Genetic pattern of the recent recovery of European otters in southern France

Abstract

We investigated how landscape affects the population genetic structure and the dispersal of the elusive European otter Lutra lutra in a contemporary colonization context, over several generations and at the level of hydrographic basins. Our study area included 10 basins located in the Cévennes National Park (CNP), at the southern front of the natural otter recovery in France. Each basin comprised 50 to 300 km of permanent rivers that were surveyed for otter presence from 1991 to 2005. Faecal samples collected in 2004 and 2005 in this area were genotyped at 9 microsatellite loci, resulting in the identification of 70 genetically distinct individuals. Bayesian clustering methods were used to infer genetic structure of the populations and to compare recent gene flow to the observed colonization. At the regional level, we identified 2 distinct genetic clusters (NE and SW; FST00.102) partially separated by ridges, suggesting that the CNP was recolonized by 2 genetically distinct otter populations. At the basin level, the genetic distance between groups of individuals in different basins was positively correlated to the mean slope separating these basins. The probable origins and directions of individual movements (i.e. migration between clusters and basin colonization inside clusters) were inferred from assignment tests. This approach shows that steep and dry lands can stop, impede or divert the dispersal of a mobile carnivore such as the otter.

Genetic structure of Eurasian otter (Lutra lutra, Carnivora: Mustelidae) populations from the western Baltic sea region and its implications for the recolonization of north-western Germany

“…The study of Dallas et al (2002) shares eleven loci with ours, and their 618 British otters yielded 76 alleles when compared to 78 alleles that we found in 225 otters. Randi et al (2003), in an analysis covering large parts of Europe (n = 102), expectedly found more alleles than we did (80 versus 74 at 10 shared loci), while two more localized studies from France (Janssens et al 2008;n = 70) and Denmark (Pertoldi et al 2001, n = 125 including otters dating back to 1883) yielded fewer (France: 36 versus 69, nine shared loci; Denmark: 31 versus 71, 10 shared loci).…”

“…In the last 20 or so years, populations have recovered and former distribution areas have been recolonized, either naturally -as, e.g. in France (Janssens et al 2008), Sweden (Bjo¨rklund and Arrendal 2008), Italy (Marcelli and Fusillo 2009) or Denmark (Elmeros et al 2006) -or by means of reintroduction as in the Netherlands (Koelewijn et al 2010), Spain (Ferrando et al 2008) and England (Stanton et al 2009). Consequently, the otterÕs global conservation status has been downgraded from ÔVulnerableÕ to ÔNear ThreatenedÕ recently (Ruiz-Olmo et al 2008).…”

An evaluation of field and noninvasive genetic methods for estimating Eurasian otter population size

“…When using spraints from wild otters, however, they reported only 24% of samples as suitable for DNA analysis (Kalz et al 2006), which indicates rather low quality of the samples and is in contradiction to zero allelic dropout. A higher frequency of ADO was reported in studies on otters on Kinmen, an island southeast of China (30.7%; Hung et al 2004), and in southern France (54.0%; Janssens et al 2008). Calculations of genotyping errors are completely missing in the papers of Prigioni et al (2006a, b) and Lanszki et al (2008).…”

“…Papers of Prigioni et al (2006a) and Lanszki et al (2008) do not include any information on the quality of samples nor any measures to mitigate genotyping errors. All further studies on otters (Arrendal et al 2007;Ferrando et al 2008;Janssens et al 2008), however, report error rates thoroughly, the first also including the testing of some factors that could affect success and error rates. As found by Hájková et al (2006) and Janssens (2006), temperature is one of the main factors affecting amplification success of spraint DNA and, therefore, sampling at low temperatures is always to be recommended.…”

“…Reported success rates, though highly variable, are generally low, ranging from 14% to 73% (Jansman et al 2001;Dallas et al 2003;Hung et al 2004;Hájková et al 2005;Janssens et al 2008;Kalz et al 2006;Prigioni et al 2006a, b;Arrendal et al 2007;Ferrando et al 2008;Lanszki et al 2008). High levels of genotyping errors are typical for NGS, especially when polymerase chain reaction (PCR) success rates are not very high, a first indication of low quality and/or quantity of extracted DNA.…”