Article published in Molecular Ecology, 2017, 26, 1498-1514.doi: 10.1111/mec.14019
1 Departamento de Biologia da Faculdade de Ciências da Universidade do Porto, Rua Campo
Alegre, 4169-007 Porto, Portugal.
2 CIBIO/InBIO, Centro de Investigacao em Biodiversidade e Recursos Genéticos da
Universidade do Porto, Instituto de Ciências Agrárias de Vairão, Rua Padre Armando Quintas 7, 4485-661 Vairão, Portugal
3 Ecology Unit, Department of Organisms and Systems Biology, University of Oviedo, C/
Catedrático Rodrigo Uría, 33071 Oviedo, Spain
4 Department of Environmental Science, Policy and Management, University of California, 130
2.1 – Abstract
Urbanization is a severe form of habitat fragmentation that can cause many species to be locally extirpated and many others to become trapped and isolated within an urban matrix. The role of drift in reducing genetic diversity and increasing genetic differentiation is well recognized in urban populations. However, explicit incorporation and analysis of the demographic and temporal factors promoting drift in urban environments are poorly studied. Here, we genotyped 15 microsatellites in 320 fire salamanders from the historical city of Oviedo (Est. 8th century) to assess the effects of time since isolation, demographic history (historical effective population size; Ne) and patch size on genetic diversity, population structure and contemporary Ne. Our results indicate that urban populations of fire salamanders are highly differentiated, most likely due to the recent Ne declines, as calculated in coalescence analyses, concomitant with the urban development of Oviedo. However, urbanization only caused a small loss of genetic diversity. Regression modelling showed that patch size was positively associated with contemporary Ne, while we found only moderate support for the effects of demographic history when excluding populations with unresolved history. This highlights the interplay between different factors in determining current genetic diversity and structure. Overall, the results of our study on urban populations of fire salamanders provide some of the very first insights into the mechanisms affecting changes in genetic diversity and population differentiation via drift in urban environments, a crucial subject in a world where increasing urbanization is forecasted.
Keywords: demography, genetic drift, genetic isolation, microsatellite, population effective
size, Salamandra salamandra.
2.2 – Introduction
Identifying the mechanisms underlying patterns of genetic diversity and structure arising from habitat loss and fragmentation continues to be a major target of ecological and evolutionary studies of natural populations (e.g. Weckworth et al. 2013; Rivera-Ortíz et al. 2015; Richardson et al. 2016). Among the many causes of habitat fragmentation, urbanization is regarded as one of the most rapid and pervasive drivers of landscape change, causing isolation and local extirpations of populations in numerous vulnerable species (McKinney 2006). Nevertheless, some species with small home range requirements are able to persist in urban environments (e.g. Noël and Lapointe 2010; Munshi-South 2012; Yamamoto et al. 2013; Rodriguez-Martínez et al. 2014; Beninde et al. 2016). These urban dwellers inhabiting high- contrast edge areas are usually confined within small remnant patches of vegetation (e.g. city
parks) surrounded by an impervious matrix of buildings and roads. Under this scenario of genetic isolation (i.e. reduced or absent gene flow) and small population size, genetic drift becomes a dominant force shaping allele frequencies, leading to a decrease in genetic diversity and an increase in genetic differentiation (Frankham 2005).
Recent empirical research in urban areas has shown indeed that urbanization significantly limits functional connectivity (gene flow) among urban populations, which, coupled with the stronger effects of drift in small populations, substantially increases genetic differentiation (Delaney et al. 2010; Munshi-South and Kharchenko 2010; Munshi-South et al. 2013; Munshi- South et al. 2016). Moreover, using genomic approaches, Munshi-South et al. (2016) found that urban populations of the white-footed mouse (Peromyscus leucopus) in New York City exhibit lower genome-wide variation compared with rural populations (see also Gortat et al. 2015), corroborating the deleterious effects of drift on genetic diversity in urban populations. However, despite the excellent work that has recently been performed on urban populations, we currently know little about the historical and population-specific factors that influence how the effects of drift are expressed in urban populations.
Two factors predominate in determining how strong the effects of drift are in isolated populations: (i) time since isolation and (ii) long-term (historical) and contemporary effective population size (Ne; Frankham 2005; Ellegren and Galtier 2016). The former determines how long drift acts without the homogenizing force of gene flow, such that populations isolated for longer periods bear more effects from drift (Frankham 2005). For the latter, phenomena such as founder effects or bottlenecks cause a pronounced decline in Ne from the ancestral population, reducing genome-wide allelic diversity. This reduction of allelic diversity is accompanied by increased genetic divergence as the effect of drift becomes stronger in small populations (Segelbacher et al. 2014; Spurgin et al. 2014; Ellegren and Galtier 2016). To our knowledge, while a handful of studies on urban populations have examined evidence of past bottlenecks (e.g. Noël and Lapointe 2010; Munshi-South and Nagy 2014), no urban genetics studies have incorporated quantitative estimates of contemporary Ne to characterize the genetic effects of urbanization. Additionally, the effects of time since isolation were only considered as a predictor of genetic diversity and differentiation in two studies (Delaney et al. 2010; Munshi-South and Nagy 2014), though both showed contrasting results, illustrating the difficulties in accurately quantifying its effects.
To investigate the combined roles of demographic and temporal effects on genetic diversity and population structure in urban systems, we studied urban populations of fire salamanders (Salamandra salamandra, Linnaeus 1758) in the historical city of Oviedo (Spain). Salamandra
range in Europe, but it has evolved to pueriparity (parturition of fully developed terrestrial juveniles) during the Pliocene–Pleistocene in northern Spain (including Oviedo; García-París
et al. 2003) and during the Holocene in two off-shore islands of NW Spain (Velo-Antón et al.
2007; Velo-Antón et al. 2012). This remarkable reproductive shift entails greater independence from surface water as the aquatic larval stage is removed (Velo-Antón et al. 2015), allowing S.
salamandra to cope with the harsh conditions of an urban environment. We sampled 12
populations within Oviedo and four outside the city to fulfil three main objectives: (i) to estimate and compare patterns of genetic diversity and genetic structure in urban and rural populations; (ii) to estimate historical and contemporary Ne of urban populations; and (iii) to quantify the contributions of population-specific traits, including demographic history, time since isolation and patch attributes to genetic differentiation and diversity.
Oviedo and S. salamandra comprise an excellent model for research on urban genetics for four key reasons. First, the nearly impervious urban matrix and low dispersal capability exhibited by fire salamanders (Schulte et al. 2007) generate an expectation that these populations have been isolated for many generations. Second, Oviedo contains populations with very low total census sizes (Álvarez et al. 2015) confined within small and discrete patches, making it straightforward to obtain representative demographic data and measure relevant variables known to be related to genetic variation (e.g. patch size; see Wang et al. 2014; Jackson and Fahrig 2016). Third, the availability of historical documents and maps of Oviedo depicting the period when particular buildings were constructed provides a clear means of estimating when urban populations became isolated within the city. Lastly, populations became isolated during distinct time frames (some more than 1000 years ago), and consequently, the short- and long-term effects of isolation can both be assessed. In our analysis, we test three main hypotheses: (i) urban populations exhibit higher genetic divergence and reduced diversity compared with rural populations; (ii) on patch–matrix landscapes such as urban environments, patch size is a strong predictor of contemporary Ne and, consequently, of the strength of contemporary drift (Ellegren and Galtier 2016); and (iii) urban populations which experienced an older decline and which were isolated for longer periods show greater genetic differentiation and reduced genetic diversity due to long-term effects of small Ne.