Gene-flow within a butterfly metapopulation: the marsh fritillary Euphydryas aurinia in western Bohemia (Czech Republic)

Marius Junker,Kamil Zimmermann,Thomas Schmitt,Martin Konvicka

doi:10.1007/s10841-021-00325-8

Abstract

In human-altered landscapes, species with specific habitat requirements tend to persist as metapopulations, forming colonies restricted to patches of suitable habitats, displaying mutually independent within-patch dynamics and interconnected by inter-colony movements of individuals. Despite intuitive appeal and both empirical and analytical evidence, metapopulations of only relatively few butterfly systems had been both monitored for multiple years to quantify metapopulation dynamics, and assayed from the point of view of population genetics. We used allozyme analysis to study the genetic make-up of a metapopulation of a declining and EU-protected butterfly, Euphydryas aurinia, inhabiting humid grasslands in western Czech Republic, and reanalysed previously published demography and dispersal data to interpret the patterns. For 497 colony x year visits to the 97 colonies known at that time, we found annual extinction and colonisation probabilities roughly equal to 4%. The genetic diversity within colonies was intermediate or high for all assessed parameters of population genetic diversity and hence higher than expected for such a habitat specialist species. All the standard genetic diversity measures were positively correlated to adult counts and colony areas, but the correlations were weak and rarely significant, probably due to the rapid within-colony population dynamics. Only very weak correlations applied to larval nests numbers. We conclude that the entirety of colonies forms a well-connected system for their majority. Especially in its core parts, we assume a metapopulation structure with a dynamic equilibrium between local extinction and recolonization. It is vital to conserve in particular these structures of large and interconnected colonies.Implications for insect conservation: Conservation measures should focus on considering more in depth the habitat requirements of E. aurinia for management plans and on stabilisation strategies for colonies, especially of peripheral ones, e.g. by habitat restoration.

Highlights

Metapopulations are sets of populations inhabiting spatially separated habitat patches and displaying mutually independent within-patch dynamics, including local extinctions, which are counterbalanced by recolonisation from occupied patches (Hanski 1999; Colombo and Anteneodo 2015)
The metapopulation dynamics was interpreted with emphasis on demographic and environmental processes with little consideration of intraspecific genetic variation and natural selection (Hanski 1999), apart from inbreeding depression reducing population growth in small populations (e.g. Saccheri et al 1998; Madsen et al 1999)
Sedentary species tend to build up high density populations, whereas mobile species can persist under lower local abundances

Summary

Introduction

Metapopulations are sets of populations inhabiting spatially separated habitat patches and displaying mutually independent within-patch dynamics, including local extinctions, which are counterbalanced by recolonisation from occupied patches (Hanski 1999; Colombo and Anteneodo 2015). Saccheri et al 1998; Madsen et al 1999) This has gradually changed, and genetic processes in metapopulations are increasingly targeted (Habel and Schmitt 2012, 2018; Hanski et al 2017). Madsen et al 2000; Újvári et al 2002; Sigaard et al (2008); Habel et al 2012, 2019), whereas dispersal among habitat patches maintains a high within-population diversity (Schmitt et al 2003, 2005a, b; Habel et al 2010; Kramp et al 2016). Sedentary species tend to build up high density populations, whereas mobile species can persist under lower local abundances (cf Habel and Schmitt 2009; Konvicka et al 2012; for an overview see Habel and Schmitt 2018). Current population genetic patterns may reflect past spatial structures in many cases, rather than recent landscape patterns (e.g. Schmitt et al 2000; Orsini et al 2008)

Methods

Results

Conclusion