RESEARCH PAPER Genetic variation and structure in the Mediterranean shrubs Myrtus communis and Pistacia lentiscus in different landscape contexts S. Nora, R. G. Albaladejo & A. Aparicio Departamento de Biolog  ıa Vegetal y Ecolog  ıa, Universidad de Sevilla, Sevilla, Spain Keywords Habitat fragmentation; insect-pollinated species; mating systems; mediterranean landscapes; population genetics; wind- pollinated species. Correspondence S. Nora, Departamento de Biolog  ıa Vegetal y Ecolog  ıa, Universidad de Sevilla, C/ Profesor Garc  ıa Gonz alez n° 2, 41012 Sevilla, Spain. E-mail: sofianora@us.es Editor R. Bekker Received: 12 July 2013; Accepted: 25 June 2014 doi:10.1111/plb.12242 ABSTRACT Studies concerning different habitat configurations can provide insights into the com- plex interactions between species’ life-history traits and the environment and can help to predict patterns in population genetics. In this study, we compared patterns of genetic variation in two Mediterranean shrub species (Myrtus communis and Pistacia lentiscus) that co-occur in populations within three contrasting landscape contexts: continuous, fragmented-connected and fragmented-isolated populations. Analysing variation at microsatellites loci, our results revealed weak responses to the landscape contexts. We rather found a population-specific response in both study species. How- ever, despite both study species sharing similar levels of genetic diversity, Myrtus dis- played higher levels of homozygosity and genetic differentiation among populations, stronger patterns of within-population spatial genetic structure, lower values of muta- tion-scaled effective population size and stronger evidence for recent genetic bottle- necks than Pistacia. This result highlights the influence of past events (e.g. historical connectivity, fluctuations in population size) and local factors (e.g. microhabitat avail- ability for recruitment, habitat quality, plant density, native fauna) and that the land- scape configuration per se (i.e. fragment size and/or isolation) might not completely determine the species’ genetic patterns. INTRODUCTION Fragmentation modifies the original configuration of natural habitats by reducing habitat extension and quality, and can eventually compromise the persistence of even common and naturally abundant species (Fahrig 2003; Honnay & Jacquemyn 2007). Aside from demographic and ecological effects (Leimu et al. 2010; Meirmans et al. 2011), fragmentation can also shape patterns of genetic diversity, both within and among populations, and influence the long-term survival of species in anthropogenic habitats (Sork & Smouse 2006). In this context, the ‘paradigm of small population genetics’ (Bacles & Jump 2011) predicts that a reduction in population size will ulti- mately lead to a decrease in genetic diversity, an increase in inbreeding (either complete inbreeding or biparental inbreed- ing) and individual homozygosity and, consequently, a reduc- tion in fitness (Ellstrand & Elam 1993; Young et al. 1996). Nevertheless, under the perspective of population genetics, the consequences of habitat fragmentation have to be examined within the interaction between species’ life-history traits (e.g. growth form, perenniality, seed and pollen dispersal, mating and breeding systems, vegetative reproduction, geographic range and longevity) that directly or indirectly affect their abil- ity to spread their genes, and the landscape context (habitat size, landscape-scale connectivity and matrix characteristics). In fragmentation studies it is therefore essential to consider all the complex mechanisms and contexts that either enhance or block gene flow (Bacles & Jump 2011). For example, the mating system, i.e. the ability to self, has been shown to be one of the drivers behind species’ vulnerabil- ity to fragmentation (Duminil et al. 2007; Aguilar et al. 2008). Indeed, self-compatible and self-incompatible species diverge in the way they retain genetic diversity, since self-compatible species tend to contain less genetic diversity within and more genetic differentiation between populations than self-incom- patible species (Ellstrand & Elam 1993; Hamrick & Godt 1996). Moreover, pollen dispersal modes are also responsible for how species react to human management and disturbance. Animal- pollinated species tend to be more prone to the effects of genetic drift in fragmented habitats, despite their inherent potential for the long-distance dispersal of pollen grains (Smouse & Sork 2004; Hughes et al. 2007). Finally, it is impor- tant to stress the role of historical events such as fluctuations in population size and past bottlenecks in shaping how genetic diversity is currently structured at different landscape scales (Vekemans & Hardy 2004; Bacles & Jump 2011). In contrast, although difficult to disentangle from species’ functional attributes, the landscape context also influences the effective population size and connectivity of populations, as well as their long-term genetic structure (Manel et al. 2003; Storfer et al. 2010). In fact, it has been stated that whenever populations are constrained to small remnants of habitat, they have high probability of becoming extinct due to environmen- tal or demographic stochasticity, natural catastrophes or reduced genetic diversity (Ewers & Didham 2006). Thus, cur- rently, habitat loss is considered to be one of the major causes Plant Biology © 2014 German Botanical Society and The Royal Botanical Society of the Netherlands 1 Plant Biology ISSN 1435-8603