Sustainability in Arid Grasslands: New Technology Applications for Management

As biodiversity, ecosystem function, and ecosystem services become more closely linked with human well-being at all scales, the study of ecology takes on increasing social, economic, and political importance. However, when compared with other disciplines long linked with human well-being, such as medicine, chemistry, and physics, the technical tools and instruments of the ecologist have generally lagged behind those of the others. This disparity is beginning to be overcome with the increasing use of biotelemetric techniques, microtechnologies, satellite and airborne imagery, geographic information systems (GIS), and both regional and global data networks. We believe that the value and efficiency of ecosystem studies can advance significantly with more widespread use of existing technologies, and with the adaptation of technologies currently used in other disciplines to ecosystem studies. More importantly, the broader use of these technologies is critical for contributing to the preservation of biodiversity and the development of sustainable natural resource use by humans. The concept of human management of biodiversity and natural systems is a contentious one. However, we assert that as human population and resource consumption continue to increase, biodiversity and resource sustainability will only be preserved by increasing management efforts—if not of the biodiversity and resources themselves, then of human impacts on them. The technologies described in this chapter will help enable better management efforts. In this context, biodiversity refers not only to numbers of species (i.e., richness) in an arbitrarily defined area, but also to species abundances within that area. Sustainability refers to the maintenance of natural systems, biodiversity, and resources for the benefit of future generations. Arid-land grazing systems support human social systems and economies in regions all over the world, and can be expected to play increasingly critical roles as human populations increase. Further, grazing systems represent a nexus of natural and domesticated systems. In these systems, native biodiversity exists side by side with introduced species and populations, and in fact can benefit from them.

Unified Framework II Ecosystem Processes: A Link Between Species and Landscape Diversity

The discipline of ecology can be subdivided into several subdisciplines, including community, ecosystem, and landscape ecology. While all the subdisciplines are important to the study of biodiversity, there is great variation in the extent to which their contributions have been analyzed. For example, the role of community ecology in biodiversity studies is well established. In community ecology, the entities of study are species that differ in their properties and generate a web of interactions that, in turn, organize the species into a community. Similar to community ecology, the contribution of landscape ecology to biodiversity is apparent. The entities of study, definable “patches,” are tangible. They differ in their properties and generate a web of interactions that organize the patches into a landscape mosaic. In contrast to community and landscape ecology, the role of ecosystem ecology in biodiversity is less apparent. In ecosystem ecology, it often is not clear what the entities are, and how they are organized. To the extent that ecosystem ecology focuses on energy flow and nutrient cycling, we can define fundamental entities as compartments and vectors in models that depict the flows of water, energy, and nutrients through communities. If we apply diversity criteria to these entities, we can use the term ecosystem diversity to refer to the number of compartments and vectors, the differences among them in type and size, and their organization in promoting energy flow or nutrient cycling. To our knowledge, ecosystem scientists have not yet developed criteria for ecosystem diversity similar to those used for species and landscape diversity. There has been some use of the term “ecosystem diversity” to refer to a diversity of ecosystems, implying a variety of habitats, landscapes, or biomes. As discussed above, we suggest that to define the role of ecosystem ecology in biodiversity studies, the approach should be to study the relationships among species, landscape, and ecosystem diversities (chapters 1 and 13). However, since the concept of ecosystem diversity awaits further development, we adopt a different approach for understanding the role of ecosystem science in biodiversity studies. In this chapter, we examine relationships among ecosystem processes, species diversity, and landscape diversity.

Biodiversity Along Core–Periphery Clines

The study of biodiversity has received wide attention in recent decades. Biodiversity has been defined in various ways (Gaston and Spicer, 1998, Purvis and Hector 2000, and chapters in this volume). Discussion regarding its definitions is dynamic, with shifts between the more traditional emphasis on community structure to emphasis on the higher ecosystem level or the lower population levels (e.g., chapters in this volume, Poiani et al. 2000). One of the definitions, proposed in the United Nations Convention on Biological Diversity held in Rio de Janeiro (1992) is “the diversity within species, between species and of ecosystems.” The within-species component of diversity is further defined as “the frequency and diversity of different genes and/or genomes . . .” (IUCN 1993) as estimated by the genetic and morphological diversity within species. While research and conservation efforts in the past century have focused mainly on the community level, they have recently been extended to include the within-species (Hanski 1989) and the ecosystem levels. The component comprising within-species genetic and morphological diversity is increasingly emphasized as an important element of biodiversity (UN Convention 1992). Recent studies suggest that patterns of genetic diversity significantly influence the viability and persistence of local populations (Frankham 1996, Lacy 1997, Riddle 1996, Vrijenhoek et al. 1985). Revealing geographical patterns of genetic diversity is highly relevant to conservation biology and especially to explicit decision-making procedures allowing systematic rather than opportunistic selection of populations and areas for in situ protection (Pressey et al. 1993). Therefore, studying spatial patterns in within-species diversity may be vital in defining and prioritizing conservation efforts (Brooks et al. 1992). Local populations of a species often differ in the ecological conditions experienced by their members (Brown 1984, Gaston 1990, Lawton et al. 1994). These factors potentially affect population characteristics, structure, and within-population genetic and morphological diversity (Brussard 1984, Lawton 1995, Parsons 1991). The spatial location of a population within a species range may be related to its patterns of diversity (Lesica and Allendorf 1995). Thus, detecting within-species diversity patterns across distributional ranges is important for our understanding of ecological and evolutionary (e.g., speciation) processes (Smith et al. 1997), and for the determination of conservation priorities (Kark 1999).

Toward a Unified Framework in Biodiversity Studies

Biodiversity has been defined as the “full variety of life on Earth” (Takacs 1996). Because biodiversity refers to a fundamental property of ecological systems, it continues to develop as a scientific frontier for exploration and discovery. Ecological systems are a mixture, or diversity, of living and nonliving entities interconnected by a web of interactions. Biodiversity is a concept used to describe the number, variety, and organization of entities in the biosphere, or a unit of the biosphere, and their relationships to each other (Gaston 1996). More importantly, biodiversity includes consideration of processes that create and maintain variation in ecological systems (Groombridge 1992). Therefore, biodiversity is not simply about entities and taxonomy. Rather, biodiversity is concerned with the diversity of species within communities, the range of ecological processes within ecosystems, and the diversity of ecosystem processes across landscape mosaics (Heywood 1994, Levin 1997). A similar perspective on biodiversity is provided by Noss (1990), who applies the concept over hierarchical levels of organization ranging from the gene to the entire biosphere, and recognizes compositional, structural, and functional approaches to biodiversity. In contrast to this broad and comprehensive scope of biodiversity suggested above, researchers and conservationists often employ a narrow definition of biodiversity shaped by their values, interests, and goals. A more precise and more widely recognized interpretation of the scope of biodiversity is required in order to promote scientific research and to develop management programs (Christensen et al. 1996). The objective of this chapter is to build on the comprehensive framework for biodiversity proposed in the introduction (chapter 1, fig.1.1) and combine it with additional insights from the other chapters. The initial framework recognizes biodiversity to include four components, and one complex output. The components are the roster of entities, the number of each kind of entity, the nature and degree of difference between the entities, and the spatial or functional organization of the entities. The outcome is the effect of biodiversity on ecosystem processes. This framework enables us to (1) map, in a more unified way, the present achievements of research in the field of biodiversity, (2) understand the complexity of biodiversity, and (3) advance the scientific basis for biodiversity management.

Management for Biodiversity: Human and Landscape Effects on Dry Environments

Biodiversity is one of the principal pillars of natural ecosystems. In fact, biodiversity can be interpreted as a manifestation of the various biotic and abiotic components of the ecological systems and their mutual interactions or as the totality or variation (chapter 19). Biodiversity applies at different realms of ecological criteria: organism (genetic/phenological), species, habitat, and landscape (Loidi 1999, Noss 1990). Historically, it was the specific assemblage of organisms—the species diversity—that attracted the attention of scientists. Later, the effect of landscape structure on biological diversity, through habitat and niche properties, became an additional focus of biodiversity research (Malanson and Cramer 1999). The impact of different disturbances on the ecosystem and community structure has also become part of the study of landscape–biodiversity interrelationships (Moloney and Levin 1996; Trabaud and Galtie 1996). In this chapter we present a third dimension that affects biodiversity: human intervention through management and land-use patterns. One may consider this dimension as another source of disturbance, but we believe that such an approach is narrow. In contrast to disturbance, management is intentional, directional, goal-oriented, and, in some cases, scientifically or professionally guided. Human societies have modified the biodiversity of their environments since prehistoric time. Traditional land use that has evolved from ancient practice usually produces highly diverse landscapes based on knowledge of old systems of land exploitation (Loidi 1999). In Scotland, for example, biodiversity was enhanced by the interactions between farmers and the woodlands surrounding the agricultural fields (Tipping et al. 1999). Modern afforestation schemes fail to create diverse woodlands similar to the ancient ones. Maintenance of biodiversity through active management has recently become an important challenge for modern conservation (Monkkonen 1999). In this chapter, we use a conceptual model of human–biodiversity relationships and apply it to water-limited systems. The model describes how ecosystem services provided to traditional and modern societies are enhanced by management actions. The ecosystem services discussed in this chapter are water accumulation, food production (mainly through primary production), and recreation potential. The essence of the model is that without external input of water, ecosystem services are controlled by relationship between landscape mosaic, ecosystem processes, and organisms.

Resource Partitioning and Biodiversity in Fractal Environments with Applications to Dryland Communities

Arid and semiarid ecosystems (drylands) often contain a higher diversity of animals and plants than would be expected from their low productivity. High spatial heterogeneity of resources and physical habitats, exhibited at a wide range of spatial scales (Rundel 1996, Holling 1992, Peterson et al. 1998), may be a major factor explaining such high diversity. For example, at extremely small scales (<10 cm), branched plant material and various soil physical processes can create spatial niches for invertebrates, cyanobacteria, and other cryptogamic organisms (Lightfoot and Whitford 1991). At somewhat larger scales (<10 m), desert shrubs may aggregate water and organic material in “islands of fertility,” yielding a highly patchy heterogeneous distribution of resources (e.g., seeds, water) for other plants and animals (Gibbens and Beck 1988, Halvorson et al. 1997, chapter 13 this volume, chapter 11 this volume). At even larger scales (>100 m), soil erosion patterns create topographic variation that locally concentrates available water and nutrients, yielding a marked heterogeneity in the distribution of productivity across the landscape (Milne 1992). These heterogeneous distributions of physical environments, biotic material, and resources are likely to have strong effects on biodiversity. Ecologists have long associated greater spatial heterogeneity with higher species diversity (MacArthur 1964; Brown 1981; May 1988). Within a particular physical environment (habitat), this association exists presumably because collections of species that use similar resources, or “guilds,” can coexist whenever they can more finely divide up space and different-sized resource “packages” (Hutchinson and MacArthur 1959, Brown 1981, 1995, Morse et al. 1985, Peterson et al. 1998). The partitioning of space and different resource patches may be constrained by the different body sizes of species within guilds (Hutchinson and MacArthur 1959, Morse et al. 1985, Belovsky 1986, 1997, Brown 1995, Siemann et al. 1996). However, the mechanism by which body size and spatial heterogeneity of habitats and resources determine species diversity remains unclear (May 1988, Brown 1995, Siemann et al. 1996, Belovsky 1997). Resource partitioning and spatial heterogeneity therefore may strongly influence diversity in drylands, where, for example, well-known guilds of granivorous vertebrates and invertebrates are structured by competition for different sizes of seeds and seed patches (Brown et al. 1979, Davidson et al. 1980, 1985).

How Can High Animal Diversity Be Supported in Low-Productivity Deserts?: The Role of Macrodetritivory and Habitat Physiognomy

On large spatial scales, species diversity is typically correlated positively with productivity or energy supply (Wright et al. 1993, Huston 1994, Waide et al. 1999). In line with this general pattern, deserts are assumed to have relatively few species for two main reasons. First, relatively few plants and animals have acquired the physiological capabilities to withstand the stresses exerted by the high temperatures and shortage of water found in deserts (reviewed by Noy-Meir 1974, Evenari 1985, Shmida et al. 1986). A second, more ecological mechanism is resource limitation. In deserts, the low and highly variable precipitation levels, high temperatures and high evapotranspiration ratios limit both plant abundance and productivity to very low levels (Noy-Meir 1973, 1985, Polis 1991d). This lack of material at the primary producer level should exacerbate the harsh abiotic conditions and reduce the abundance of animals at higher trophic levels by limiting the types of resources and their availability. Animal abundance should be even further reduced because primary productivity is not only low, but also tends to be sporadic in time and space (MacMahon 1981, Crawford 1981, Ludwig 1986). Herbivores should have difficulties tracking these variations (e.g., Ayal 1994) and efficiently using the available food resources. Hence, herbivore populations in deserts have low densities relative to other biomes (Wisdom 1991) and most of the primary productivity remains unused (Crawford 1981, Noy-Meir 1985). This low abundance of herbivores should propagate through the food web and result as well in lower abundance of higher trophic levels. The number of individuals and the number of species are not always positively correlated; in particular, some examples of low diversity at high productivity with high densities are well documented (e.g., salt marshes, reviewed by Waide et al. 1999). However, several distinct mechanisms have led to the expectation that when productivity and the number of individuals are low, the number of species is also likely to be low. First, within trophic levels, the “statistical mechanics” model of Wright et al. (1993) may operate. In this model, the amount of energy present determines the probability distribution of population sizes for the members of the species pool in a region.

The Effects of Grazing on Plant Biodiversity in Arid Ecosystems

Conventional wisdom views heavy grazing as the major cause of desertification in semiarid and arid areas of Africa, Asia, and Australia (see, e.g., Acocks 1953, Jarman and Bosch 1973, Sinclair and Fryxell 1985, Middleton and Thomas 1997). Nowhere is the effect of heavy grazing more apparent than in the Sahel of Africa (Sinclair and Fryxell 1985). This land denudation has resulted in a negative feedback loop via decreased soil nutrient status and increased soil albedo (due to lower vegetation cover), causing increased evaporation and decreased precipitation, which in turn reduces the stocking capacity of the land, further exacerbating the negative effects of grazing (Schlesinger et al. 1990). A less dramatic result of overgrazing is a long-term decline in agricultural productivity. For example, the arid Karoo region of South Africa has experienced no climatic change over the last two centuries, yet there has been a 50% decline in stocking rates in seven of eight magisterial districts from 1911 to 1981 (Dean and McDonald 1994). These authors ascribe this decline to heavy grazing that reduced palatable plant populations and hence the carrying capacity of the vegetation in the long term. These examples of the negative effects of grazing in arid ecosystems lie in stark contrast with a large number of African studies that compared the effects of commercial (privately owned) and communal (subsistence, no private ownership) ranching on vegetation and soils (e.g., Archer et al. 1989, Tapson 1993, Scoones 1995, Ward et al. 1999a,b, reviewed by Behnke and Abel 1996). In spite of 5–10-fold higher stocking rates on communal ranches, few studies have shown differences in effects on biodiversity, plant species composition and soil quality between these ranching types (Archer et al. 1989, Tapson 1993, Scoones 1995, Ward et al. 1999a,b—fig. 14.1). Similarly, studies of grazing in Mediterranean semiarid grasslands (reviewed by Seligman 1996) and Middle Eastern arid rangelands (Ward et al. 1999b) show that the effects of grazing on biodiversity are relatively small. A consensus has developed in recent years that arid grazing ecosystems are nonequilibrial, event-driven systems (see, e.g., O’Connor 1985, Venter et al. 1989, Milchunas et al. 1989, Parsons et al. 1997).

Species Diversity and Ecosystem Processes in Water-Limited Systems

There are many relationships between ecosystem properties and species (Jones and Lawton, 1995) with the potential links described by five hypotheses: 1. The null hypothesis claims that there is no effect of species diversity on ecosystem processes. The following hypotheses imply biological mechanisms. 2. The diversity–stability hypothesis predicts that ecosystem productivity and recovery increase as the number of species increases (Johnson et al. 1996). 3. The rivet hypothesis predicts a threshold in species richness, below which ecosystem function declines steadily and above which changes in species richness are not reflected by changes in ecosystem function (Ehrlich and Ehrlich 1981; Vitousek and Hooper 1993). 4. The redundant species hypothesis states that species loss has little effect on ecosystem processes if the losses are within the same functional group (Walker 1992) 5. The idiosyncratic response hypothesis suggests that as diversity changes so do ecocosystem processes (Lawton 1994, Lawton and Brown 1994). There have been both field and laboratory attempts to test these hypotheses, (Naeem and Li 1998), however, the interpretation and the generality of the results remain contentious (Tilman 1999). A fundamental reason for such uncertainty is that the hypotheses are not driven by a comprehensive theory of the relationship between species properties and ecosystem processes (Tilman et al., 1997). We propose that the foundations for the necessary theory are in models of the distribution of resources and their utilization by organisms. This is because ecosystem processes such as primary production, decomposition, mineralization, and evapotranspiration are dependent on the processing of resources by the species that are producers, consumers, and decomposers. A theory that links the direct participation of species in ecosystem processes may resolve differences among the various hypotheses or identify how they complement each other. From a community perspective, a theory of resource utilization is based on two alternative assumptions: 1. The rate of ecosystem processes is determined by the few species that are most efficient in using and converting resources. For example, in a desert system, dominant species are those that are proficient in using water for biomass production or in converting inorganic matter into organic materials.

Unified Framework I: Interspecific Interactions and Species Diversity in Drylands

The goal of this chapter is to delineate how abiotic conditions, regional processes, and species interactions influence species diversity at local scales in drylands. There is a very rich literature that bears on this topic, but here we focus on mechanisms that promote or constrain local diversity and ask how these factors apply to deserts. We ask, “What is different about deserts, relative to other habitats, in their patterns of diversity, temporal variability in productivity, and spatial heterogeneity?” We assess how such differences might modify extant theory, and sketch relevant examples. Compared with other biomes, productivity, population densities, and community biomass are much lower in deserts, and temporal heterogeneity is typically higher. Do these differences imply distinct ecological processes and patterns in deserts? Or, do processes operate in deserts in similar ways as in tropical forests or grasslands? For example, it is often assumed that abiotic factors are more important in deserts. If so, how do abiotic factors modify biotic interactions? How do we integrate physical and biotic interactions? More generally, we ask what should be the main goals and approaches of a research program to understand the role of species interactions in determining community structure in drylands, as modified by abiotic factors and regional processes. . . . What Is Different About Drylands? . . . Deserts are traditionally perceived as relatively simple ecosystems harboring low species diversity. Yet increasing evidence suggests that desert communities can be highly diverse and complex. To our knowledge the only systematic analysis of the relative diversity in desert versus nondesert communities was compiled by Polis (1991a). These data suggest that patterns differ widely among taxonomic groups. In some cases, deserts support high diversity, comparable to or even higher than nonarid areas (see Polis 1991b). For example, while avian (Wiens 1991) and anuran (Woodward and Mitchell 1991) diversities are low compared with other biomes, desert annual plants show extremely high species diversity (Inouye 1991). Ants, succulent plants, lizards, scorpions, and tenebrionid beetles also have relatively high diversity in deserts (Polis 1991a–c, Wiens 1991). But, while very high diversity may occur, local diversity varies greatly in space and time (e.g., ants and annual plants: Danin 1977, Inouye 1991, MacKay 1991).