency for either species to increase from low density if perturbed there, and so there is no stable coexistence.

very similar effects to feedback loops through resources. As consequence Holt coined the term "apparent competition" [41]. While previously we focused on species differences based on their use of resources, equally we can focus on species differences based on their patterns of susceptibility to the various predator species of the guild in question. Thus, the idea of a

Within this expanded concept of a niche, the niche overlap measure *ρ* is extended to predator overlap too. Indeed, the MacArthur consumer-resource model extends to a consumer-resource-predator model, with *ρ* depending on both resource and predator overlap, measuring resource-sensitivity and predator-sensitivity similarities between species [16]. For example, *ρ* would be lowered if one species were more sensitive to changes in resource abundance while the other species were more sensitive to changes in predatorabundance. The extended fitness measure, *κ*, accounts for overall predation susceptibility as

species' niche can be expanded to include susceptibility to predator species [3, 5].

Fig. 4. Foodweb for the consumer-resource-predator model. As for Fig. 1, but with

interspecific and intraspecific density dependence, and thus include both resource competition and apparent competition. Of most importance, the relationship (1) between competition coefficient ratios, fitness ratios and niche overlap still holds, and the coexistence conditions (2) remain valid. Thus, no longer can competition and predation be regarded as having very different effects on species coexistence. We should not think of competition as the primary factor limiting coexistence, with predation modifying what competition does. Instead, the view emerging corresponds to Holt's [38] advice that what is possible for

*ij*, and 

*jj*, become coefficients of

well as resource harvesting ability.

predators signified by the diamond symbols.

competition is possible for predation too.

In this fuller concept of the niche, competition coefficients

The significance of neutral models is their ability to predict some patterns of diversity seen in nature, for example in very diverse communities, neutral models have successfully reproduced the observed frequencies of species with different degrees of commonness and rarity [8]. To produce any pattern at all, however, they need to go beyond the strict deterministic description of equations like those in Table 1, and generally include demographic stochasticity. They thus recognize the finiteness of population size, and include independent chance events in the lives of individuals that determine individual fates [7]. For a model like the MacArthur consumer-resource equations, this would mean that the abundances of the species would be constantly randomly perturbed up and down the line 11 1 22 2 *N N* 1 , or with more than two species, over the hyperspace <sup>1</sup> *jj j N* . Extinctions do occur in such models, and so in modern versions they include stochastic speciation and immigration to balance species losses [7].

The predictions of neutral models have often not held up [9-12], and neutral models do not give predictions about the differential effects that environmental change often has on the species in a guild [13], which are critical for conservation and management questions. On the other hand, models in which individuals of different species are not identical in their ecology can reproduce at least some of the patterns of multispecies diversity predicted by neutral models [34]. Neutral models provide a simpler mathematical route to some predictions, and in this way have highlighted some of the processes that lead to some observed patterns in nature, but these same processes are able to produce these patterns in nonneutral models as well.

#### **2.3 Multitrophic diversity maintenance**

Similar species can interact with one another by feedback loops through resources, or by feedback loops through predators. Naturally, they may also interact directly by interfering [35] with or facilitating each other's activities [36], or through intraguild predation [37]. However, the focus here is on the much misunderstood and critical area of feedback loops through predators [38]. Fig. 4 expands the foodweb model of Fig. 1 to include predators of the species in our guild of interest. Much emphasis has focused on idea that competition limits diversity of guild members, while predation modifies what competition does. For example, the keystone species idea was developed around the concept that predation on a competitive dominant would prevent competitive exclusion [39, 40].

The keystone species idea can be view directly within the competition framework above if it is assumed that the predator, in causing mortality, increases the resource maintenance requirement of the dominant species. This is a natural expectation because if a species has higher mortality, to persist it will need higher reproduction, and higher resource consumption to fuel that reproduction. The increased maintenance requirement for the dominant then decreases its *κ* value making it more comparable to that of other species, potentially enabling coexistence. For example, in the two-species case, the coexistence conditions (2) might become satisfied [6].

The above keystone species discussion focuses just on the mortality that the predators cause, and neglects the feedback loops associated with them. In causing mortality, the predator benefits and can build up in density, with the potential then of inflicting greater mortality. The feedback loops from guild members (the prey) to guild members arise because higher densities allow predator numbers to increase, increasing mortality on guild members. As Holt [38, 41] pointed out many years ago, these feedback loops through predators can have

The significance of neutral models is their ability to predict some patterns of diversity seen in nature, for example in very diverse communities, neutral models have successfully reproduced the observed frequencies of species with different degrees of commonness and rarity [8]. To produce any pattern at all, however, they need to go beyond the strict deterministic description of equations like those in Table 1, and generally include demographic stochasticity. They thus recognize the finiteness of population size, and include independent chance events in the lives of individuals that determine individual fates [7]. For a model like the MacArthur consumer-resource equations, this would mean that the abundances of the species would be constantly randomly perturbed up and down

*jj j N* . Extinctions do occur in such models, and so in modern versions they include

The predictions of neutral models have often not held up [9-12], and neutral models do not give predictions about the differential effects that environmental change often has on the species in a guild [13], which are critical for conservation and management questions. On the other hand, models in which individuals of different species are not identical in their ecology can reproduce at least some of the patterns of multispecies diversity predicted by neutral models [34]. Neutral models provide a simpler mathematical route to some predictions, and in this way have highlighted some of the processes that lead to some observed patterns in nature, but these same processes are able to produce these patterns in

Similar species can interact with one another by feedback loops through resources, or by feedback loops through predators. Naturally, they may also interact directly by interfering [35] with or facilitating each other's activities [36], or through intraguild predation [37]. However, the focus here is on the much misunderstood and critical area of feedback loops through predators [38]. Fig. 4 expands the foodweb model of Fig. 1 to include predators of the species in our guild of interest. Much emphasis has focused on idea that competition limits diversity of guild members, while predation modifies what competition does. For example, the keystone species idea was developed around the concept that predation on a

The keystone species idea can be view directly within the competition framework above if it is assumed that the predator, in causing mortality, increases the resource maintenance requirement of the dominant species. This is a natural expectation because if a species has higher mortality, to persist it will need higher reproduction, and higher resource consumption to fuel that reproduction. The increased maintenance requirement for the dominant then decreases its *κ* value making it more comparable to that of other species, potentially enabling coexistence. For example, in the two-species case, the coexistence

The above keystone species discussion focuses just on the mortality that the predators cause, and neglects the feedback loops associated with them. In causing mortality, the predator benefits and can build up in density, with the potential then of inflicting greater mortality. The feedback loops from guild members (the prey) to guild members arise because higher densities allow predator numbers to increase, increasing mortality on guild members. As Holt [38, 41] pointed out many years ago, these feedback loops through predators can have

stochastic speciation and immigration to balance species losses [7].

competitive dominant would prevent competitive exclusion [39, 40].

1 , or with more than two species, over the hyperspace

the line 11 1 22 2 

nonneutral models as well.

**2.3 Multitrophic diversity maintenance** 

conditions (2) might become satisfied [6].

<sup>1</sup> 

 *N N* 

very similar effects to feedback loops through resources. As consequence Holt coined the term "apparent competition" [41]. While previously we focused on species differences based on their use of resources, equally we can focus on species differences based on their patterns of susceptibility to the various predator species of the guild in question. Thus, the idea of a species' niche can be expanded to include susceptibility to predator species [3, 5].

Within this expanded concept of a niche, the niche overlap measure *ρ* is extended to predator overlap too. Indeed, the MacArthur consumer-resource model extends to a consumer-resource-predator model, with *ρ* depending on both resource and predator overlap, measuring resource-sensitivity and predator-sensitivity similarities between species [16]. For example, *ρ* would be lowered if one species were more sensitive to changes in resource abundance while the other species were more sensitive to changes in predatorabundance. The extended fitness measure, *κ*, accounts for overall predation susceptibility as well as resource harvesting ability.

Fig. 4. Foodweb for the consumer-resource-predator model. As for Fig. 1, but with predators signified by the diamond symbols.

In this fuller concept of the niche, competition coefficients *ij*, and *jj*, become coefficients of interspecific and intraspecific density dependence, and thus include both resource competition and apparent competition. Of most importance, the relationship (1) between competition coefficient ratios, fitness ratios and niche overlap still holds, and the coexistence conditions (2) remain valid. Thus, no longer can competition and predation be regarded as having very different effects on species coexistence. We should not think of competition as the primary factor limiting coexistence, with predation modifying what competition does. Instead, the view emerging corresponds to Holt's [38] advice that what is possible for competition is possible for predation too.

Some ideas about spatial and temporal variation, especially those under the heading of disturbance, have viewed variation in the same way that has often coloured thinking about predation: competition has been seen as limiting diversity, with variation in time and space as modifying or even nullifying what competition does [57-60]. Many of these views are influenced by the idea that variation in time and space make ecological communities "nonequilibrial," while niche ideas are presumed to depend on the concept of equilibrium

Theoretical studies converge on the conclusion that variation in time and space are best viewed as providing more opportunities for the niches of species to be distinguished, defined by the responses of species to the changing conditions they encounter [3, 5]. This is most easily understood with respect to spatial variation [50, 62-64]. If species favor different habitats, and their resources are consumed and regenerated locally in those habitats, the habitats provide separate feedback loops, just like separate resources. Separate feedback loops with respect to predators are less likely, because predators are often wider ranging than their prey and so prey densities in one habitat can affect predation in other habitats, but this case has been developed theoretically [38]. The same would be true of resources too if resources moved between habitats, or if habitats changed their status over time on a shorter time scale than resource changes, i.e. if there were rapid spatio-temporal

Niche theory for spatial and temporal variation has developed techniques for assessing the extent to which spatial and temporal variation can separate the niches of species [20, 22]. This is the concept of covariance between environment and competition and its generalization to include apparent competition [51, 52]. Such covariances assess whether density-dependent feedback loops change with the environmental conditions under consideration, and can thus separate the niches of species. Based on them, powerful techniques for definitively testing diversity maintenance hypotheses based on spatial and temporal have been proposed [20] and implemented in some cases [22, 66]. Extension of these ideas for other niche-based diversity maintenance mechanisms should be possible, and might well lead to much progress on deciding between the various diversity maintenance

Although it is not possible at the present time to say with confidence which of the many possible diversity maintenance ideas applies in any particular system, these ideas share enough features in common that some general advice is possible. Our principal example has been MacArthur's consumer-resource model, and its extension to a consumer-resourcepredator model, for which average fitness and overlap measures are easy to define. Per capita growth rates in these models are linear functions of the densities of the various species. Models similar in spirit, but having nonlinear relationships, have been studied extensively by Tilman [17, 67], especially for plants limited by nonbiological resources, and compared with others by Chase and Leibold [3]. Although these simple measures of average fitness and niche overlap are not available, the underlying concepts remain applicable [3]. Other models of resource and apparent competition in both constant and variable environments have yielded measures of fitness and niche overlap, when they have been sought explicitly [5, 29, 47, 51, 52, 68], clearly demonstrating how these ideas extend

environmental variation [65]. Similar issues apply to temporal variation [5].

[61]. Nothing could be further from the truth [3, 5].

ideas that have been proposed.

**3. Conservation and management** 

Relationships with resources and with predators can equally promote or limit diversity depending on the circumstances. Coexistence is promoted if different guild members have different relationships with their resources or different relationships with their predators, and exclusion is promoted if these relationships tend to be similar. Differences with respect to either resource sensitivity or predator susceptibility lower *ρ*, while similarities in either respect increase *ρ*. As noted above, an extra complication is that complementary relationships between species between sensitivity to resources and susceptibility to predators also lowers *ρ*, and is referred to in the literature as a competition-predation tradeoff [42].

Feedback loops through resources and through predators naturally do have some differences. For example, predators fitting the most common notion as species that hunt and kill prey species, are often larger than their prey, or of comparable size, and are normally less species rich than the prey they focus on, while the resources of their prey may be more species rich. Opportunities for niche distinctions between species due to their relationships with their predators would therefore seem less than the opportunities due their relationships with their resources. However, predatory behavior can be complex [43], and when predators have frequency-dependent behavior, a single predator has a similar effect in terms of the coexistence of prey species to several predators that do not have this frequency-dependent behavior, but are instead specialists on particular species. Moreover, interpreting "predators" as natural enemies more generally, such as diseases and parasites, makes predators appear every bit as able as resources to define distinct niches for their prey species.

The potential symmetry between resources and predators in promoting coexistence or exclusion depends also on the relative strengths of these processes. The niche overlap measure *ρ* takes these relative strengths into account. This can be seen in the formula in Table 1, where the reciprocal of the renewal rate of a resource weights its importance in the overlap measure. Slow renewal means that the measure is more strongly affected by consumption, and therefore more strongly contributes to resource competition. Similarly, in the extended overlap measure to include predation, the ability of a predator to build up in response to prey consumption weights its importance, and determines its contribution to apparent competition [16]. Of most significance, if resource separation is strong, but resource competition as a whole is weak relative to apparent competition, then that resource separation will have little effect on coexistence. Whether the species coexist depends on whether their niches are distinguished by predation. If their niches are not distinguished by predation, then they will have difficulty coexisting despite substantial separation at the level of resources. Naturally, the opposite conclusion is reached if it is resource competition that is strong relative to apparent competition.

#### **2.4 Variation in time and space**

All natural environments vary substantially in time or space in at least some major ecologically significant ways. Environmental variables and population densities nearly always vary substantially in both time and space. Empirical studies show this, but at great variance with the realities of nature, attention of theory to these facts historically was not a mainstream endeavor [44], and sadly that remains true today. Despite the absence of mainstream attention, considerable theory on the role of variation in time and space has been developed [14, 17, 19, 20, 38, 45-52]. Although the influence of this theory on empirical studies has not yet matched its potential importance, there is now a growing body of related empirical studies [53-56]

Relationships with resources and with predators can equally promote or limit diversity depending on the circumstances. Coexistence is promoted if different guild members have different relationships with their resources or different relationships with their predators, and exclusion is promoted if these relationships tend to be similar. Differences with respect to either resource sensitivity or predator susceptibility lower *ρ*, while similarities in either respect increase *ρ*. As noted above, an extra complication is that complementary relationships between species between sensitivity to resources and susceptibility to predators also lowers *ρ*, and is referred to in the literature as a competition-predation

Feedback loops through resources and through predators naturally do have some differences. For example, predators fitting the most common notion as species that hunt and kill prey species, are often larger than their prey, or of comparable size, and are normally less species rich than the prey they focus on, while the resources of their prey may be more species rich. Opportunities for niche distinctions between species due to their relationships with their predators would therefore seem less than the opportunities due their relationships with their resources. However, predatory behavior can be complex [43], and when predators have frequency-dependent behavior, a single predator has a similar effect in terms of the coexistence of prey species to several predators that do not have this frequency-dependent behavior, but are instead specialists on particular species. Moreover, interpreting "predators" as natural enemies more generally, such as diseases and parasites, makes predators appear

The potential symmetry between resources and predators in promoting coexistence or exclusion depends also on the relative strengths of these processes. The niche overlap measure *ρ* takes these relative strengths into account. This can be seen in the formula in Table 1, where the reciprocal of the renewal rate of a resource weights its importance in the overlap measure. Slow renewal means that the measure is more strongly affected by consumption, and therefore more strongly contributes to resource competition. Similarly, in the extended overlap measure to include predation, the ability of a predator to build up in response to prey consumption weights its importance, and determines its contribution to apparent competition [16]. Of most significance, if resource separation is strong, but resource competition as a whole is weak relative to apparent competition, then that resource separation will have little effect on coexistence. Whether the species coexist depends on whether their niches are distinguished by predation. If their niches are not distinguished by predation, then they will have difficulty coexisting despite substantial separation at the level of resources. Naturally, the opposite conclusion is reached if it is resource competition that

All natural environments vary substantially in time or space in at least some major ecologically significant ways. Environmental variables and population densities nearly always vary substantially in both time and space. Empirical studies show this, but at great variance with the realities of nature, attention of theory to these facts historically was not a mainstream endeavor [44], and sadly that remains true today. Despite the absence of mainstream attention, considerable theory on the role of variation in time and space has been developed [14, 17, 19, 20, 38, 45-52]. Although the influence of this theory on empirical studies has not yet matched its potential importance, there is now a growing body of related

every bit as able as resources to define distinct niches for their prey species.

is strong relative to apparent competition.

**2.4 Variation in time and space** 

empirical studies [53-56]

tradeoff [42].

Some ideas about spatial and temporal variation, especially those under the heading of disturbance, have viewed variation in the same way that has often coloured thinking about predation: competition has been seen as limiting diversity, with variation in time and space as modifying or even nullifying what competition does [57-60]. Many of these views are influenced by the idea that variation in time and space make ecological communities "nonequilibrial," while niche ideas are presumed to depend on the concept of equilibrium [61]. Nothing could be further from the truth [3, 5].

Theoretical studies converge on the conclusion that variation in time and space are best viewed as providing more opportunities for the niches of species to be distinguished, defined by the responses of species to the changing conditions they encounter [3, 5]. This is most easily understood with respect to spatial variation [50, 62-64]. If species favor different habitats, and their resources are consumed and regenerated locally in those habitats, the habitats provide separate feedback loops, just like separate resources. Separate feedback loops with respect to predators are less likely, because predators are often wider ranging than their prey and so prey densities in one habitat can affect predation in other habitats, but this case has been developed theoretically [38]. The same would be true of resources too if resources moved between habitats, or if habitats changed their status over time on a shorter time scale than resource changes, i.e. if there were rapid spatio-temporal environmental variation [65]. Similar issues apply to temporal variation [5].

Niche theory for spatial and temporal variation has developed techniques for assessing the extent to which spatial and temporal variation can separate the niches of species [20, 22]. This is the concept of covariance between environment and competition and its generalization to include apparent competition [51, 52]. Such covariances assess whether density-dependent feedback loops change with the environmental conditions under consideration, and can thus separate the niches of species. Based on them, powerful techniques for definitively testing diversity maintenance hypotheses based on spatial and temporal have been proposed [20] and implemented in some cases [22, 66]. Extension of these ideas for other niche-based diversity maintenance mechanisms should be possible, and might well lead to much progress on deciding between the various diversity maintenance ideas that have been proposed.

#### **3. Conservation and management**

Although it is not possible at the present time to say with confidence which of the many possible diversity maintenance ideas applies in any particular system, these ideas share enough features in common that some general advice is possible. Our principal example has been MacArthur's consumer-resource model, and its extension to a consumer-resourcepredator model, for which average fitness and overlap measures are easy to define. Per capita growth rates in these models are linear functions of the densities of the various species. Models similar in spirit, but having nonlinear relationships, have been studied extensively by Tilman [17, 67], especially for plants limited by nonbiological resources, and compared with others by Chase and Leibold [3]. Although these simple measures of average fitness and niche overlap are not available, the underlying concepts remain applicable [3]. Other models of resource and apparent competition in both constant and variable environments have yielded measures of fitness and niche overlap, when they have been sought explicitly [5, 29, 47, 51, 52, 68], clearly demonstrating how these ideas extend

This outcome can be understood intuitively: one species increases in abundance when a valuable habitat type becomes more available, and, due to dispersal between habitat types, its abundance everywhere is increased. It thus has greater competitive effects on other species in all habitats. From the perspective of the theory above, average fitnesses and niche overlap change in the same way as when resource supply is changed. Similar affects apply to temporal environmental variation, an issue of particular concern with climate change [78]. For example, desert annual plants are believed to coexist with one another because different species are favored in different years, depending on the weather [19, 55]. Changes in weather patterns can change the relative *κ* values of the species if the weather becomes on average more favorable for some than for others. Changes in disturbance regimes can have

Although systems where coexistence relies on disturbance are sometimes regarded as nonequilibrial [80], they do not escape these general priniciples. Individual species may be adapted to survive and even take advantage of temporal change, but there is every reason to expect that individual species, and the system as a whole, are sensitive to changes in the average frequency and intensity of environmental events, such as disturbances [81, 82]. Surviving and taking advantage of temporal change may rely on life-history attributes such as dormancy, but may also depend on dispersal to and from refuge habitats that escape change either permanently or at given time (e.g when a fire sweeps through) [83]. Maintaining landscape connectivity, for example through maintenance of dispersal

Finally, we come to the important question of trophic structure. Recognition that predation and competition can have similar roles in diversity maintenance, and that the effects that each has can be undermined by the other, depending on their relative strengths, means that maintenance of trophic structure is vital for conservation [85]. There is much concern about loss of large predators in many systems [86-89]. These predators are often valued for their own sake. Conservation measures focused on them might aim merely for their persistence in nature, rather than for maintenance of the roles that they have in systems [90, 91]. For example, practices in the United States relating to wolf restoration face contentious arguments over the mere persistence of wolves, versus maintaining their numbers so that the ecosystem roles are restored as well [92, 93]. Discussion of trophic cascades, which are increasingly seen in marine systems, as the larger species are eliminated, focus much on the overall abundance of a given trophic level [94, 95]. However, the theory discussed here emphasizes that decimating one trophic level will affect the diversity maintenance roles that it has in other trophic levels for which is either a predator trophic level or a resource trophic

Taken together, these principles and examples reemphasize that species should not be managed in isolation. Factors improving the situation for one set of species may degrade it for others. Maintaining trophic structure, physical environmental structure, disturbance regimes, and spatial connectivity are all common sense ideas that are backed up by the

similar effects through changes in the patterns of spatio-temporal variation [79].

corridors, is especially important in such circumstances [84].

level, as has been found empirically in some systems [96].

This research was supported by NSF grant DEB-0542991.

theory of diversity maintenance mechanisms.

**4. Conclusions** 

**5. Acknowledgements** 

beyond MacArthur's [27] beginnings. Moreover, Shigesada [69] provides an explicit representation of spatial competition models in the MacArthur framework.

The fact that different species do differ from one another ecologically means that changes in the environment do affect different species differently or to different degrees. Under the individual-species approach to conservation and management, this would simply mean that different species are at risk to different degrees from some change in the environment due to human activities. Under the interactive model given here, changes in the environment need not directly affect a species for it to be affected indirectly through its linkages with other species. The fitness measures *κ* reveal this most simply. A mortality rate of a particular species might be changed by hunting, elimination of a predator, or changes in physical stress in the environment. That change would have a direct effect on that species, either increasing or decreasing its abundance and perhaps putting it at risk, but that change would also have effects on a species' relationships with others. In particular, in the MacArthur model, we see that the ratio of the *κ* values would be affected. An increase in the *κ* ratio in one species favor disadvantages another, potentially leading to its extinction.

One example of this *κ*-ratio analysis is its application to understanding invasive species of large effect [30, 70]. A large *κ* value in a particular system would allow a species to invade that system, but would have also have the effect of depressing native species and potentially driving them extinct. There are a number of ways that an invasive species might achieve a larger *κ* value including low susceptibility to natural enemies in the invaded system [71], advantages in resource harvesting [72], and lower sensitivity to competition than the native species [73]. Alternatively, an invasive species might have a larger *κ* value than native species because habitat degradation has increased stresses on native species, lowering their *κ* values, and rendering the system vulnerable to invasion [30, 74]. Viewed from another perspective, a native species might become endangered not because some environmental change has directly affected it, but because other guild members have had their *κ* values elevated by changes that directly benefited them.

Naturally, changes in resource supply also have important effects [75]. Increasing the carrying capacity of a resource that one species depends on directly increases that species' *κ* value, and so can increase its ability to compete with another species that does not depend on that resource. This other species would thus be disadvantaged, potentially driving it extinct. However, not just average fitness, *κ*, but niche overlap, *ρ*, would be affected by the increase in supply of a particular resource because *ρ* depends on resource carrying capacities (see Table 1). In particular, *ρ* would decrease with an increase in the carrying capacity of a resource that only one species uses. Nevertheless, the change in the *κ* value has the larger effect confirming the conclusion that changes directly benefiting only one species will harm its competitors indirectly. Effects like this are most clearly seen for invasive species where an increase in resource supply, for example nitrogen deposition for an invasive plant species [76, 77], gives advantages to invaders, depressing native species. However, these concerns apply not only between invaders and native species but between native species where factors helping just some species may well negatively affect others in the same guild [75].

These various effects stemming from changes in resource richness have their counterparts in habitat availabilities because habitat variation in space can have effects similar to resource diversity, as explained above in the section on variation in time and space. Thus, increasing the availability of a habitat type specific to one species, can negatively affect another species that does not use that habitat type, but shares other habitat types with the species that does.

beyond MacArthur's [27] beginnings. Moreover, Shigesada [69] provides an explicit

The fact that different species do differ from one another ecologically means that changes in the environment do affect different species differently or to different degrees. Under the individual-species approach to conservation and management, this would simply mean that different species are at risk to different degrees from some change in the environment due to human activities. Under the interactive model given here, changes in the environment need not directly affect a species for it to be affected indirectly through its linkages with other species. The fitness measures *κ* reveal this most simply. A mortality rate of a particular species might be changed by hunting, elimination of a predator, or changes in physical stress in the environment. That change would have a direct effect on that species, either increasing or decreasing its abundance and perhaps putting it at risk, but that change would also have effects on a species' relationships with others. In particular, in the MacArthur model, we see that the ratio of the *κ* values would be affected. An increase in the *κ* ratio in

One example of this *κ*-ratio analysis is its application to understanding invasive species of large effect [30, 70]. A large *κ* value in a particular system would allow a species to invade that system, but would have also have the effect of depressing native species and potentially driving them extinct. There are a number of ways that an invasive species might achieve a larger *κ* value including low susceptibility to natural enemies in the invaded system [71], advantages in resource harvesting [72], and lower sensitivity to competition than the native species [73]. Alternatively, an invasive species might have a larger *κ* value than native species because habitat degradation has increased stresses on native species, lowering their *κ* values, and rendering the system vulnerable to invasion [30, 74]. Viewed from another perspective, a native species might become endangered not because some environmental change has directly affected it, but because other guild members have had their *κ* values

Naturally, changes in resource supply also have important effects [75]. Increasing the carrying capacity of a resource that one species depends on directly increases that species' *κ* value, and so can increase its ability to compete with another species that does not depend on that resource. This other species would thus be disadvantaged, potentially driving it extinct. However, not just average fitness, *κ*, but niche overlap, *ρ*, would be affected by the increase in supply of a particular resource because *ρ* depends on resource carrying capacities (see Table 1). In particular, *ρ* would decrease with an increase in the carrying capacity of a resource that only one species uses. Nevertheless, the change in the *κ* value has the larger effect confirming the conclusion that changes directly benefiting only one species will harm its competitors indirectly. Effects like this are most clearly seen for invasive species where an increase in resource supply, for example nitrogen deposition for an invasive plant species [76, 77], gives advantages to invaders, depressing native species. However, these concerns apply not only between invaders and native species but between native species where factors helping just some species may well negatively affect others in the same guild [75]. These various effects stemming from changes in resource richness have their counterparts in habitat availabilities because habitat variation in space can have effects similar to resource diversity, as explained above in the section on variation in time and space. Thus, increasing the availability of a habitat type specific to one species, can negatively affect another species that does not use that habitat type, but shares other habitat types with the species that does.

representation of spatial competition models in the MacArthur framework.

one species favor disadvantages another, potentially leading to its extinction.

elevated by changes that directly benefited them.

This outcome can be understood intuitively: one species increases in abundance when a valuable habitat type becomes more available, and, due to dispersal between habitat types, its abundance everywhere is increased. It thus has greater competitive effects on other species in all habitats. From the perspective of the theory above, average fitnesses and niche overlap change in the same way as when resource supply is changed. Similar affects apply to temporal environmental variation, an issue of particular concern with climate change [78]. For example, desert annual plants are believed to coexist with one another because different species are favored in different years, depending on the weather [19, 55]. Changes in weather patterns can change the relative *κ* values of the species if the weather becomes on average more favorable for some than for others. Changes in disturbance regimes can have similar effects through changes in the patterns of spatio-temporal variation [79].

Although systems where coexistence relies on disturbance are sometimes regarded as nonequilibrial [80], they do not escape these general priniciples. Individual species may be adapted to survive and even take advantage of temporal change, but there is every reason to expect that individual species, and the system as a whole, are sensitive to changes in the average frequency and intensity of environmental events, such as disturbances [81, 82]. Surviving and taking advantage of temporal change may rely on life-history attributes such as dormancy, but may also depend on dispersal to and from refuge habitats that escape change either permanently or at given time (e.g when a fire sweeps through) [83]. Maintaining landscape connectivity, for example through maintenance of dispersal corridors, is especially important in such circumstances [84].

Finally, we come to the important question of trophic structure. Recognition that predation and competition can have similar roles in diversity maintenance, and that the effects that each has can be undermined by the other, depending on their relative strengths, means that maintenance of trophic structure is vital for conservation [85]. There is much concern about loss of large predators in many systems [86-89]. These predators are often valued for their own sake. Conservation measures focused on them might aim merely for their persistence in nature, rather than for maintenance of the roles that they have in systems [90, 91]. For example, practices in the United States relating to wolf restoration face contentious arguments over the mere persistence of wolves, versus maintaining their numbers so that the ecosystem roles are restored as well [92, 93]. Discussion of trophic cascades, which are increasingly seen in marine systems, as the larger species are eliminated, focus much on the overall abundance of a given trophic level [94, 95]. However, the theory discussed here emphasizes that decimating one trophic level will affect the diversity maintenance roles that it has in other trophic levels for which is either a predator trophic level or a resource trophic level, as has been found empirically in some systems [96].

#### **4. Conclusions**

Taken together, these principles and examples reemphasize that species should not be managed in isolation. Factors improving the situation for one set of species may degrade it for others. Maintaining trophic structure, physical environmental structure, disturbance regimes, and spatial connectivity are all common sense ideas that are backed up by the theory of diversity maintenance mechanisms.

#### **5. Acknowledgements**

This research was supported by NSF grant DEB-0542991.

[21] A. M. Siepielski and M. A. McPeek, "On the evidence for species coexistence: a critique

[22] A. L. W. Sears and P. Chesson, "New methods for quantifying the spatial storage effect: an illustration with desert annuals," *Ecology,* vol. 88, pp. 2240-2247, 2007. [23] J. A. Endler, *Natural Selection in the Wild*. Princeton, New Jersey: Princeton University

[24] E. D. Brodie, A. J. Moore, and F. J. Janzen, "Visualizing and quantifying natural selection," *Trends in Ecology and Evolution,* vol. 10, pp. 313-318, 1995. [25] E. L. Charnov, Life History Invariants: some explorations of symmetry in evolutionary

[26] P. B. Adler, J. HilleRisLambers, and J. M. Levine, "A niche for neutrality," *Ecology* 

[27] R. MacArthur, "Species packing and competitive equilibrium for many species,"

[28] P. Chesson, "MacArthur's consumer-resource model.," *Theoretical Population Biology,* 

[29] R. Haygood, "Coexistence in MacArthur-style consumer–resource models," *Theoretical* 

[30] K. Shea and P. Chesson, "Community ecology theory as a framework for biological invasions," *Trends in Ecology and Evolution,* vol. 17, pp. 170-176, 2002. [31] P. L. Chesson and R. R. Warner, "Environmental variability promotes coexistence in lottery competitive systems," *The American Naturalist,* vol. 117, pp. 923-943, 1981. [32] P. Chesson, "Multispecies competition in variable environments," *Theoretical Population* 

[33] P. Chesson and M. Rees, "Commentary on Clark et al . ( 2 0 0 7 ): Resolving the

[34] D. W. Purves and S. W. Pacala, "Ecological drift in niche-structured communities:

[35] P. Amarasekare, "Interference competition and species coexistence," *Proceedings of the Royal Society of London Series B-Biological Sciences,* vol. 269, pp. 2541-2550, 2002. [36] R. M. Callaway, Positive Interactions and Interdependence in Plant Communities.

[37] R. D. Holt and G. A. Polis, "A theoretical framework for intraguild predation," *The* 

[38] R. D. Holt, "Spatial heterogeneity, indirect interactions, and the coexistence of prey

[39] R. T. Paine, "Food web complexity and species diversity," *American Naturalist,* vol. 100,

[40] J. Lubchenco, "Plant species diversity in a marine intertidal community: importance of

[41] R. D. Holt, "Predation, apparent competition, and the structure of prey communities,"

herbivore food preference and algal competitive abilities," *American Naturalist,* vol.

neutral pattern does not imply neutral process," in *Biotic Interactions in the Tropics*, B. D., M. Pinard, and S. Hartley, Eds. Cambridge, UK: Cambridge University Press,

biodiversity paradox," *Ecology Letters,* vol. 10, pp. 659-661, 2007.

of the coexistence program," *Ecology,* vol. 91, pp. 3153-3164, 2010.

Fundacion BBVA, 2008, pp. 119-164.

*Letters,* vol. 10, pp. 95-104, 2007.

*Biology,* vol. 45, pp. 227-276, 1994.

vol. 37, pp. 26-38, 1990.

2005, pp. 103-138.

pp. 65-75, 1966.

112, pp. 23-39, 1978.

Dordrecht :: Springer,, 2007.

*American Naturalist,* vol. 149, pp. 745-764, 1997.

species," *American Naturalist,* vol. 124, pp. 377-406, 1984.

*Theoretical Population Biology,* vol. 12, pp. 197-229, 1977.

ecology. Oxford: Oxford University Press, 1993.

*Theoretical Population Biology,* vol. 1, pp. 1-11, 1970.

*Populaiton Biology,* vol. 61, pp. 215-223, 2002.

Press, 1986.

Camacho, A. Elosegui, C. Gracia, M. Estrada, J. C. Senar, and J. M. Gili, Eds. Bilbao:

#### **6. References**


[1] M. L. Rosenzweig, *Species Diversity in Space and Time*. Cambridge: Cambridge University

[2] J. Roughgarden, "Community structure and assembly," in *Perspectives in Ecological* 

[3] J. M. Chase and M. A. Leibold, *Ecological Niches: linking classical and contemporary* 

[4] S. A. Levin, "Community equilibria and stability, and an extension of the competitive exclusion principle," *American Naturalist,* vol. 104, pp. 413-423, 1970. [5] P. Chesson, "Mechanisms of maintenance of species diversity," *Annual Review of Ecology* 

[6] [P. Chesson and N. Huntly, "The roles of harsh and fluctuating conditions in the dynamics of ecological communities," *The American Naturalist,* vol. 150, pp. 519-553, 1997. [7] S. P. Hubbell, *The unified neutral theory of biodiversity and biogeography*. Princeton, NJ:

[8] I. Volkov, J. R. Banavar, S. P. Hubbell, and A. Maritan, "Patterns of relative species abundance in rainforests and coral reefs," *Nature,* vol. 450, pp. 45-49, 2007. [9] C. J. Chu, Y. S. Wang, G. Z. Du, F. T. Maestre, Y. J. Luo, and G. Wang, "On the balance

[10] P. B. Adler, "Neutral models fail to reproduce observed species-area and species-time relationships in Kansas grasslands," *Ecology,* vol. 85, pp. 1265-1272, 2004. [11] C. K. Kelly, M. G. Bowler, O. Pybus, and P. H. Harvey, "Phylogeny, niches, and relative abundance in natural communities," *Ecology,* vol. 89, pp. 962-970, 2008. [12] T. Wiegand, S. Gunatilleke, and N. Gunatilleke, "Species associations in a heterogeneous Sri lankan dipterocarp forest," *American Naturalist,* vol. 170, pp. E67-E95, 2007. [13] E. G. Leigh, "Neutral theory: a historical perspective," *Journal of Evolutionary Biology,* 

[14] D. Tilman, "Niche tradeoffs, neutrality, and community structure: A stochastic theory

[15] P. Amarasekare, "The geometry of coexistence," *Biological Journal of the Linnean Society,* 

[16] P. Chesson and J. J. Kuang, "The interaction between predation and competition"

[17] D. Tilman, *Resource Competition and Community Structure*. Princeton, New Jersey:

[18] P. Amarasekare, M. F. Hoopes, N. Mouquet, and M. Holyoak, "Mechanisms of

[19] P. Chesson, R. L. E. Gebauer, S. Schwinning, N. Huntly, K. Wiegand, M. S. K. Ernest,

[20] P. Chesson, "Quantifying and testing species coexistence mechanisms," in *Unity in* 

coexistence in competitive metacommunities," *The American Naturalist,* vol. 164, 2004.

A. Sher, A. Novoplansky, and J. F. Weltzin, "Resource pulses, species interactions and diversity maintenance in arid and semi-arid environments," *Oecologia,* vol. 141,

*Diversity: Reflections on Ecology after the Legacy of Ramon Margalef*, F. Valladares, A.

of resource competition, invasion, and community assembly," *Proceedings of the* 

between niche and neutral processes as drivers of community structure along a successional gradient: Insights from alpine and sub-alpine meadow communities,"

*approaches*. Chicago: The University of Chicago Press, 2003.

*Theory*, J. Roughgarden, R. M. May, and S. A. Levin, Eds. Princeton, N.J.: Princeton

**6. References** 

Press, 1995.

University Press, 1989, pp. 203-226.

*and Systematics,* vol. 31, pp. 343–66, 2000.

*Annals of Botany,* vol. 100, pp. 807-812, 2007.

Princeton University Press, 2001.

vol. 20, pp. 2075-2091, 2007.

vol. 71, pp. 1-31, 2000.

pp. 236-257, 2004.

*National Academy of Science USA,* 2004.

*Nature*, Vol 456, pp 235-238, 2008.

Princeton University Press, 1982.

Camacho, A. Elosegui, C. Gracia, M. Estrada, J. C. Senar, and J. M. Gili, Eds. Bilbao: Fundacion BBVA, 2008, pp. 119-164.


[61] M. Huston, "A general hypothesis of species diversity," *The American Naturalist,* vol.

[62] P. Amarasekare and R. M. Nisbet, "Spatial heterogeneity, source-sink dynamics, and

[63] P. L. Chesson, "Coexistence of competitors in spatially and temporally varying

[64] P. Chesson, "General theory of competitive coexistence in spatially-varying environments," *Theoretical Population Biology,* vol. 58, pp. 211-237, 2000. [65] R. E. Snyder, "Spatiotemporal population distributions and their implications for

[66] A. L. W. Sears, "Quantifying the effects of spatial environmental variation on dynamics

[67] D. Tilman, *Plant Strategies and the Dynamics and Structure of Plant Communities*.

[68] R. E. Snyder, E. T. Borer, and P. Chesson, "Examining the relative importance of spatial and nonspatial coexistence mechanisms," *American Naturalist,* vol. 166, pp. E75-E94, 2005. [69] N. Shigesada, "Spatial distribution of rapidly dispersing animals in heterogeneous

[70] A. S. MacDougall, B. Gilbert, and J. M. Levine, "Plant invasions and the niche," *Journal* 

[71] R. M. Keane and M. J. Crawley, "Exotic plant invasions and the enemy release

[72] K. Petren and T. Case, "An experimental demonstration of exploitation competition in

[73] R. M. Callaway and E. T. Aschehout, "Invasive plants versus their new and old neighbors: a mechanism for exotic invasion," *Science,* vol. 290, pp. 521-523, 2000. [74] A. A. Sher and L. A. Hyatt, "The disturbed resource-flux invasion matrix: a new framework for patterns of plant invasion," *Biological Invasions,* vol. 1, pp. 107-114, 1999. [75] C. M. Clark, E. E. Cleland, S. L. Collins, J. E. Fargione, L. Gough, K. L. Gross, S. C.

[76] M. L. Brooks, "Effects of increased soil nitrogen on the dominance of alien annual plants in the Mojave Desert," *Journal of Applied Ecology,* vol. 40, pp. 344-353, 2003. [77] D. A. Wedin and D. Tilman, "Influence of nitrogen loading and species composition on the carbon balance of grasslands," *Science,* vol. 274, pp. 1720-1723, 1996. [78] J. R. Malcolm, C. R. Liu, R. P. Neilson, L. Hansen, and L. Hannah, "Global warming

[79] S. Vuilleumier, C. Wilcox, B. J. Cairns, and H. P. Possingham, "How patch

[80] J. H. Connell, M. D. Lowman, and I. R. Noble, "Subcanopy gaps in temperate and tropical forests," *Austalian Journal of Ecology,* vol. 22, pp. 163-168, 1997.

Pennings, K. N. Suding, and J. B. Grace, "Environmental and plant community determinants of species loss following nitrogen enrichment," *Ecology Letters,* vol. 10,

and extinctions of endemic species from biodiversity hotspots," *Conservation* 

configuration affects the impact of disturbances on metapopulation persistence,"

*Theoretical Population Biology,* vol. 28, pp. 263-287, 1985.

Princeton, New Jersey: Princeton University Press, 1988.

environments.," *Lect. Notes Biomath.,* vol. 54, pp. 478-501., 1984.

hypothesis," *Trends in Ecology and Evolution,* p. this issue, 2002.

an ongoing invasion," *Ecology,* vol. 77, pp. 118-132, 1996.

the local coexistence of competing species," *The American Naturalist,* vol. 158, pp.

environments: a look at the combined effects of different sorts of variability,"

species coexistence in a variable environment," *Theoretical Population Biology,* vol.

of natural plant populations: field tests for covariance between response to the environment and competition," in *Population biology* Davis: University of California,

113, pp. 81-101, 1979.

572-584, 2001.

72, pp. 7-20, 2007.

Davis, 2004, 155 pp.

pp. 596-607, 2007.

*of Ecology,* vol. 97, pp. 609-615, 2009.

*Biology,* vol. 20, pp. 538-548, 2006.

*Theoretical Population Biology,* vol. 72, pp. 77-85, 2007.


[42] J. P. Grover and R. D. Holt, "Disentangling resource and apparent competition: realistic

[43] D. W. Stephens and J. R. Krebs, *Foraging Theory*. Princeton, N. J.: The Princeton

[44] D. R. Strong, "Density vagueness: abiding the variance in the demography of real

[45] Y. Iwasa and J. Roughgarden, "Interspecific competition among metapopulations with space-limited subpopulations," *Theoretical Population Biology,* vol. 30, pp. 194-214, 1986. [46] D. Tilman, "Competition and biodiversity in spatially structured habitats," *Ecology,* vol.

[47] R. E. Snyder and P. Chesson, "How the spatial scales of dispersal, competition, and

[48] B. M. Bolker, S. W. Pacala, and C. Neuhauser, "Spatial dynamics in model plant

[49] P. Amarasekare, "Competitive coexistence in spatially structured environments: a

[50] S. Muko and Y. Iwasa, "Incomplete mixing promotes species coexistence in a lottery

[51] J. J. Kuang and P. Chesson, "Interacting coexistence mechanisms in annual plant

[52] P. Chesson and J. J. Kuang, "The storage effect due to frequency-dependent predation

[53] P. B. Adler, S. P. Ellner, and J. M. Levine, "Coexistence of perennial plants: an embarrassment of niches," *Ecology Letters,* vol. 13, pp. 1019-1029, 2010. [54] P. B. Adler, J. HilleRisLambers, P. C. Kyriakidis, Q. F. Guan, and J. M. Levine, "Climate

[55] A. L. Angert, T. E. Huxman, P. Chesson, and D. L. Venable, "Functional tradeoffs

[57] J. H. Connell, "Diversity and the coevolution of competitors, or the ghost of

[58] G. E. Hutchinson, "The paradox of the plankton," *The American Naturalist,* vol. 95, pp.

[59] J. H. Connell, "Diversity in tropical rain forests and coral reefs," *Science,* vol. 199, pp.

[60] P. L. Chesson and T. J. Case, "Overview: nonequilibrium community theories: chance,

variability, history, and coexistence," in *Community Ecology*, J. Diamond and T. J.

synthesis," *Ecology Letters,* vol. 6, pp. 1109-1122, 2003.

*Population Biology,* vol. 77, pp. 56-70, 2010.

*Animal Ecology,* vol. 69, pp. 201-213, 2000.

competition past," *Oikos,* vol. 35, pp. 131-138, 1980.

Case, Eds. New York: Harper and Row, 1986, pp. 229-239.

353-376, 1998.

University Press, 1986.

75, pp. 2-16, 1994.

pp. 359-368, 2003.

164, 2010.

12798, 2006.

137-145, 1961.

1302-1310, 1978.

2003.

164, pp. 633-650, 2004.

Harper & Row Publishing, 1986.

models for plant-herbivore communities," *Journal of Theoretical biology,* vol. 191, pp.

populations.," in *Community Ecology*, J. Diamond and T. e. Case, Eds. New York:

environmental heterogeneity interact to affect coexistence," *American Naturalist,* vol.

communities: What do we really know?," *American Naturalist,* vol. 162, pp. 135-148,

model with permanent spatial heterogeneity," *Theoretical Population Biology,* vol. 64,

communities: frequency-dependent predation and the storage effect " *Theoretical* 

in multispecies plant communities," *Theoretical Population Biology,* vol. 78, pp. 148-

variability has a stabilizing effect on the coexistence of prairie grasses," *Proceedings of the National Academy of Sciences of the United States of America,* vol. 103, pp. 12793-

determine species coexistence via the storage effect," *Proceedings of the National Academy of Sciences of the United States of America,* vol. 106, pp. 11641-11645, 2009. [56] P. Amarasekare, "Spatial dynamics in a host-multiparasitoid community," *Journal of* 


**4** 

*Hungary* 

**Using Multiple Linear Regression** 

**Avian Species Imperilment in** 

R. Eliot Crafton and Brandon P. Anthony *Department of Environmental Sciences & Policy,* 

 *Central European University, Budapest* 

**Sub-Saharan Africa and Europe** 

**Models to Identify Factors Underlying** 

Determining what factors influence the threats faced by the world's flora and fauna is of key importance to conservation biologists (Cardillo et al., 2008; Davies et al., 2006; Smith et al., 2003; see Spangenberg, 2002). A plethora of research has been directed at this effort and has looked extensively at biological and anthropogenic factors, including social and socioeconomic conditions (e.g. Holland et al., 2009; Huby et al., 2006; Kerr & Currie, 1995; Lenzen et al., 2009; McKee et al., 2003). This chapter intends to supplement the existing literature by utilizing updated data to address this issue from a primarily socio-economic perspective for birds in a selection of sub-Saharan African and European countries. We generate several models using multiple linear regression to test the explanatory power of a host of variables, including human population density (HPD) per km2, Corruption Perception Index (CPI) score (as a proxy for governance), GDP per capita, and the average degrees from the equator. In addition, the results are considered in light of projected changes to HPD levels

Bird species are currently being impacted by several threats, resulting in the need for conservationists to address a wide range of issues (Brooke et al., 2008). These include landuse change, habitat destruction, invasive species, unsustainable exploitation, climate change, and insufficient governance (Brooke et al., 2008; Butchart, 2008; Lemoine et al., 2007; Lenzen et al., 2009; Reif et al., 2008; Smith et al., 2003; Thomas et al., 2004). These threats have the potential to impact the quality or quantity of available resources, directly impact the population, or change the conditions that a species may face. In addition, while these threats exist in isolation, many are correlated or exist in a cascading fashion (for example climate change can lead to habitat destruction). While these threats, and others, can be enumerated, the underlying drivers of these pressures are often rooted in socio-economic conditions, including corruption, HPD, and poverty level (Kerr & Currie, 1995; Pandit & Laband, 2009;

Understanding how these socioeconomic factors influence conservation agendas has been the focus of several authors, occasionally with contradictory results. While correlation can

**1. Introduction** 

for the year 2050 (United Nations [UN], 2008).

Pellegrini & Gerlagh, 2004).

