Ecological pork: Novel resources and the trophic reorganization of an ecosystem

Part of the attraction of ecology and its related intellectual challenges is its dynamic complexity. Darwin's (1) entangled bank metaphor captured the essence; it wasn't just that everything was connected to everything else but rather that specific relationships mattered. Darwin found that mowed turf contained approximately twice as many species as unmowed turf; a logical chain of influence was argued to link the probability of clover pollination and seed production to the negative influence of domestic cats on mice, themselves major predators of bumble bees doing the pollinating. Since Darwin's time an ever-increasing array of field experiments and observations has substantiated the dynamic consequences of these connections, especially in aquatic systems (2–6). These and many comparable examples share the common feature of highlighting the importance of apex predators and therefore top-down trophic control of population dynamics and community structure. In many cases, chains of indirect interactions among species with shared predators intensify top-down control. Effects are particularly acute when these relationships are asymmetric; one species may be a significant energy source for a predator without being regulated by it, thus strengthening the predatory impacts on other prey (7).

Change is commonplace in ecosystems with awareness heightened by the consequences of introductions of exotic or nonnative species or range expansions of those native to the system. The research contribution by Roemer et al. in this issue of PNAS (8) adds a novel dimension to the complexity of responses to biological invasions. The presence of feral, nonnative pigs seems to have enabled golden eagles to colonize some of California's Channel Islands, resulting in restructuring of an island food web and the approaching extinction of an endangered native fox. A hypothesis is proposed that native spotted skunks compete with foxes; piglets bolster eagle predation on foxes, thus releasing skunks from competitive suppression, leading to their recent numerical increase. The study is exceptional for its demonstration that birds, taxa that are often discounted within terrestrial ecosystems as important apex predators (9) that benefit from highly productive alternative prey (piglets), can increase to the point that their predation on native foxes becomes a significant conservation concern. Darwin's interest in chains of interactions and the related indirect positive or negative consequences once again is justified.

The presence of feral, nonnative pigs seems to have enabled golden eagles to colonize some of California's Channel Islands.

The Roemer et al. study is unusual on a number of grounds. Long-term interest in the biology of the Channel Islands provides an important historical perspective. The islands themselves could be perceived as replicates, although this attribute was minimally developed. The authors also support their conclusions with an intriguing and broad sampler of ecology's tools: natural history, including sniffing live-trapped golden eagles for the telltale odors of skunk encounters; a three trophic-level modeling effort based on Lotka-Volterra dynamics, which supports some but not all the population trajectories; intense long-term live-trapping and mark-recapture efforts that revealed the reciprocal trends in fox and skunk densities; diet studies of what this intriguing mix of carnivores eat; bioenergetics estimates of predation rates by interacting species; and stable isotope analyses (N and C) to suggest energy sources independent of the carcass count on which the eagle diet is based.

The temporal and spatial breadth (1993–1999, with two separate sites on Santa Cruz), and the essential commitment underlying this study are impressive—hundreds of hours trapping, marking, and releasing foxes and skunks and the endless details of determining diets and finding corpses. Similar to any detailed, condensed report, some questions remain unanswered, and others are roused. For instance, the stable isotope analyses suggest that the eagle diet consists of more than just piglets, foxes, and an occasional skunk. It seems likely that sea birds such as cormorants and pelicans, available and probably increasing since 1968 (10) following recovery from the DDT-induced eggshell thinning, provide alternative prey. The assumption that foxes and spotted skunks compete based solely on calculated dietary overlap and consumption rates is unconvincing, because it does not address whether the overlapping resources are in short supply and therefore competed for. In fact, is this premise even necessary to their argument? Their simulation model, although correctly portraying the inverse population trajectories of eagles and foxes, fails for skunks. Feral pigs have inhabited these islands for over 150 years. What delayed the colonization of golden eagles until 1994? In the early 1900s, the Channel Islands were home to vigorous populations of bald eagles. In fact, they became the quarry of island sheep farmers who were losing their lambs to marauding eagles (11). Were lambs bolstering bald eagles in ways similar to the piglet-golden eagle example? Observed agonistic encounters between these eagle species suggest no pattern of dominance (12). The long-delayed golden eagle invasion suggests unrecognized dimensions to their interesting story. For instance, island foxes seem especially vulnerable to ectoparasite infections (13) and even diseases spread by agricultural animals (11). And what hidden role do the pigs play? This species is widely documented to be environmentally destructive at most sites to which they've been introduced (14, 15). Many unanswered questions are stimulated by the case study described by Roemer et al.

What broader context best describes this contribution? One message is that intervention is possible; eagles can be live-trapped and translocated just like problem bears. If the authors' calculations and intuition are correct, fox populations should respond positively and rapidly to this manipulation. On the other hand, the bottom-up alternative of pig elimination is doubly attractive (but perhaps less practical or immediate); it would test directly one prediction of their simulation and remove an environmentally destructive introduced species. Thus, Roemer et al. pose an intriguing hypothesis to describe the ecological disruption of pigs on the relationships within a suite of vertebrate predators on California's Channel Islands.

The more dramatic problems in conservation today often involve ecological reorganization.

We offer two perspectives. Ecology is a discipline in which historical, deep, spatially localized context matters. Effective fishery management is thwarted often by debate over what stock baseline should be used (16). A similar observation could apply to the Channel Islands, on which in the early 1900s sheep and pigs were common, house cats ran wild, and foxes were “… not as plentiful as formerly,” perhaps because scabies, contracted from sheep, was causing them to go blind (11). Ecological change on these islands surely is ongoing, and it remains an open question whether the results of intensive study since 1993 are independent of the site's long history.

Perhaps the broadest perspective we can offer is that ecologists should be increasingly alert to the possibility that ecosystem properties can change abruptly, even catastrophically, and reversal/recovery may not be simple (17). Further, perturbation to almost any member of a system whether native or exotic should be suspect as a potential driver of ecosystem alteration. For instance, the recent increase of bald eagles on Washington state's outer coast bears dire implications for resident sea birds (18). A well intentioned stocking of opossum shrimp into Flathead Lake, Montana, led to collapse of zooplankton and salmon, decline in bear and eagle populations, and a dwindling number of fall visitors to Glacier National Park (19). Eutrophication of the Baltic Sea, which is providing limited nutrients to the benthic plant assemblage, has produced sweeping changes at higher trophic levels (20). Killer whales along the Aleutian chain, by feeding on sea otters (21), have begun to reverse the patterns of a classic trophic cascade (3).

The more dramatic problems in conservation today often involve ecological reorganization produced by resource subsidies that alter a species' role in community organization. For example, soil nitrification by invasive legumes in Hawaii enhances invasion success of other plants (22). Snow geese populations are on an ecological rampage in Hudson Bay because of the trophic bounty represented by agricultural crops in the Great Plains, causing vegetation damage so extensive that it is visible from space (23). Household and feral domestic cats kill millions of songbirds and small mammals annually (24). Not only are cats nonnative, many are lavishly subsidized, yet Crooks and Soule (25) have shown that when landscape parcels are sufficiently extensive, native coyotes can control cat numbers to the benefit of songbirds. These studies and many comparable ones featuring other vertebrates, especially fish but also invertebrates (4, 26), demonstrate all too clearly the dynamic pliability of ecosystems. Roemer et al. have illustrated yet another pathway contributing to change.

Were these results happenstance? Were they predictable? Are they reversible? These are concerns requiring greater appreciation and understanding of ecological complexity. Will more informative data, better models, and adaptive, novel management practices provide substantive guidance? The golden eagle-piglet-fox interaction should contribute to this lofty question.