“Life is a train of moods like a string of beads, and, as we pass through them, they prove to be many-colored lenses which paint the world their own hue, and each shows only what lies in its focus.” Ralph Waldo Emerson, Experience, 1844.
The concept of looking at nature through multiple lenses to see different things is not new and has been long recognized. As always, the devil is in the details. Recent developments in analytical tools and the embracement of an integrative metapopulation concept and the newly emergent field of functional biogeography, are allowing exciting new insights to be made by population ecologists that have direct bearing on our understanding of the effects of environmental change on biodiversity patterns.
The metapopulation concept posits that isolated populations of organisms are connected through dynamics of dispersal and extinction. Across a landscape, areas of suitable habitat occur, which at one point in time may or may not host a viable population of a particular species. I study this concept with terrestrial plants, and have asked what environmental conditions determine suitable habitat for metapopulations.
Much of the foundational work in this topic was conducted on butterfly populations in meadows across otherwise forested habitat. Regardless of study organism, embracement of this concept has been enough to make population ecologists realize that studying single populations may give only a limited view on generalities of ecology and evolution. Indeed, taking this concept on board, has led population ecologists to want to predict in which areas of suitable habitat across the landscape a new population may establish.
“There’s no getting away from field work!”
There are obvious conservation and management implications that result from being able to predict the geographical distribution of a species, whether an invasive exotic spreading across the globe, or an endangered organism. Unfortunately, just knowing where a species or a group of species may occur across the landscape is not enough. Individuals in some populations may have low fitness and their populations may be barely hanging on. For some species such as potential island colonizers, it has been proposed that limited ability to colonize vacant habitat patches may be due to the occurrence of closely related species occupying a similar niche.
Important ‘missing pieces’ from a full understanding of the metapopulation puzzle have been through inclusion of population growth rate estimates and incorporation of species evolutionary relationships (i.e., their phylogenic ancestry). Population ecologists have been toiling away making fitness estimates of their species of interest in the field. Systematists, on the other hand, have been grinding it out in the lab to generate the molecular data necessary to construct phylogenetic trees to help classify their species.
Community ecologists studying multispecies assemblages, as a third-dimensional angle to this question, have been working with geographers to develop species distribution models. It is only recently that the analytical tools have emerged that allow these groups of scientists to collaborate and address questions of common interest about metapopulations.For example, Cory Merow and colleagues have recently shown how Bayesian models can be used to propagate uncertainty estimates in the application of integral projection models (IPMs) to forecast growth rates as part of predictive demographic distribution models (transition matrix models could also be used). In other words, species geographic distribution predictions can be improved by accounting for population-level fitness estimates.
In another study, Oluwatobi Oke and colleagues have shown how phylogenetic relationships among 66 co-occurring species in populations across a metapopulation structured landscape of Canadian barrens can improve understanding of species distribution patterns. The basis for Oke et al.’s phylogenetic patterns among their species was the large angiosperm supertree based upon nucleotide sequence data of three genes from over 500 species.
The basis for all of the work described above are precise and accurate estimates of individual fitness and population growth rates. There’s no getting away from field work! Methods for carrying out the field work component of these studies, to allow the use of modern statistical methods including Bayesian analysis, IPMs, and transition matrix models, have to be planned and carried out with care. We have come a long way in the last decade in enabling population studies to scale up to address fundamental questions at higher levels of the ecological hierarchy.
The field of population demography is moving fast. For example, the recent launch of the COMPADRE Plant Matrix Database, with accurate demographic information for over 500 plant species in their natural settings worldwide, will make addressing these scale-related types of comparative evolutionary and ecological questions even more tractable in the future.