Genetically informed ecological niche models improve climate change predictions

Abstract

Background: Although pelagic seabirds are broadly recognised as indicators of the health of marine systems, numerous gaps exist in knowledge of their at-sea distributions at the species level. These gaps have profound negative impacts on the robustness of marine conservation policies. Correlative modelling techniques have provided some information, but few studies have explored model development for non-breeding pelagic seabirds. Here, I present a first phase in developing robust niche models for highly mobile species as a baseline for further development. Methodology: Using observational data from a 12-year time period, 217 unique model parameterisations across three correlative modelling algorithms (boosted regression trees, Maxent and minimum volume ellipsoids) were tested in a time-averaged approach for their ability to recreate the at-sea distribution of non-breeding Wandering Albatrosses (Diomedea exulans) to provide a baseline for further development. Principle Findings/Results: Overall, minimum volume ellipsoids outperformed both boosted regression trees and Maxent. However, whilst the latter two algorithms generally overfit the data, minimum volume ellipsoids tended to underfit the data. Conclusions: The results of this exercise suggest a necessary evolution in how correlative modelling for highly mobile species such as pelagic seabirds should be approached. These insights are crucial for understanding seabird-environment interactions at macroscales, which can facilitate the ability to address population declines and inform effective marine conservation policy in the wake of rapid global change. ABSTRACT Boosted regression trees; digital accessible knowledge; distribution modelling; Maxent; minimum volume ellipsoids; pelagic seabird distribution; Diomedea exulans KEYWORDS

Forecasting range shifts of a cold‐adapted species under climate change: are genomic and ecological diversity within species crucial for future resilience?

“…Despite the fact that comparative empirical studies at the intraspecific level are lacking for cold-adapted species, equivalent studies across a range of plant groups suggest that taxa occupying a broader niche will be less affected by future climate change than taxa restricted to narrower niches (Thuiller et al 2005, Broennimann et al 2006, Gottfried et al 2012, Duputié et al 2015, Carrillo-Angeles et al 2016. region-or lineage-specific modeling), also demonstrated differential responses at the intraspecific level (Pearman et al 2010, D'Amen et al 2013, Ikeda et al 2017. These lineages may be adapted to local climatic conditions, thus their potential for adaptation to future climatic conditions may vary greatly among them (Hereford 2009, Benito-Garzón et al 2011, Shaw and Etterson 2012.…”

“…However, predicting the extent and directionality of future distributional shifts remains challenging. Moreover, spatial variation in genetic and ecological diversity differentially affects locally adapted populations within species, but such variation has been incorporated in future predictions only recently (Fitzpatrick and Keller 2015, Gotelli and Stanton-Geddes 2015, Moran et al 2015, Hällfors et al 2016, Ikeda et al 2017. Moreover, spatial variation in genetic and ecological diversity differentially affects locally adapted populations within species, but such variation has been incorporated in future predictions only recently (Fitzpatrick and Keller 2015, Gotelli and Stanton-Geddes 2015, Moran et al 2015, Hällfors et al 2016, Ikeda et al 2017.…”

“…Most of these phylogeographic lineages differ substantially from each other both in the environmental conditions they currently experience and in their adaptive potential to changing environments (Hereford 2009), thus predictions about their persistence that do not account for such differences may fail to identify potential risks or benefits of climate change to each lineage (Pearman et al 2010, Ikeda et al 2017. region-or lineage-specific approach; D' Amen et al 2013, Jeffress et al 2013, Schwalm et al 2016, Ikeda et al 2017. region-or lineage-specific approach; D' Amen et al 2013, Jeffress et al 2013, Schwalm et al 2016, Ikeda et al 2017.…”

Spatial and temporal heterogeneity in climate change limits species' dispersal capabilities and adaptive potential

“…Nonetheless, the theoretical models used to predict the consequences of global climate change are developing rapidly and employing an increasingly diverse array of information to forecast changes in population dynamics (Bush et al 2016). For instance, genetically and physiologically informed ecological niche models have recently been developed and are a major improvement, as they incorporate information on population genetic structure and variation in levels of local adaptation into projections of future range shifts (Ikeda et al 2017). Similarly, efforts to make use of spatially explicit population data, such as avian abundance and population trend information generated by the Breeding Bird Survey and Christmas Bird Count, are an important development that can help identify vulnerable local populations (Langham et al 2015).…”

A novel tool to assess the effect of intraspecific spatial niche variation on species distribution shifts under climate change

“…Niche-based models are then used to explore potential shifts in species distributions under future climate conditions (Elith, Kearney, & Phillips, 2010;Franklin, 2013). For instance, the most frequently used approaches consider a species as a fixed entity in the modelling process and assume that all the populations respond to changing climate conditions in a similar way throughout the occupied geographical range (Ikeda et al, 2017;Moran, Hartig, & Bell, 2016;Pearman, Guisan, Broennimann, & Randin, 2008). For instance, the most frequently used approaches consider a species as a fixed entity in the modelling process and assume that all the populations respond to changing climate conditions in a similar way throughout the occupied geographical range (Ikeda et al, 2017;Moran, Hartig, & Bell, 2016;Pearman, Guisan, Broennimann, & Randin, 2008).…”