Vertebrate invasions don’t seem to conform to the norm

Blog written by Marcus Lashley. Read the full article here.

Biological invasions are one of the biggest threats to biodiversity globally. Because of the pervasive threat invasions pose, understanding basic principles of invasion and how those invasions affect biodiversity is of primary interest to ecologist and conservationist. A fundamental scale-dependent relationship has been observed repeatedly across plant and invertebrate invasions. That is, invasion commonly suppresses biodiversity at small or local scales but as the spatial scale considered increases, biodiversity is either not affected or facilitated at large scales. This could happen for several reasons but most commonly, invaders regulate the dominance of a native species which can release other species in the same community ultimately resulting in the scale-dependent pattern. The scale-dependent pattern has been generalized to include vertebrate invasions but because they are often more difficult to study, to our knowledge the scale-dependence hypothesis has not been explicitly tested in a vertebrate system.

There are three basic requirements to test the hypothesis: 1) a vertebrate invasion, 2) an estimate of biodiversity with and without the vertebrate invader, and 3) an estimate of biodiversity with and without the invader across spatial scales. In the United States, one of the most problematic nonnative vertebrates is the feral pig (Sus scrofa). Feral pigs are problematic because of their foraging behavior and their taste for agricultural crops. Interestingly, in agroecosystems, land clearing often results in fragmentation of forests leaving relatively isolated forest fragments similar in composition but varying in size. Feral pigs use those forest fragments for cover while not foraging on surrounding crop fields. Thus, an agroecosystem where feral pigs have invaded may provide all three basic needs to test the scale dependence hypothesis if forest fragments have a predictable species-area relationship.

We established camera traps in 36 forest fragments ranging across four orders of magnitude in area to estimate species richness of forest fragments. With camera traps, we knew many species would not be detected, however, it is common to take a subset of biodiversity in these types of studies given it is nearly impossible to tally all species present. In all, we detected 41 species, including feral pigs in 11 of the forest fragments. We took the 25 forest fragments without feral pigs and ran a simple linear model to determine that the number of species detected was very well predicted by the area of the forest fragment. The number of species detected increased as the forest fragment area increased. In other words, based on the area of a forest fragment, we can accurately predict how many species should be detected. Using that species-area relationship as a basis for how many species should be detected in a forest fragment of a given area, we then used a similar model to evaluate the number of species detected in forest fragments where feral pigs had invaded. If feral pigs are negatively affecting species richness in a scale dependent pattern, we should expect the predicted species richness across scale to have two basic properties as it relates to uninvaded fragments: 1) the y-intercept of the line should be lower than in the uninvaded estimate and 2) the slopes of the lines should not be parallel. Indeed, in forest fragments where feral pigs had invaded the y-intercept was 26% lower indicating we detected 26% fewer species than should have been expected given the fragment area. The slopes of the two lines were parallel indicating this reduced species richness occurred across the range in forest fragment size when pigs had invaded. Collectively, those results do not support the scale dependence hypothesis in this vertebrate invasion calling into question how generalizable the scale dependence hypothesis is to vertebrate invasions.

Log–Log relationship between species richness and forest fragment area in the Mississippi Alluvial Valley invaded (solid line and solid points) and absent (broken line and hollow points) of feral swine.

In this study, we did not establish causation, so it would not be appropriate to assume that feral pigs cause the reduction in biodiversity. However, our observations are similar to other causative studies which show invasions reduce diversity by 19-27%. Moreover, feral pigs are competitors and predators of many of the native wildlife in this ecosystem, so it is plausible that they do suppress biodiversity through those mechanisms. Ultimately, more research is needed to determine if feral pigs are causing this reduction in biodiversity and to determine if the lack of scale dependence applies to other vertebrate invasions.