Discussion
Phenotypic plasticity involves phenotypic changes associated with environmental conditions, and may favor the establishment or persistence of organisms in changing environments (Ghalambor et al. , 2007). Consequently, plasticity may potentially affect the selective pressures that a population encounters and, as a result, its evolutionary trajectory (Oster & Alberch, 1982; Bonduriansky et al. , 2012). Our experiment illustrates how within- and transgenerational plasticity can ameliorate the impact of stressful thermal conditions on physiological and fitness-related traits. In this context, our main results can be summarized as follows. First-generation flies subjected to a variable and, stressful thermal environment exhibited higherCTmax when compared against their counterparts maintained at a constant temperature. This plastic and seemingly adaptive response came at costs, however, since these flies also exhibited lower survival rates, fecundity (Figure 3), and, ultimately,\(R_{0}\) and Tg (Figure 2). Interestingly, these maladaptive plastic responses were less evident in their offspring (Figures 2 and 3), suggesting partial compensation mediated to some degree by transgenerational plasticity.
We aimed to compare the response of the flies under variable and constant thermal environments using both direct measures of fitness and physiological proxies of fitness, such as thermal tolerance, and results were dramatically different (Figure 3). Contrary to results by Nyamukondiwa et al. (Nyamukondiwa et al. , 2018) who evaluated the influence of thermal variability on heat tolerance, our results show that CTmax was positively affected by thermal variability, although with adverse effects on fitness.
These results are intriguing because the temperature peak in the variable thermal environment (32ºC) was substantially lower than the estimated CTmax (~ 39ºC), and yet this temperature was clearly stressful and impacted survival (Figure 3). Although the impact of temperature peak and time of exposure to thermal extremes on organisms cannot be disentangled, the differences in exposure time might explain this counterintuitive result, since the temperature range that organisms can tolerate is associated with the duration of thermal stress (Rezende et al. , 2014). Consequently, prolonged exposure to high and yet less extreme temperatures elicited an increase in CTmax at a cost in survival and, more importantly, in fecundity rates that are suggestive of a trade-off since less energy could be allocated to reproduction. These results agree with Folguera et al. , (2009) who reported that high environmental thermal amplitude experienced by terrestrial isopods increased the synthesis of stress-inducible heat-shock proteins (HSP), but at a metabolic energy cost with negative effects on longevity and growth rate. Not only is the production of HSP metabolically expensive (e.g., protein biosynthesis represent nearly 30-50% of total cellular energy consumption(Krebs & Loeschcke, 1994; Krebs & Feder, 1997, 1998)), but also they require ATP to maintain the structural integrity of other proteins (Hochachka & Somero, 2002). In this sense, although plastic responses may mitigate the adverse effects of thermal stress, their compensatory effects might be limited by energetic trade-offs (Pigliucci, 2001; Bozinovic et al. , 2016). Consequently, several studies work with the assumption that higher heat tolerance is a beneficial trait (Cavieres et al. , 2016; Sørensen et al. , 2016; Salachan & Sørensen, 2017; Salinas et al. , 2019), but here we show that this response was accompanied by a decrease in survival and fecundity, highlighting the importance of incorporating direct measures of fitness in physiological studies in order to have a broad understanding of the implications of phenotypic changes in response to environmental inputs.
Interestingly, fitness cost of living under a variable temperature decreased significantly in the second generation, providing evidence of partial compensation to a stressful thermal environment. This cross-generational compensatory response involves a 133% increase inR0 in the F1 in comparison to P (R0 = 259 in VV versus 111 eggs/female in V), but values were still 25 % lower than in flies reared at a constant temperature (R0 = 325 eggs/female). In this context, our results agree with previous studies that have documented that parental experience modifies the response to environmental input in their offspring. For instance, rapid compensatory responses in tolerance and/or reproductive output have been described in the marine polychaeteOphryotrocha labronica subjected to a low CO2environment (Rodríguez-Romero et al. , 2016) in Daphnia magna (Gustafsson et al. , 2005) raised with toxic cyanobacteria or coral reef fish Acanthochromis polycanthus (Donelson et al. , 2016) exposed to high temperature (see also(Jensen et al. , 2014; Thor & Dupont, 2015). Overall, these studies suggest that populations can respond rapidly to pronounced environmental changes, providing putative evidence that non-genetic inheritance might underlie observed responses to rapid changes in climatic conditions (Rando & Verstrepen, 2007).
The potential impact of selection should not be dismissed, however. Recovery of reproductive output reported in the literature and in our study might result from the synergistic effects of within-generation plasticity and genetic adaptation(Rodríguez-Romero et al. , 2016). As has been recently pointed out, estimates of transgenerational plasticity can be biased due to selection and this is the case even in half- or full-sibs designs(Santos et al. , 2019). Consequently, these effects are particularly relevant in studies dealing with responses to stressful environments employing outbred populations (e.g., this study or results for O. labronica (Rodríguez-Romero et al. , 2016)), which is a problem often neglected in the literature of transgenerational effects (Ho & Burggren, 2010; Burggren, 2014). For instance, the drop in survival and fecundity in P flies from V may impose strong selection and observed responses in F1 could partly reflect adaptive responses in the Darwinian sense. While the growing evidence indicates that natural populations can respond rapidly to environmental changes, resulting in full or partial compensatory responses to an environmental stress, caution is warranted regarding inferences on the mechanistic basis underlying these responses (Santoset al. , 2019).
To summarize, here we describe how an outbred population of D. melanogaster responds rapidly changing thermal conditions within and across generations. Our analyses provide evidence of a trade-off between thermal tolerance and fitness components such as fecundity in parental flies and pronounced albeit incomplete compensatory responses in their offspring. Similar approaches are necessary to extend studies from within- generational responses to responses across multiple generations. In this context, we urge future research to be tailored to specific climatic scenarios or geographic regions, aiming to build explanations and predict, in the near future, the potential responses of natural populations to ongoing global warming.