**2.4 Runoff generation mechanisms**

Surface runoff and erosion in semi-arid areas are the result of various factors associated with rainfall (duration and intensity), soil (moisture, cracking, crusting, and soil infiltration capacity), plant cover (density) and terrain relief. During the study period, 46 precipitationrunoff-sediment yield events were recorded and their main characteristics and the corresponding hydraulic responses were evaluated. 55% of the rainfall events duration were less than 60 minutes in duration and 66% between 18h00 and 06h00. Rainfall peak rate ranged between 9.14 and 137.16 mm.h-1. In 56% of events peak rate surpassed 40 mm.h-1 and in 5 events it exceeded 90 mm.h-1. The runoff coefficient is the relationship between surface runoff and rainfall levels during the event. Observed values of runoff coefficient, rainfall peak rate and precipitation height are presented for the beginning and the end of the rainy periods in Figures 4(a) and 4(b), respectively.

33% of the observed events didn't produce runoff. Observed runoff coefficients were lower than 0.1 for 82% of events, which indicate high soil water storage capacity. Only 5 events exhibited runoff coefficients higher than 0.2; these higher values were possibly due to influence of antecedent rainfall, soil water storage capacity and the density of vegetation cover.

Plot-Scale Experimental Studies 159

decrease in sediment supply are reflected in an empirical relationship that seems to be more

Fig. 5. Empirical relationships between sediment yield and precipitation height in 2006 and

Flora in the plot is formed by the *Caatinga* biome composed of xerophilous species. These species possess mechanisms to adapt and cope with water dryness spells and their physiological processes are conditioned to water availability. Climate factors and the soil water availability are determinants of natural ecosystem functioning. Accordingly, two distinct landscape scenarios may be observed during the drought and rainy periods. In the dry season, annual species typically become absent (herbaceous); permanent species survive due to their root structure and ability to store water during this period. Another adaptation of the permanent species in the area is their ability to lose their leaves when water is scarce to avoid water loss through transpiration. Figures 6(a) and 6(b) illustrate the vegetation

Vegetation cover in the plot is most dense at the end of the rainy season. A survey of existing species in the plot identified 31 individuals of the *Mimosa Tenuiflora* species (medium-sized tree) and 15 *Cróton campestris* (shrub). A predominance of annual species was observed, whose life cycle (germination, flowering, fruiting and death) is completed in less than one year. Table 3 depicts the observed vegetation species in the plot during the

The results from the analyses highlight the stark difference between the soil hydraulic properties, faunal activity and vegetation cover in the rainy season and dry season. This is consistent with other studies in semi-arid regions, which have also found that water availability is the key driving factor of biological and geomorphological processes (Cammeraat, 2002; Cerda 2002). In contrast, Cammeraat (2002) found that in a humid

independent from precipitation characteristics.

landscape scenario during the wet and dry seasons.

**2.6 Summary of plot-scale experimental observations** 

2007.

**2.5 Vegetation cover** 

rainy period.

(b)

Fig. 4. Runoff coefficient, precipitation height and rainfall peak rate at the (a) beginning of the rainy period and (b) end of the rainy period.

Furthermore, these values above 0.2 were observed at the beginning of the rainy period when the vegetation cover density was low. Indeed, vegetation density increased as the rainy period progressed, thereby increasing infiltration capacity and soil water storage. During the rainy period, runoff coefficients were lower than 0.05, and seemed to be independent from rainfall characteristics. This demonstrates the role of native vegetation in improving infiltration capacity and soil water storage. Also, a feedback relationship seemed to control the regeneration of annual species influenced by soil moisture, intense faunal activity and seed supply followed the first week of the wet season. The graph in Figure 5 presents empirical relationships involving sediment yield and precipitation height in 2006 (plot installation) and 2007. The impact of disturbance to the soil surface during plot installation is clearly visible. In 2007, undisturbed natural conditions in the plot and a

Time

Time

Precipitation, PRR

Precipitation, PRR

(a)

12/2 17/2 22/2 27/2 4/3

Precipitation height (mm) Runoff coefficient Peak rainfall rate (mm/h)

(b)

0,0 140

1/3 31/3 30/4 30/5 29/6 29/7 28/8 27/9 27/10 26/11

Precipitation height (mm) Runoff coefficient Peak rainfall rate (mm/h)

Furthermore, these values above 0.2 were observed at the beginning of the rainy period when the vegetation cover density was low. Indeed, vegetation density increased as the rainy period progressed, thereby increasing infiltration capacity and soil water storage. During the rainy period, runoff coefficients were lower than 0.05, and seemed to be independent from rainfall characteristics. This demonstrates the role of native vegetation in improving infiltration capacity and soil water storage. Also, a feedback relationship seemed to control the regeneration of annual species influenced by soil moisture, intense faunal activity and seed supply followed the first week of the wet season. The graph in Figure 5 presents empirical relationships involving sediment yield and precipitation height in 2006 (plot installation) and 2007. The impact of disturbance to the soil surface during plot installation is clearly visible. In 2007, undisturbed natural conditions in the plot and a

Fig. 4. Runoff coefficient, precipitation height and rainfall peak rate at the (a) beginning of

the rainy period and (b) end of the rainy period.

0,0 0,1 0,2 0,3 0,4 0,5 0,6

> 0,1 0,2 0,3 0,4 0,5 0,6 0,7

Runoff coefficient

Runoff coeficient

decrease in sediment supply are reflected in an empirical relationship that seems to be more independent from precipitation characteristics.

Fig. 5. Empirical relationships between sediment yield and precipitation height in 2006 and 2007.

#### **2.5 Vegetation cover**

Flora in the plot is formed by the *Caatinga* biome composed of xerophilous species. These species possess mechanisms to adapt and cope with water dryness spells and their physiological processes are conditioned to water availability. Climate factors and the soil water availability are determinants of natural ecosystem functioning. Accordingly, two distinct landscape scenarios may be observed during the drought and rainy periods. In the dry season, annual species typically become absent (herbaceous); permanent species survive due to their root structure and ability to store water during this period. Another adaptation of the permanent species in the area is their ability to lose their leaves when water is scarce to avoid water loss through transpiration. Figures 6(a) and 6(b) illustrate the vegetation landscape scenario during the wet and dry seasons.

Vegetation cover in the plot is most dense at the end of the rainy season. A survey of existing species in the plot identified 31 individuals of the *Mimosa Tenuiflora* species (medium-sized tree) and 15 *Cróton campestris* (shrub). A predominance of annual species was observed, whose life cycle (germination, flowering, fruiting and death) is completed in less than one year. Table 3 depicts the observed vegetation species in the plot during the rainy period.

#### **2.6 Summary of plot-scale experimental observations**

The results from the analyses highlight the stark difference between the soil hydraulic properties, faunal activity and vegetation cover in the rainy season and dry season. This is consistent with other studies in semi-arid regions, which have also found that water availability is the key driving factor of biological and geomorphological processes (Cammeraat, 2002; Cerda 2002). In contrast, Cammeraat (2002) found that in a humid

Plot-Scale Experimental Studies 161

**Strata Family Scientific name Popular name** 

Amaranthaceae *Froelichia* 

mimosoideae *Mimosa tenuiflora* Jurema Preta

Euphorbiaceae *Acalypha communis* Algodãozinho Asteraceae *Hyptis suaveolens* Alfazema-braba

Leg. Papilionoideae Stylozanthes Stylozanthes

Leg. Mimosoideae *Mimosa ursina* Jureminha Turneraceae *Turnera subulata* Chanana Rubiaceae *Diodia teres* Quebra-tijela

Asteraceae Centratherum

Poaceae *Aristida* 

The runoff coefficient values were generally quite low for the study site, with a third of events not resulting in any runoff and the majority (82%) of events producing values less than 0.1. The 5 events that produced runoff coefficient values over 0.2 occurred at the beginning of the rainy period when vegetation cover was low. It was observed that as the rainy period progressed, the vegetation density increased, along with increased infiltration capacity and soil water storage, and consequently the runoff coefficients dropped to below 0.05. Indeed, the saturated hydraulic conductivity rates in the rainy period were approximately six times higher than that observed during the drought period. Also, the soils beneath the permanent plant species were found to have higher infiltration capacity rates. Another important observation from this study was the higher sediment yields in 2006 following the installation of the plot, compared to 2007. In 2007, the soils were relatively undisturbed and accordingly there was a decrease in sediment yield. A significant increase in the number of arthropods was also observed during the rainy season. This phenomenon between microbial and hydrological processes in arid and semi-arid environments was explored by Belnap et al. (2005) using the *trigger-transfer-reserve-pulse* framework (Ludwig et al., 1997). Under this framework, rainfall can be considered as the *trigger* which results in the

Table 3. Observed vegetation species in the plot during the rainy period.

Malvaceae Pavonia cancellata Malva-rasteira

*humboldtiana* Ervanço

*Sida rhombifolia* Relógio

punctatum Perpétua-roxa

*adscensionu L.* Capim Panasco

Euphorbiaceae *Cróton campestris* Velame Sterculiaceae *Waltheria bracteosa* Corre-campo Sterculiaceae *Waltheria indica* Malva-branca

Arboreo Leguminosae-

Shrubby

Annual species

temperate climate (Luxembourg) these processes were dominated by water surplus. In this semi-arid Brazilian plot study, the soils were composed of gravel, fine, medium and coarse sand, and also rock fragments, which provides natural roughness to the soil surface. This natural roughness is particularly important in reducing water and soil loss during the dry season, when vegetation cover is sparse.

Fig. 6. Vegetation landscapes in the (a) drought and (b) rainy periods.

temperate climate (Luxembourg) these processes were dominated by water surplus. In this semi-arid Brazilian plot study, the soils were composed of gravel, fine, medium and coarse sand, and also rock fragments, which provides natural roughness to the soil surface. This natural roughness is particularly important in reducing water and soil loss during the dry

Fig. 6. Vegetation landscapes in the (a) drought and (b) rainy periods.

(a)

(b)

season, when vegetation cover is sparse.


Table 3. Observed vegetation species in the plot during the rainy period.

The runoff coefficient values were generally quite low for the study site, with a third of events not resulting in any runoff and the majority (82%) of events producing values less than 0.1. The 5 events that produced runoff coefficient values over 0.2 occurred at the beginning of the rainy period when vegetation cover was low. It was observed that as the rainy period progressed, the vegetation density increased, along with increased infiltration capacity and soil water storage, and consequently the runoff coefficients dropped to below 0.05. Indeed, the saturated hydraulic conductivity rates in the rainy period were approximately six times higher than that observed during the drought period. Also, the soils beneath the permanent plant species were found to have higher infiltration capacity rates. Another important observation from this study was the higher sediment yields in 2006 following the installation of the plot, compared to 2007. In 2007, the soils were relatively undisturbed and accordingly there was a decrease in sediment yield. A significant increase in the number of arthropods was also observed during the rainy season. This phenomenon between microbial and hydrological processes in arid and semi-arid environments was explored by Belnap et al. (2005) using the *trigger-transfer-reserve-pulse* framework (Ludwig et al., 1997). Under this framework, rainfall can be considered as the *trigger* which results in the

Plot-Scale Experimental Studies 163

on several scales. It is estimated that the global area dedicated to agriculture has increased five times over the last 200 years, prompted by population growth and higher food demand (UNEP, 1995). On the other hand, reservoir construction and the damming of water and sediment significantly reduce the amount of sediment reaching floodplains and estuaries. A recent survey using long-term records of large basins subjected to the impact of human activity found that, in some cases, increased sediment in river systems may not impact deltas and estuaries (Dai & Tan, 1996; Walling, 2000; Walling & Fang, 2003) due to sediment retention in reservoirs located upstream. Thus, erosion processes reflect the combined action of climate factors and disturbances in the basin as a result of

Urbanization may also cause substantial changes in hydrologic behavior and erosive processes (Taylor, 2007). Land occupation of urban areas brings together the production of liquid and solid residues that, if not adequately collected, may be detrimental to water and sediment quality. Urbanization is associated with building construction and infrastructure service. Paved surfaces in urban environments result in lower amounts of water infiltration, which, in turn, can produce adverse social and economic impacts such as floods. In developing countries, it can be observed that urban development does not commonly occur in line with infrastructure and urban services investments. Often public services such as health and education are inadequate and planning and provisions to prevent or cope with extreme events (e.g. prevention measures and land occupation control) are lacking. In addition, urban occupation generates sediment contaminated by toxic substances (heavy metals, pesticides, oils, organic compounds), which can adhere to the fine fractions in the fluvial environment (Robertson et al., 2003; Lecoanet et al., 2003). Primary sources of sediment contamination in urban areas are domestic sewage and the construction of buildings and roadways. The presence of contaminants significantly affects aquatic organisms that feed on the sediment, which highlights the implications of land management

The research of this study was supported by CNPq–Technologic and Scientific National Development Council/Science and Technology Ministry/Brazilian Government. This

Abrahams, A.D., Li G, Krishnam C. & Atkinson, J.F. (1998). Predicting sediment transport by

Abrahams, A.D., Parsons, A.J., Wainwright, J. (1995). Effects of vegetation change on interrill

Belnap, J., Welter, J.R., Grimm, N.B., Barger, N., Ludwig, J.A. (2005) Linkages between

interrill overland flow on rough surfaces, *Earth Surface Processes and Landforms*, Vol.

runoff and erosion, Walnut Gulch, southern Arizona, *Geomorphology*, Vol. 13, pp.

microbial and hydrologic processes in arid and semiarid watersheds, Ecology, Vol.

unsustainable human activities.

on other parts of the system.

support is gratefully acknowledged.

23, pp. 481-492.

86, pp. 298-307.

37-48.

**4. Acknowledgment** 

**5. References** 

*transfer* of resources such as water, nutrients and soil to the receiving patch (referred to as the *reserve*) downslope. Patches within a semi-arid landscape are typically formed by plants, under which soils tend to have higher organic matter, nutrients and microbial activity. The rainfall and subsequent transfer of materials to the patch triggers a *pulse* of biological activity, which in turn produces positive feedbacks including the formation of stronger or new soil aggregates that improve soil stability and water infiltration (Belnap et al., 2005). Human activities that disrupt this positive feedback loop between the abiotic and biological activities, for example native vegetation clearance or overgrazing, can lead to negative impacts on the system. For example, removal of vegetation will reduce organic matter input to the soil, which can lead to decreased microbial activity and poorer soil structure and lower soil storage capacity.
