**3. Biodiversity in studies of BEF**

studies is called the stage "What?", in which scientists seek only to know what are the species that occur in a particular location. In the second stage, the stage called "How?", is characterized by the attempts to search for patterns of biodiversity, and the third stage (the stage "Why?"), seeks to explain the factors that lead to the patterns observed before. Also according to [1], when the book was written, biodiversity studies remained "emphatically in the second stage of development, with more discussions on the measures and standards, than with issues

In just over a decade after the publication of Gaston, several mechanisms that determine and influence biodiversity were discussed [1], and today the statement made by the author would certainly be very wrong. Many studies have been carried out in different spatial and temporal scales [1], using several different biological systems and in different regions of the world. Much remains to be done, but it is not risky to affirm that little knowledge is still to be generated regarding the mechanisms responsible for determining biodiversity, especially with respect

Around the same time of the publication of the book by Gaston, there was a shift in the view of scientists on biodiversity: the thought that biodiversity was an expression of abiotic environmental conditions gave way to the recognition that the properties of the environment were also affected by the biota [4]. This recognition spurred the search for the elucidation of the effect of the loss of biodiversity on ecosystem functioning, generating more than 50 different hypotheses to explain how this relationship would be [5]. Ecosystem functioning can be understood as a set of biogeochemical processes and ecosystem functions [6], responsible for the flow of matter and energy, and it is directly related to the dynamics of resources and

The hypotheses that explain the effect of biodiversity on ecosystem functioning may be classified into three major classes [8]. Within the first class fall the cases which species are assumed as redundant, and the loss of some species may be compensated by the presence of others, which perform the same function. Thus, to some extent, there would be no reduction of ecosystem functioning due to the lost of species and, on the other hand, there was no increase in the ecosystem functioning when species are added. In the second category of hypotheses are the cases when species are singular or unique, such as, for example, key species. According to this hypothesis, species lack redundancy and the loss or addition of some species would cause drastic changes to ecosystem functioning. Finally, there are hypotheses in which the ecosystem functioning effects of the loss or gain of species does not depend neither on the number nor the identity of the species present, but on the conditions under which this loss

These model-hypotheses formed the necessary structure to the experimental tests of the relationship between biodiversity and ecosystem functioning [5]. The initial tests involved theoretical approaches, as well as the use of simplified micro and mesocosm laboratory experiments. Posterior studies incorporated actual environmental variation through observa‐ tional and manipulative experiments, allowing higher applicability in public policy manage‐

occurs, so that the effects of species on the functioning become idiosyncratic.

to the definition of biodiversity as the number of species in a given area.

related to the mechanisms."

34 Biodiversity - The Dynamic Balance of the Planet

the stability of the ecosystem [7].

ment and biodiversity conservation [9].

Biodiversity may be estimated through several different metrics, and the most traditional approach is to access it via taxonomic diversity [10]. Later in the studies of BEF, functional diversity started to be used as an additional metric [11, 12] and, more recently, phylogenetic diversity has been included as a proposal of metric to biodiversity [13-15].

Taxonomic diversity, in turn, may be translated by species abundance, richness and compo‐ sition [16, 17] parameters that are easily accessed and that may give basic information to the generation of diversity indices (such as Simpson or Shannon). These diversity indices are useful to synthesize and compare the biodiversity in different environments or sites [18, 19]. However, the indiscriminate use of diversity indices, and as a goal in itself, rather than using them as a useful metric of diversity, prompted several authors to avoid its use in favor of species richness as a metric. Species may be identified from morphological or genetic traits and, when using taxonomic diversity as a metrics, it is assumed that the differences among species are determined by these aspects. Nevertheless, some authors suggest that taxonomic diversity may not be the most adequate metric [12] to evaluate the effect of organisms on ecosystem functioning. To better evaluate such effects, it would be necessary to use the functional diversity, even though it is expected that, with an increase in species richness, there would be an increase in functional diversity [20].

Functional diversity may be understood as the group of characteristics, of species or organisms, responsible by altering one or more aspects of ecosystem functioning [21]. Such characteristics may be related to the abilities that organisms have to engage or to alter ecosystem processes such as seed dispersal [22], pest biological control [23], pollination [24], nutrient cycling [25], decomposition [26], productivity [27, 28], amongst others. However, when inferring functional diversity from species richness, it is assumed that the relationship between species number and niche occupation is linear, which usually does not occur in nature [7].

Therefore, the use of species richness as an estimate of functional diversity has been criticized, and the estimate of functional diversity has been achieved by species classification by their trophic level, guild, as well as physiological and phenotypic characteristics. Functional diversity is usually used to estimate the biodiversity of plant communities, classifying plant species according to their physiognomy, phenology or photosynthetic pathways. Animals are frequently grouped in guilds based on their consumption, but commonly with a low level of resolution, due to the weak knowledge of their biology. Such characteristics are frequently considered to determine functional diversity because they supposedly relate to aspects of the niches occupied and, consequently, they may express the effects of the organisms in ecosystem processes [21, 29]. From the analysis of these characteristics it is possible to determine and to include species in functional groups, which assemble organisms that fulfill similar functions and, consequently, have similar effects on the ecosystem [30, 22]. This approach is the most usual when using functional diversity in BEF studies.

Functional diversity is considered an estimate that may express more powerfully the effects of biodiversity on ecosystem functioning, because it refers to those biodiversity components that directly affect in how ecosystem operates [22].

There are two questions regarding the use functional diversity as an estimate of biodiversity. The first relates to its teleological use, which involve purpose, such as some earlier discussion of the hypothesis of redundancy [31]. These discussions were full of anthropocentric analogies, describing redundant species as passengers in an automobile [32] or words in a phrase [31]. The replacement of purpose with process, as suggested by [33] and [31], "retains the intuition that if something is functional, it must do something". Therefore, to be functional, species must interact with ecological processes, a relationship that brings us to the second possible question regarding the use of functional diversity as an estimate of biodiversity: is it a tautological discussion?

propose that, since these characteristics may influence ecosystem processes, the phylogenetic history may mirror more accurately the effects of biodiversity on ecosystem functioning. This may occur because closely related species tend to occupy similar niches, and thus may play a similar role in the ecosystem processes. Consequently, communities composed by species that encompass a higher phylogenetic diversity (or more distantly related species), would also encompass a wider range of niches. These communities would be more efficient communities in maintaining ecosystem functioning (Figure. 1), due to the higher complementarity of species

Biodiversity and Ecosystem Functioning: a Conceptual Model of Leaf Litter Decomposition

http://dx.doi.org/10.5772/57396

37

There are two main reasons to use phylogenetic diversity as a relevant biodiversity estimate in comprehension of ecosystem functioning, instead of functional groups: (1) the removal or addition of a functionally redundant species may have effects of ecosystem processes, highlighting important functional among species, non-captured by functional groupings; and (2) in the case of prediction of change in ecosystem processes like productivity, for example, functional groups may explain as much as categories of randomly chosen groups [34].

Phylogenetic diversity may be defined, in general, as the sum of phylogenetic branches that ling species [34], although several methods of estimating phylogenetic diversity may be recognized [35]. These metrics may be classified into two basic types. The type I metrics begin by calculating a distinctness score for all species of a regional phylogeny and, after this calculation, the distinctness scores of the focal subset of species are summed (or any other function to be used), to produce the phylogenetic diversity metric. Contrarily, the type II metrics use a local phylogeny, from which the distinctness scores are calculated to the focal

Important considerations regarding the choice of these estimates are presented [41], based on the available data, listing the metrics that are typically used in community ecology studies. These are: (1) Phylogenetic diversity (PD): sum of all branches lengths in the portion of a phylogenetic tree connecting the focal set of species; (2) Mean phylogenetic distance (MPD): mean phylogenetic distance between each pair of species in the focal set of species; (3) Sum of phylogenetic distances (SPD): sum of phylogenetic distances between each pair of species (MPD) multiplied by the number of species pairs; (4) Mean nearest neighbor distance (MNND): mean phylogenetic distance from each species to its closest relative in the focal species set. These metrics are primarily used on conservation biology studies that focus on species conservation, supporting conservation decisions. Nevertheless, in relation to biodiversity and ecosystem functioning studies we do not have a standard metric for usage and this lack comprises an important issue in BEF studies that uses phylogenetic diversity as biodiversity estimate. We recommend to researchers who want to use PD as biodiversity estimate, to search for a powerful background in choosing the metric. We expect that with the increase of studies using PD as metric in BEF researches will result in an improvement of using this metric.

However, even though the estimates of phylogenetic diversity may be promising in BEF studies, there are still no solid evidences about its relationship with ecosystem functioning [15], except those reported by [14]. These latter authors propose a model that partitions the effects

of biodiversity into phylogenetic effects and other community properties.

effect on ecosystem processes [33].

species [35].

Because functional diversity is an estimate derived from the relationship between species and ecological processes, it stays clear to us that an observed relationship between ecosystem functioning and functional biodiversity must be positive and highly significant. There are even suggestions that ecological processes and ecological function might be treated as synonymous [31], increasing the possibility of a positive relationship in the studies relating ecosystem functioning and functional diversity.

The above issues must be considered when using functional diversity as an estimate of biodiversity in studies relating ecosystem function and biodiversity. The definition of func‐ tional diversity is somewhat established since the proposal by [31], but this metric would only correlate two variables that are a priori related. The metric could be attractive to achieve a definition of the functional groups that are more prevalent in the effects on ecosystem processes, but the metrics alone lack an explanatory power. It is still necessary to question why biodiversity, if it estimated independently of functionality, interferes increasing or decreasing ecosystem functioning.

There has been a suggestion that animal traits that are predicted to influence ecosystem processes must be defined a priori [13]. The actual effect of each functional group on ecosystem processes is then tested and the predictions may be refined accordingly. The above authors also propose a hypothetical relationship between functional diversity and the number of taxa or groups. They suggest a loose relationship between functional diversity and species richness, and adding more families within a trophic level would add more functional trait variation than adding more species within a genus or family. Consequently, the increase in taxa variation would represent an increase of the expected functional diversity, and a priori classification of functional groups could be achieved by a metric that could capture phylogenetic diversity.

Recent studies observed the existence of a relationship between functional and phylogenetic diversities, mainly when considering functional groups as an estimate of functional diversity [16, 30, 34]. According to [30], assemblages with a higher phylogenetic diversity present higher functional diversity, possibly due to the complementarity of ecosystem functions among clades, contrasting with the expected redundancy among species that possess common ancestors.

Species functional characteristics, as well as all other traits, appear along their evolutionary history, affecting the way species distribute and relate to each other. Therefore, these traits tend to be shared among the species that have a common ancestor [30]. The latter authors propose that, since these characteristics may influence ecosystem processes, the phylogenetic history may mirror more accurately the effects of biodiversity on ecosystem functioning. This may occur because closely related species tend to occupy similar niches, and thus may play a similar role in the ecosystem processes. Consequently, communities composed by species that encompass a higher phylogenetic diversity (or more distantly related species), would also encompass a wider range of niches. These communities would be more efficient communities in maintaining ecosystem functioning (Figure. 1), due to the higher complementarity of species effect on ecosystem processes [33].

There are two questions regarding the use functional diversity as an estimate of biodiversity. The first relates to its teleological use, which involve purpose, such as some earlier discussion of the hypothesis of redundancy [31]. These discussions were full of anthropocentric analogies, describing redundant species as passengers in an automobile [32] or words in a phrase [31]. The replacement of purpose with process, as suggested by [33] and [31], "retains the intuition that if something is functional, it must do something". Therefore, to be functional, species must interact with ecological processes, a relationship that brings us to the second possible question regarding the use of functional diversity as an estimate of biodiversity: is it a tautological

Because functional diversity is an estimate derived from the relationship between species and ecological processes, it stays clear to us that an observed relationship between ecosystem functioning and functional biodiversity must be positive and highly significant. There are even suggestions that ecological processes and ecological function might be treated as synonymous [31], increasing the possibility of a positive relationship in the studies relating ecosystem

The above issues must be considered when using functional diversity as an estimate of biodiversity in studies relating ecosystem function and biodiversity. The definition of func‐ tional diversity is somewhat established since the proposal by [31], but this metric would only correlate two variables that are a priori related. The metric could be attractive to achieve a definition of the functional groups that are more prevalent in the effects on ecosystem processes, but the metrics alone lack an explanatory power. It is still necessary to question why biodiversity, if it estimated independently of functionality, interferes increasing or decreasing

There has been a suggestion that animal traits that are predicted to influence ecosystem processes must be defined a priori [13]. The actual effect of each functional group on ecosystem processes is then tested and the predictions may be refined accordingly. The above authors also propose a hypothetical relationship between functional diversity and the number of taxa or groups. They suggest a loose relationship between functional diversity and species richness, and adding more families within a trophic level would add more functional trait variation than adding more species within a genus or family. Consequently, the increase in taxa variation would represent an increase of the expected functional diversity, and a priori classification of functional groups could be achieved by a metric that could capture phylogenetic diversity.

Recent studies observed the existence of a relationship between functional and phylogenetic diversities, mainly when considering functional groups as an estimate of functional diversity [16, 30, 34]. According to [30], assemblages with a higher phylogenetic diversity present higher functional diversity, possibly due to the complementarity of ecosystem functions among clades, contrasting with the expected redundancy among species that possess common

Species functional characteristics, as well as all other traits, appear along their evolutionary history, affecting the way species distribute and relate to each other. Therefore, these traits tend to be shared among the species that have a common ancestor [30]. The latter authors

discussion?

functioning and functional diversity.

36 Biodiversity - The Dynamic Balance of the Planet

ecosystem functioning.

ancestors.

There are two main reasons to use phylogenetic diversity as a relevant biodiversity estimate in comprehension of ecosystem functioning, instead of functional groups: (1) the removal or addition of a functionally redundant species may have effects of ecosystem processes, highlighting important functional among species, non-captured by functional groupings; and (2) in the case of prediction of change in ecosystem processes like productivity, for example, functional groups may explain as much as categories of randomly chosen groups [34].

Phylogenetic diversity may be defined, in general, as the sum of phylogenetic branches that ling species [34], although several methods of estimating phylogenetic diversity may be recognized [35]. These metrics may be classified into two basic types. The type I metrics begin by calculating a distinctness score for all species of a regional phylogeny and, after this calculation, the distinctness scores of the focal subset of species are summed (or any other function to be used), to produce the phylogenetic diversity metric. Contrarily, the type II metrics use a local phylogeny, from which the distinctness scores are calculated to the focal species [35].

Important considerations regarding the choice of these estimates are presented [41], based on the available data, listing the metrics that are typically used in community ecology studies. These are: (1) Phylogenetic diversity (PD): sum of all branches lengths in the portion of a phylogenetic tree connecting the focal set of species; (2) Mean phylogenetic distance (MPD): mean phylogenetic distance between each pair of species in the focal set of species; (3) Sum of phylogenetic distances (SPD): sum of phylogenetic distances between each pair of species (MPD) multiplied by the number of species pairs; (4) Mean nearest neighbor distance (MNND): mean phylogenetic distance from each species to its closest relative in the focal species set. These metrics are primarily used on conservation biology studies that focus on species conservation, supporting conservation decisions. Nevertheless, in relation to biodiversity and ecosystem functioning studies we do not have a standard metric for usage and this lack comprises an important issue in BEF studies that uses phylogenetic diversity as biodiversity estimate. We recommend to researchers who want to use PD as biodiversity estimate, to search for a powerful background in choosing the metric. We expect that with the increase of studies using PD as metric in BEF researches will result in an improvement of using this metric.

However, even though the estimates of phylogenetic diversity may be promising in BEF studies, there are still no solid evidences about its relationship with ecosystem functioning [15], except those reported by [14]. These latter authors propose a model that partitions the effects of biodiversity into phylogenetic effects and other community properties.

There are three aspects in the studies of BEF relationship that have to be distinguished. The first aspect is the evaluation of which would be the appropriate metrics for biodiversity and for ecosystem functioning. This evaluation involves both methodological and mechanistic considerations, but certainly may affect the results of BEF relationship evaluation, for nonbiological reasons. Therefore, one has to be especially careful to distinguish actual mechanistic relationships from tautological correlations, disconnected from biology itself. The second aspect of BEF studies is the evaluation of BEF relationship hypotheses, which necessarily involve regression models that must be tested against null hypothesis of no relationship, or spurious relationship due exclusively to methodological or mathematical issues.

These two first aspects of BEF studies have been often confused, and may lead to unwarranted conclusions. For example, studies on the relationship of phylogenetic diversity in relation to ecosystem functioning [12], have shown that phylogenetic diversity is more relevant to predict ecosystem functioning, at least when evaluated by primary productivity, than crude species richness or diversity. The authors argue for evolutionary reasons for this pattern, and further suggest that thus one should favor phylogenetic diversity rather than species diversity, as a metric to evaluate BEF. This should not, however, be interpreted as a test of the BEF relation‐ ship, but rather a methodological refinement, prior to an actual hypotheses testing.

The third aspect of BEF studies, which is the one that we proposed to illustrate in this chapter, regards the explanations for the actual relationship between biodiversity and ecosystem functioning. Disregarding the metrics used to depict biodiversity and the shape of the relationship between biodiversity and ecosystem function, the question why does ecosystem function vary with biodiversity remains. The mechanistic processes by which these two community parameters relate are then described below, using decomposition as the ecosystem function and the variety of species that compose the litter, the so-called litter mixture, as the biodiversity metrics.

The physico-chemical environment is related to climate, especially humidity and temperature [45, 47]. The climate may indirectly affect litter decomposition, changing litter characteristics, or directly, controlling the activity of decomposing organisms. Plant chemical composition may result from soil formation and from nutrient cycling, and both are regulated by the climate [47]. Therefore, nutrient mineralization may be more accelerated in hot and humid climates, resulting in higher nutrient concentration in litter, increasing its degradability and decompo‐ sition [48]. In addition, the direct effect of physico-chemical environment on decomposition occurs from favoring the activity of decomposer organisms by the higher temperature and

**Figure 1.** Relationship between functional and phylogenetic diversities in a community, considering the existence of a phylogenetic signal to the functional characteristics in question. PD: Phylogenetic diversity. Closely related species are more similar (symbols differ only in color) than distantly related species (symbols differ in shape). Adapted from [15].

Biodiversity and Ecosystem Functioning: a Conceptual Model of Leaf Litter Decomposition

http://dx.doi.org/10.5772/57396

39

The second main factor regulating litter decomposition is the quality of organic matter from litter, which is frequently associated to leaf degradability [46]. There is not a unanimity regarding a valid index of litter degradability, although certain nutrient concentrations have been usually associated to higher quality [50]. High nitrogen (N) and phosphorus (P) concen‐ trations, as well as high proportions of easily degradable carbon compounds, such as sugars, have been associated to better litter quality [50], which translates into higher decomposition rates. Contrarily, less degradable carbon compounds (such as lignin), decrease degradability and litter quality, because these compounds require higher energy from decomposers to break

humidity, increasing the rates of litter decomposition and nutrient release [48, 49].

the organic matter [46].
