Rethinking the Center-Periphery Hypothesis

The intuitive and oft-cited center-periphery hypothesis (CPH) predicts that genetic diversity and demographic rates will decline when observing populations from the core to the edge of a species distribution. Although the idea of geographically marginal populations being less fit than core populations is deeply entrenched and often assumed to be true in the minds of many ecologists, a thorough analysis has been lacking to actually test the major tenets of CPH. Pironon et al. (2017) sought to remedy this by reviewing 200+ empirical studies on CPH. Remarkably, Pironon et al. found only limited support for CPH. Approximately half of the reviewed studies demonstrated a decrease in genetic variation in core vs. edge populations. Even more interesting, only 20% - 30% of studies found populations to exhibit decreasing demographic rates from core vs. edge populations. Therefore, a majority of the reviewed cases did not support the two primary postulates of CPH. The key takeaway from this review is that geographic marginality is often not the same as ecological marginality. A refinement of CPH is therefore in order. Pironon et al. suggest the integration of ecological and geographic factors as a way forward beyond the traditional CPH. For example, a species distribution during glaciation will likely have a different distribution of geographic and ecological cores and peripheries compared to its post-glacial distributions. Therefore, a population that was previously a geographic and ecological core may now be geographically marginal, but still an ecological core. The interactions between historical and present local ecological factors may provide a better explanation for the variation observed in population genetics and demographic rates as opposed to one factor or the other. The findings of Pironon et al. (2017) underline the necessity of thinking at smaller ecological scales and larger temporal scales if there is to be a more complete understanding of the mechanisms which lead to the emergence of geographic species distributions.

Citation:

Pironon S., Papuga G., Villellas J., Angert A.L., Garcia M.B., and Thompson J.D. (2017). Geographic variation in genetic and demographic performance: new insights from an old biogeographical paradigm. Biological Reviews. 92: 1877 - 1909.

Interactive Range Limit Theory

A number of different hypotheses exist to explain why we see species distribution limits where we do. A straightforward explanation is the species interaction abiotic stress hypothesis (SIASH), which states that as abiotic conditions become less stressful, biotic factors become more important for determining range limits. However, counter-examples to SIASH abound, showing instances where distributions do not follow the predictions of this hypothesis. Enter interactive range limit theory (iRLT). Proposed by Siren and Morelli (2020), iRLT builds extensions onto the predictions made by SIASH. Namely, iRLT allows for abiotic factors to enhance or lessen biotic stress at the low-abiotic-stress end of a distribution and for biotic factors to enhance or lessen abiotic stress at the high-abiotic-stress end of a distribution (Figure 1). This expansion on SIASH covers the most of the cases that have been noted to violate SIASH predictions. An interesting question which emerges is why do some species fall under the simpler framework of SIASH as opposed to the sometimes counter-intuitive iRLT? Is there a way we can group the species that conform to SIASH vs iRLT?

One important limitation of iRLT is that the evidence provided for it relies nearly exclusively on animal systems. While Siren and Morelli (2020) suggest iRLT should also apply to plants, evidence provided in their article is sparse. There are, however, plant examples given in other papers addressing range limit hypotheses which seem to suggest some species may conform to the patterns predicted by iRLT. A more direct test of iRLT in plants would be necessary to draw strong conclusions, though.

Citation:

Siren, A.P.K. and Morelli, T.L. (2020) Interactive range-limit theory (iRLT): An extension for predicting range shifts. Journal of Animal Ecology, 89(4), 940-954. https://doi.org/10.1111/1365-2656.13150

Figure 1 Highlighting the differences in the predictions made by SIASH versus iRLT. Adapted from Siren and Morelli 2020.

A potential counter-example to SIASH from distribution models

The Species Interaction - Abiotic Stress Hypothesis (SIASH) posits that, where abiotic conditions are stressful, species interactions become less important in determining distributions. Species distribution models are common tools ecologists use to make predictions about where a species is likely to occur. Heikkinen et al. (2007) tested the effect of inclusion of biotic and land-cover variables on distribution model performance for high-latitude owl species. While the express purpose of this study was not to test SIASH, it is an interesting case study which presents a potential counter-example to the position of SIASH.

In their study, Heikkinen et al. made a baseline distribution model for four owl species which contained only climatic predictor variables.. As treatments, the authors also created models where land-cover and biotic interaction predictor variables were included in addition to climatic variables. It was found that when land-cover and biotic interactions were included in models, performance of said models improved significantly. These results imply that biotic interactions were taking place amongst the boreal bird species modeled.

While the purpose of this paper was to demonstrate the general utility of including land-cover and biotic variables in models, there are also interesting implications from a SIASH perspective. Each of the bird species tested exist in the extremely high latitudes of Scandinavia. SIASH would predict that minimal important biotic interactions should occur at high latitudes, however the improvement of model predictions resulting from the inclusion of biotic interaction variables suggests otherwise. This study therefore demonstrates that there may be limits to generalizing SIASH across species. At the same time, caution must be applied when interpreting species distribution model outcomes in the context of SIASH, as the methods used provide a purely correlational assessment of the relationships between the boreal owls and their environments.

Paper reference:

Heikkinen R., Luoto M., Virkkala R., Pearson R., Korber J.H. (2007). Biotic interactions improve prediction of boreal bird distributions at macro-scales. Global Ecology and Biogeography. 16(6): 754-763.