Sensory trait variation in an echolocating bat suggests roles for both selection and plasticity

Abstract

BackgroundAcross heterogeneous environments selection and gene flow interact to influence the rate and extent of adaptive trait evolution. This complex relationship is further influenced by the rarely considered role of phenotypic plasticity in the evolution of adaptive population variation. Plasticity can be adaptive if it promotes colonization and survival in novel environments and in doing so may increase the potential for future population differentiation via selection. Gene flow between selectively divergent environments may favour the evolution of phenotypic plasticity or conversely, plasticity itself may promote gene flow, leading to a pattern of trait differentiation in the presence of gene flow. Variation in sensory traits is particularly informative in testing the role of environment in trait and population differentiation. Here we test the hypothesis of ‘adaptive differentiation with minimal gene flow’ in resting echolocation frequencies (RF) of Cape horseshoe bats (Rhinolophus capensis) across a gradient of increasingly cluttered habitats.ResultsOur analysis reveals a geographically structured pattern of increasing RF from open to highly cluttered habitats in R. capensis; however genetic drift appears to be a minor player in the processes influencing this pattern. Although Bayesian analysis of population structure uncovered a number of spatially defined mitochondrial groups and coalescent methods revealed regional-scale gene flow, phylogenetic analysis of mitochondrial sequences did not correlate with RF differentiation. Instead, habitat discontinuities between biomes, and not genetic and geographic distances, best explained echolocation variation in this species. We argue that both selection for increased detection distance in relatively less cluttered habitats and adaptive phenotypic plasticity may have influenced the evolution of matched echolocation frequencies and habitats across different populations.ConclusionsOur study reveals significant sensory trait differentiation in the presence of historical gene flow and suggests roles for both selection and plasticity in the evolution of echolocation variation in R. capensis. These results highlight the importance of population level analyses to i) illuminate the subtle interplay between selection, plasticity and gene flow in the evolution of adaptive traits and ii) demonstrate that evolutionary processes may act simultaneously and that their relative influence may vary across different environments.

Environmental correlates of geographic divergence in a phenotypic trait: A case study using bat echolocation

“…This was also the case for R. ferrumequinum (Jiang et al., 2015) in Asia and for R. capensis in southern Africa. Similar to R. damarensis , R. capensis has a distribution that extends into the more arid western and northwestern regions of South Africa (Neumann & Bamford, 2015; Odendaal et al., 2014) and its RF was also not correlated with RH (Odendaal et al., 2014). This may be because both R. damarensis (Figure 5b) and R. capensis (Odendaal et al., 2014) occupy more arid regions (Table 1; compare with Mutumi et al., 2016) and in such areas temperature is a better predictor of RF than RH.…”

“…Several studies have revealed lineage diversification within species in the western half of southern Africa as a result of climate‐induced changes in biomes (Bauer & Lamb, 2005; Matthee & Flemming, 2002). Comparative studies in the western half of southern Africa revealed lineage diversification and clades within species in a wide range of animals such as reptiles (da Silva & Tolley, 2013), insects (Pitzalis & Bologna, 2010), and mammals (Matthee & Robinson, 1996; Willows‐Munro & Matthee, 2011; du Toit, Jansen Van Vuuren, Matthee, & Matthee, 2012) including horseshoe bats (Jacobs et al., 2013; Odendaal et al., 2014; Taylor et al., 2012). It appears therefore that local adaptation to climate in combination with an interruption of gene flow resulting from IbE may explain phenotypic divergence in R. damarensis and the regions fauna.…”

“…Akaike information criterion corrected differences (Δ i ) and Akaike weights ( w i ) were used to identify the best candidate models (Amar, Koeslag, Malan, Brown, & Wreford, 2014; Odendaal et al., 2014). Models with the lowest AICc and highest Akaike values were considered the most parsimonious, and the differences in AIC scores (Δ i ) were calculated to determine the likelihood that a given model was the best approximating model relative to other candidate models (Amar et al., 2014; Odendaal et al., 2014; Symonds & Moussalli, 2011). A model with a (Δ i ) value of zero was considered the best approximating model, and those with values of up to two were regarded as good as the best model (Symonds & Moussalli, 2011).…”

“…Sensory signals (e.g., acoustic signals) are important in the context of lineage diversification (Mutumi, Jacobs, & Winker, 2016; Odendaal, Jacobs, & Bishop, 2014; Slabberkoorn & Smith, 2002) because they are essential to the survival and reproduction of animals that rely on such signals for orientation and foraging as well as in mate choice and assortative mating (Bolnick & Kirkpatrick, 2012; Coyne & Orr, 2004). These signals are propagated through the environment, and specific environmental conditions can influence the evolutionary trajectory of the signaling system.…”

Listening carefully: increased perceptual acuity for species discrimination in multispecies signalling assemblages

“…This species emits resting frequency echolocation calls (RF, calls emitted by rhinolophid bats when stationary and hunting from a perch; Neuweiler et al, 1987;Schnitzler, 1968) that vary by more than 10 kHz across its distribution range ( Fig. 1; Odendaal, Jacobs, & Bishop, 2014). The lowest resting peak frequency, 75 kHz, is found in the northwestern part of its distribution, and the highest, 86 kHz, in the southeast.…”