Evaluating habitat suitability and connectivity for a recolonizing large carnivore

Heather Hemmingmoore,Malin Aronsson,Mikael Åkesson,Jens Persson,Henrik Andrén

doi:10.1016/j.biocon.2019.108352

Heather Hemmingmoore, Malin Aronsson + Show 3 more

Open Access

https://doi.org/10.1016/j.biocon.2019.108352

Copy DOI

Abstract

The conservation of wide-ranging species presents challenges in a world of intensified human land use, forcing animals to occupy and recolonize human-modified landscapes. Although identifying suitable habitat and ensuring connectivity are important in supporting natural recolonization, these actions are rarely validated due to difficulties in monitoring such events. In Sweden, the Eurasian lynx (Lynx lynx) is now recolonizing its former range, after centuries of persecution. We investigated resource selection based on telemetry data from 108 lynx monitored over 20 years. We assessed the differences between the established population in central Sweden and the recolonizing population in southern Sweden, and between established and dispersing individuals. We found that models based on central Sweden successfully identified core habitat patches for establishment in southern Sweden, validated after recolonization. We also found that lynx selected for higher habitat suitability during the recolonization phase, and that dispersing individuals were less selective than established lynx. Using cost-distance analysis, we assessed connectivity between central and southern Sweden, and found that landscape permeability was higher when based on dispersing lynx compared to established lynx. Altogether, our findings suggest that when landscapes are sufficiently similar between source and recolonization areas, resource selection information from an established population can be useful for managers seeking to facilitate recolonization of wide-ranging species. We recommend more frequent use of validation during and after recolonization events, to improve our common understanding of habitat suitability and connectivity modeling, and therefore to enable more active management of recolonization events.

Highlights

Some large carnivores are currently recovering in human-dominated landscapes (Chapron et al, 2014; Gantchoff and Belant, 2017; LaRue and Nielsen, 2016); there is a growing need to predict their establishment and to facilitate management actions that help ensure their long-term viability and mitigate potential conflicts (Redpath et al, 2013)
Large carnivore movements are increasingly constrained by human land use, forcing them to occupy or cross human-modified landscapes in search of suitable habitat (Fahrig, 2001; Fahrig, 2007; Tucker et al, 2018)
Suitable habitat patches are often identified from resource selection functions (RSFs) (Boyce et al, 2002; Manly et al, 2007), and the potential for connectivity is evaluated by assessing the cost of passing through different habitat types based on RSF results (Abrahms et al, 2017; Beier et al, 2008; Chetkiewicz and Boyce, 2009; Sawyer et al, 2011)

Summary

Introduction

Some large carnivores are currently recovering in human-dominated landscapes (Chapron et al, 2014; Gantchoff and Belant, 2017; LaRue and Nielsen, 2016); there is a growing need to predict their establishment and to facilitate management actions that help ensure their long-term viability and mitigate potential conflicts (Redpath et al, 2013). Suitable habitat patches are often identified from resource selection functions (RSFs) (Boyce et al, 2002; Manly et al, 2007), and the potential for connectivity is evaluated by assessing the cost of passing through different habitat types based on RSF results (Abrahms et al, 2017; Beier et al, 2008; Chetkiewicz and Boyce, 2009; Sawyer et al, 2011). As dispersal events often involve long distance movements (Fahrig, 2007; Tucker et al, 2018), and resource requirements may differ between established and dispersing individuals (Abrahms et al, 2017; Chetkiewicz et al, 2006), connectivity should ideally be assessed using actual dispersal events (Beyer et al, 2010). Studies on habitat connectivity often use data from resident individuals, which could overestimate landscape resistance when dispersers are more inclined than residents to use the lowquality matrix between core habitat patches (Abrahms et al, 2017; Jackson et al, 2016)

Methods

Results

Discussion

Conclusion