We demonstrate collapse of niche width in a generalist predator because of ecosystem fragmentation using a novel metric based on stable isotope ratios. It has been proposed that stable isotopes may provide a robust measure of niche width (Bearhop et al. 2004
; Layman et al. 2007
), and we provide the first quantitative measure to this end. Because the measure reflects relative variation both in trophic position and diversity of resource use, it may be a powerful approach in the study of both intraspecific and community-wide trophic structure. In the present study, niche width (i.e. TA) of grey snapper populations is significantly smaller in fragmented ecosystems because of contraction along both δ13
C and δ15
N niche dimensions. We discuss the underlying ecological reasons that drive niche width collapse, and explore implications of this collapse for the stability of the food webs in fragmented habitats.
Niche width collapse tracks the decrease in potential prey diversity in fragmented systems measured by either taxonomic or functional classifications. This reduction in prey diversity is the direct and indirect result of four ecosystem-level effects induced by fragmentation. First, ecosystem fragmentation disrupts metapopulation dynamics of prey populations, thereby reducing the supply of larvae, juveniles and adults to upstream areas (Gonzalez et al. 1998
; Pringle 2006
). Second, overall ecosystem size (volume of water and/or surface area of inudated wetland) can be reduced because of shallower water depths and increased sedimentation that fragmentation induces. Third, variation in physico-chemical conditions increases, especially in temperature and salinity, because of a lack of tidal flushing (Valentine et al. 2007b
). And fourth, partially due to each of the preceding three factors, there is an overall decline in abundance of structural flora (e.g. seagrasses and macroalgae) that provide important habitat for a diverse suite of estuarine and marine fauna (Adams et al. 2006
; McCall & Rakocinski 2007
). Isolating the relative individual contribution of each of these mechanisms with respect to particular impacts on food web structure is difficult, but one net result, i.e. an overall reduction in prey diversity, is obvious and significant.
Reductions in prey diversity in fragmented ecosystems contribute directly to niche width collapse of top predators because predator individuals are constrained in their ability to choose among potential prey items. As such, in fragmented ecosystems, there is less scope for either chance (e.g. patchy prey distribution) or specialization (Bolnick et al. 2003
) to result in individuals that exploit different sub-sets of available prey taxa. Variation in resource use among individuals is reflected by greater intraspecific variability in isotopic signatures (e.g. Cross Harbour population in ). But when prey diversity is low, all top predator individuals (both conspecifics and other species) must exploit the same taxa, thereby collapsing intraspecific niche variation (Marsh Harbour population in ). This phenomena may be widespread following ecosystem fragmentation, as well as following other anthropogenic impacts, because biodiversity loss typically follows such disturbances (Fahrig 2003
In our example, collapse in niche width was a function of reduction in diversity of basal resources supporting the food webs (represented by variation in δ13
C), as well as a reduction in trophic level variation (δ15
N) among organisms in fragmented food webs. Whereas terrestrial fragmentation does not necessarily change underlying characteristics of interior portions of fragmented patches (Fahrig 2003
), aquatic ecosystem fragmentation, by definition, intrinsically alters overall physico-chemical characteristics (Pringle 2001
). The most well-acknowledged instance of this phenomena is in freshwater ecosystems where dams alter hydrologic connectivity and subsequently affect food web structure and ecosystem function in both up- and downstream areas (Poff et al. 1997
). In the present example, fragmented estuarine tidal creeks have altered upstream physico-chemical conditions, increased sedimentation and altered nutrient cycling, with the result being a net decrease in basal resource diversity in fragmented ecosystems. With fewer primary consumer niches available in fragmented creeks (e.g. no seagrass epiphytes or macroalgae on which to feed), there is an overall reduction in primary consumer diversity and thus fewer potential prey for top predators. This phenomena is represented by the reduction in community-wide δ13
C range in fragmented systems (), and is reflected in the decreased niche width of the top predator species.
Niche width collapse also is related to decreased variability in δ15
N within food webs of fragmented creeks. For example, in the two representative webs depicted in , the range of δ15
N is 2.9‰ larger in the unfragmented system. According to assumed values of fractionation for trophic transfers within food webs (e.g. 3.4 or 2.5, respectively, Post 2002b
; Layman et al. 2005
), the difference in δ15
N range approximates a full trophic level of variation. As such, there seems to be c
. 1 additional trophic level in an unfragmented ecosystem. This observation is supported by functional classifications of prey items, as unfragmented systems support intermediate predators (e.g. pelagic planktivores) that are not available in fragmented systems. This pattern is consistent with the ‘insertion mechanism’ that can affect relative trophic position of top predators through the addition of an intermediate link in a food chain (Post et al. 2000b
; Post 2002a
). We extend this idea to suggest that the insertion mechanism may affect the trophic position of individuals within
a population of conspecifics. That is, if particular individuals specialize (sensuBolnick et al. 2003
) on a prey item that is only found in unfragmented creeks, those individuals may have a higher trophic position within the web than is possible for conspecifics in a fragmented system. This partially drives the overall increase in population niche width in unfragmented systems.
Other studies have found that population niche width expands when resources become scarce or intraspecific competition otherwise increases (Wilson & Turelli 1986
; Bolnick 2001
; Svanbäck & Persson 2004
). Such studies are typically conducted in systems with a variety of alternative resources among which consumers could choose, e.g., in lakes where fish can utilize distinct pools of benthic and limnetic resources (Svanbäck & Persson 2004
). In such systems, consumers can switch to suboptimal resources when preferred prey become scarce. Yet in the instance considered herein, predators do not have a distinct alternative prey resource to exploit because diversity of prey decreases with fragmentation, as is reflected in the representative fragmented food web in . When anthropogenic impacts significantly reduce available resource pools, population niche width expansion cannot ensue in response to resource scarcity because prey choice is so constrained.
Based on > 3500 individual isotope samples, hundreds of stomach content analyses, and extensive quantitative floral and faunal surveys, we are able to depict conceptualized energy flow pathways in typical fragmented and unfragmented creek systems (). Three critical observations can be derived from these conceptual trophic architectures. First, generalist top predators (grey snapper, as well as other species such as barracuda, needlefish and tarpon) have intermediate positions along a δ13
C niche axis, as they integrate across multiple energy flow pathways (Vander Zanden & Vadeboncoeur 2002
; Rooney et al. 2006
). By feeding on a diverse suite of prey with variable trophic roles, their resulting position along the δ13
C niche axis is intermediate relative to other taxa within the web. Second, a more narrow range of consumer values along the δ13
C axis in the fragmented systems is a direct function of a single basal resource pool (microalgal mat) dominating the substrate. This pool serves as the ultimate source of carbon for all consumer species. Third, in unfragmented systems, hundreds (and perhaps thousands) of unique energy flow pathways connect the base of the food web to top predators, because of the increased diversity of basal resource pools, greater abundance of intermediate prey taxa, and increased incidence of omnivory. In fragmented systems, there is a drastic overall simplification of energy flow pathways and web architecture.
Figure 4 Conceptual depictions of the ‘sink’ food webs (i.e. all energy flow pathways that culminate in grey snapper) in a typical fragmented and unfragmented creek system, as based on > 3500 individual isotope samples, hundreds of stomach (more ...)
The importance of numerous, heterogeneous, energy flow pathways has been implicated as a primary mechanism stabilizing food webs (McCann et al. 1998
; Post et al. 2000a
; Rooney et al. 2006
). Energy flow ‘asymmetry’ occurs when top predators utilize multiple pathways that differ in productivity and turnover rate, thereby coupling energy flows that cycle at different temporal scales and providing stability to food webs (Rooney et al. 2006
). Niche width collapse is a direct indication that such asymmetry is lost in fragmented habitats. In fragmented systems, pathways of energy flow are constrained to originate with a single basal resource pool and pass through fewer intermediate consumers. The end result is an overall homogenization in energy flow pathways and ultimately a less stable food web structure (sensuRooney et al. 2006
Even though many generalist species can persist in highly fragmented ecosystems, ecological roles of these species may be altered significantly. Niche width collapse of the generalist species results from an overall simplification of food web structure which may, in turn, render the top predators more susceptible to population fluctuations and more likely to face extinction through time (Tilman et al. 1994
). Because alteration in the role of top predators can have cascading effects throughout entire food webs (Terborgh et al. 2001
), collapse in niche width ultimately can have fundamental effects on ecosystem function. With continuing erosion of local and global biodiversity, ecologists should continue to move beyond documenting the presence/absence of species in altered ecosystems, and instead work to better understand how anthropogenic impacts intrinsically alter ecological roles of species.