Evidence for different thermal ecotypes in range centre and trailing edge kelp populations

Highlights

• Warmer trailing edge kelp populations are more thermotolerant than cooler range centre populations.

• Differences were fixed across season and population's exhibit restricted gene flow, indicating local adaptation.

• Local adaptation may make species responses to future climate changes more complex than previously thought.

Abstract

Determining and predicting species' responses to climate change is a fundamental goal of contemporary ecology. When interpreting responses to warming species are often treated as a single physiological unit with a single species-wide thermal niche. This assumes that trailing edge populations are most vulnerable to warming, as it is here where a species' thermal niche will be exceeded first. Local adaptation can, however, result in narrower thermal tolerance limits for local populations, so that similar relative increases in temperature can exceed local niches throughout a species range. We used a combination of common garden temperature heat-shock experiments (8–32 °C) and population genetics (microsatellites) to identify thermal ecotypes of northeast Atlantic range centre and trailing edge populations of the habitat-forming kelp, Laminaria digitata. Using upregulation of hsp70 as an indicator of thermal stress, we found that trailing edge populations were better equipped to tolerate acute temperature shocks. This pattern was consistent across seasons, indicating that between-population variability is fixed. High genetic structuring was also observed, with range centre and trailing edge populations representing highly distinct clusters with little gene flow between regions. Taken together, this suggests the presence of distinct thermal ecotypes for L. digitata, which may mean responses to future warming are more complex than linear range contractions.

Introduction

Temperature is one of the most important drivers of ecological patterns and processes (Hutchins, 1947), dictating where a species can exist and how well it performs throughout its distribution (Brown, 1984; Dunson and Travis, 1991; Gaston, 2003). Rising global temperatures, from anthropogenic greenhouse gas emissions, have already resulted in the altered performance and poleward range migrations of a range of biota and is set to continue as climate change advances (Walther et al., 2002; Parmesan and Yohe, 2003; Burrows et al., 2011; Sunday et al., 2012; Poloczanska et al., 2013). As range migrations and altered performances can have serious implications for the structure and functioning of entire ecosystems (Walther et al., 2002, Parmesan, 2006, Doney et al., 2011), understanding the effect of rising temperatures on species performance and distributions is a key goal in climate change ecology. However, if this is to be achieved then we must first understand the physiological traits and mechanisms that govern existing species distributions.

When forecasting future distributions, species are often treated as a single homogenous unit (Pearman et al., 2010; Reed et al., 2011), with populations assumed to exhibit similar thermal limits throughout the species' range. Therefore, thermal safety margins, the buffer between experienced temperatures and a species upper thermal limits (see Deutsch et al., 2008; Bennett et al., 2015), are assumed greatest at the range centre and lowest at trailing edges. As such, range centre populations are generally considered to be less vulnerable to predicted warming trends than trailing edge populations, where thermal safety margins will be exceeded first (Thomas et al., 2006; Thomas, 2010). Along temperature gradients, however, thermal tolerances are not always consistent between populations, as local adaptation and/or phenotypic plasticity can result in thermal limits being different across a species biogeographic range (Sanford and Kelly, 2011). This may result in more complex responses to warming than simple linear range contractions from trailing edges (Sanford and Kelly, 2011; Valladares et al., 2014; Bennett et al., 2015; Pontes-da-Silva et al., 2018), meaning central populations may also be vulnerable to ongoing warming.

Within the context of decadal scale warming, the significance of intraspecific variation in thermal niche largely depends on the mechanisms responsible (i.e. plasticity vs. adaptation). If the response is plastic then species are likely to be able to keep pace with climate warming, but if responses are a result of adaptation then the pace of warming is likely to be too fast for natural selection (Jump and Penuelas, 2005; Quintero and Wiens, 2013). Gaining an understanding of gene flow can provide valuable insight into whether adaptation or plasticity is favoured. Where gene flow is greater than the selection gradient, there is likely to be little selection for local thermal adaptation, resulting in a single plastic phenotype (García-Ramos and Kirkpatrick, 1997; Kirkpatrick and Barton, 1997), whereas restricted gene flow facilitates the development of local ecotypes (Endler, 1977).

Sessile organisms, that cannot modify their behaviour, rely on physiological mechanisms, underpinned by modulation of gene expression, to mediate periods of thermal stress. Whilst the mechanisms themselves are evolutionary conserved, patterns in gene expression are heritable and can vary considerably between populations (López-Maury et al., 2008) resulting in population level differences in thermal tolerance (e.g. Henkel and Hofmann, 2008). Therefore, analysis of variation in gene expression offers a powerful tool to identify thermal set points of a population that are often apparent before higher level physiological differences are observed. The Heat Shock Response (HSR) is perhaps the most well studied mechanism for identifying differences in thermal physiology. When organisms are challenged by elevated temperatures that result in protein denaturation and aggregation they rapidly upregulate a suite of molecular chaperones known as heat shock proteins (HSPs). These HSPs preserve normal cell function by ensuring appropriate protein folding during translation (Frydman, 2001), membrane stability and transport (Hartl and Hayer-Hartl, 2002) and protein refolding (Hendrick and Hartl, 1993).

Kelps are large brown habitat-forming seaweeds that form extensive forests along rocky coastlines in temperate and subpolar regions (Steneck et al., 2002; Smale et al., 2013; Teagle et al., 2017). Kelp forests rank among the world's most productive and extensive habitats, being distributed along one-quarter of the world's coastlines and rivalling the productivity of tropical rainforests (Leith and Whittaker, 1975; Mann, 1973). Kelp exist over vast temperature gradients and whilst long distance dispersal is possible (e.g. Fraser et al., 2018) the majority of spores settle within a few meters of the parent alga (e.g. Norton, 1992; Kendrick and Walker, 1995). Therefore, local adaptation may be a common feature throughout kelp distributions. Given that climate mediated range shifts have already been observed in kelp forests across the world (Marba and Duarte, 2010; Wernberg et al., 2016; Krumhansl et al., 2016) and are predicted to continue as warming progresses (e.g. Martínez et al., 2012; Jueterbock et al., 2013; Khan et al., 2018; Assis et al., 2018) understanding whether intraspecific variation in thermal niche may make central populations vulnerable to future warming trends has direct relevance for management and conservation of kelp forest ecosystems.

In this study we investigated intraspecific variation in the Heat Shock Response (HSR) of Laminaria digitata (Hudson) J.V. Lamouroux, a common transatlantic kelp. Specifically, we compared populations from two thermally distinct regions in the United Kingdom (representing East Atlantic range centre and trailing edge populations) and conducted experiments at the coolest and warmest times of year, to characterise intraspecific and intra-annual variability in HSRs. We also used neutral microsatellite markers to gain an understanding of gene flow between populations. By adopting a multi-pronged experimental approach we aimed to determine (i) whether populations show differentiation in thermal niche and (ii) whether such differences are likely a product of plasticity or adaptation.