Effects of nutrient and water restriction on thermal tolerance: A test of mechanisms and hypotheses

Abstract

Nutritional deprivation or desiccation can influence thermal tolerance by impacting the insects' ability to evaporatively cool, maintain cell membrane integrity and conduct protective or repair processes. Recovery from chilling is also linked to the re-establishment of iono- and osmo-regulatory homeostasis. Here, using Mediterranean fruit fly (Ceratitis capitata, Diptera: Tephritidae), we manipulated water and nutrient availability to test the mechanistic expectation that changes in whole organism lipid and water content can elicit variation in cold or heat tolerance (scored as chill coma recovery time and heat knockdown time). We measured body condition (body water and lipid content) as well as heat shock protein 70 gene (hsp70) and protein (HSP70) levels. A significant reduction in body water content with water restriction did not translate into differences in chill coma recovery. When nutrient restriction was coupled with water deprivation, this resulted in a significant reduction (− 54%) of heat knockdown time in females but male flies were unaffected. There was no evidence for an hsp70 or HSP70 response under any of the stress treatments and therefore no correlation with heat or cold tolerance. Heat hardening decreased all hsp levels. Therefore, although body water and total body lipid content differed between the treatment groups, the contribution of these factors to thermal tolerance was inconsistent with mechanistic expectations in heat knockdown time and insignificant for chill coma recovery. These results therefore highlight that the effects of resource restriction on thermal limits in insects are mechanistically more complex than previous models of stress resistance have suggested.

Introduction

Traits of thermal tolerance or resistance are strongly correlated with insect species' geographic distributions (reviewed in Addo-Bediako et al., 2000), and assessing the factors influencing these traits in the field is important (e.g. Sinclair and Roberts, 2005, Cooper et al., 2008, Terblanche et al., 2011). Most studies use laboratory-based methodology to control for a host of intrinsic and extrinsic factors that can influence thermal tolerance estimates, such as recent thermal history (e.g. season, acclimation), age or sex (e.g. Fischer et al., 2010). The impact of body condition on tolerance estimates is typically not the subject of investigation (but see e.g. Terblanche et al., 2008, Overgaard et al., 2012). As the temporary deprivation of food and/or water can have downstream impacts on organismal performance and fitness, tolerating these stressors is essential for insects in heterogeneous landscapes. Mathematical models have also highlighted how thermal traits may be influenced by nutritional and water status at the organismal level (Kearney et al., 2013). Complex interactions and fitness trade-offs occur between stressors in at least some instances (reviewed in Hoffmann et al., 2003, Mellanby, 1932, Bubliy et al., 2012a, Bubliy et al., 2012b, Overgaard et al., 2012, Kellermann et al., 2013, Karl et al., 2014, Boardman et al., 2015, Scharf et al., 2016). The different approaches typically employed, and range of stressors considered may elicit very different underlying physiological processes, or similar trait responses may arise from distinctly different mechanisms. Thus, determining the mechanistic underpinnings of multi-stressor responses to changing body condition is a crucial avenue for forecasting species responses to changing environments.

There are at least three major expectations for how stress resistance may respond under changing nutritional or hydration conditions at the organismal level. First, the organism may already have the necessary cellular biochemistry in place to fully withstand the stress, and therefore no effect would be detected on thermal tolerances. Second, exposure to stress (e.g. food or water deprivation) may result in upregulation of protective (or repair) mechanisms that allows the organisms to better withstand subsequent thermally stressful conditions. Finally, biochemical or physiological responses may be upregulated but are insufficient to compensate for the subsequent thermal stress. Predictable trait responses can be generated to differentiate among these possibilities.

Desiccation prior to a heat or cold tolerance assay would generally be expected to reduce overall thermal tolerance by restricting available water resources that underlie multiple physiological traits (reviewed in Zera and Harshman, 2001). However, some studies have found a positive association between prior desiccation stress and heat tolerance (e.g. Benoit et al., 2009, Bubliy et al., 2012a), largely attributed to the upregulation of heat shock proteins (Hsps) that act as molecular chaperones for damaged proteins (Feder and Hofmann, 1999). However, the association is equivocal depending on stress type and exposure duration (e.g. Feder et al., 1992, Silbermann and Tatar, 2000, Boardman et al., 2013, Sørensen et al., 2013). Desiccation has been associated with improved cold hardiness by potentially increasing the osmolality of the tissue, which in turn, decreases the temperature at which the organism freezes (supercooling point; reviewed in Zachariassen, 1985). Cryoprotective dehydration, whereby an organism loses water during cold stress to avoid ice formation in tissues, is an important low temperature survival strategy found in many insects (e.g. Sørensen and Holmstrup, 2011; reviewed in Chown and Terblanche, 2006). However, if the organism does not freeze, then this increase in osmolality may detrimentally affect the individuals' ability to regain ion homeostasis, the latter of which is the major hypothesis describing insects' inability to recover from chill coma (e.g. Coello Alvarado et al., 2015, Olsson et al., 2016, reviewed in Overgaard and MacMillan, 2017).

Nutritional stress might influence thermal tolerance primarily through reductions in body lipid levels. Lipid levels and the composition thereof (e.g. long vs. short chains, number of double vs. single bonds) are thought to underpin the link between starvation and thermal tolerance in Drosophila (e.g. Hoffmann and Harshman, 1999, Hoffmann et al., 2005). For example, a negative association is found between starvation resistance and cold tolerance due to the importance of lipids in both traits (e.g. Hoffmann et al., 2005). Reductions in lipid levels are generally associated with starvation across all organisms (McCue, 2010) but may result in increased water uptake (e.g. Raubenheimer and Gäde, 1993) which can be beneficial for improving heat tolerance by increasing the water stores available for evaporative cooling or generally withstanding desiccation stress (Chown and Nicholson, 2004).

Here we provide a systematic, comprehensive assessment of these potential mechanisms that have been hypothesised to underlie the influence of water and nutritional status on thermal tolerance using the Mediterranean fruit fly, Ceratitis capitata (Wiedemann) (Diptera: Tephritidae). As a global agricultural pest there is considerable interest in better understanding its basal and plastic climatic stress resistance, and the underlying mechanisms thereof (De Meyer et al., 2008, Malacrida et al., 2007, Nyamukondiwa et al., 2010). Previous studies have measured chill coma recovery time and heat knockdown time (Weldon et al., 2011), shown a pronounced decrease in heat tolerance (measured as CTmax) following prior starvation (Nyamukondiwa and Terblanche, 2009) and plastic responses to prior thermal exposures (Nyamukondiwa et al., 2010) that were attributed to variation in hsp70 mRNA expression (Kalosaka et al., 2009, but see Pujol-Lereis et al., 2014). We exposed C. capitata to a full factorial experimental design of long term (i.e. several days as adults) restriction of nutrients, water and the combined restriction of both nutrient and water and then assayed chill coma recovery time and heat knockdown time. We also estimated body water and body lipid amounts, and the relative amount of mRNA hsp70, HSP70 protein and total protein concentration to explain any potential thermal tolerance responses. We predicted that restricting water or nutrient levels would have a detrimental effect on heat tolerance and, by contrast, be beneficial for cold tolerance. Furthermore, it can be predicted that if thermal tolerance is set in some mechanistic way by an organism's body condition, then combined stressors of nutrient and water restriction should have a greater reduction on thermal limits than either stressor applied individually, and these latter should reduce thermal limits more than a control or reference group. Heat tolerance is likely to be impacted by reduced water stores by limiting evaporative cooling and/or reducing the pool of water available to survive prolonged stress. Nutrient restriction may also be detrimental for heat tolerance by reducing the amount of stored lipids that can be metabolised for water. The change in saturation of cell membranes caused by reduced nutrient levels would be beneficial for cold tolerance. Reduced water levels would benefit cold tolerance by reducing the formation of ice within the tissue. However, if the nutrient restriction results in an increased uptake in water, heat tolerance may benefit from increased structural integrity of cell membranes i.e. through lipid saturation, and more available body water for evaporative cooling. Cross-tolerance may also be detected if prior depletion of nutrients or water results in heat shock protein production.