Cold tolerance, water balance, energetics, gas exchange, and diapause in overwintering brown marmorated stink bugs

Highlights

• Halyomorpha halys are chill-susceptible and die when overwintered outside.

• Diapausing H. halys have lower supercooling points, but unchanged cold tolerance.

• Diapausing H. halys have lower water loss and metabolic rates.

• H. halys likely rely on sheltered microhabitats to survive winter in Ontario.

Abstract

Halyomorpha halys (Hemiptera: Pentatomidae) is an emerging pest which established in Ontario, Canada, in 2012. Halyomporpha halys overwinters in anthropogenic structures as an adult. We investigated seasonal variation in the cold tolerance, water balance, and energetics of H. halys in southwestern Ontario. We also induced diapause in laboratory-reared animals with short daylength at permissive temperatures and compared cold tolerance, water balance, energetics, and metabolism and gas exchange between diapausing and non-diapausing individuals. Halyomorpha halys that overwintered outside in Ontario all died, but most of those that overwintered in sheltered habitats survived. We confirm that overwintering H. halys are chill-susceptible. Over winter, Ontario H. halys depressed their supercooling point to c. −15.4 °C, and 50% survived a 1 h exposure to −17.5 °C. They reduce water loss rates over winter, and do not appear to significantly consume lipid or carbohydrate reserves to a level that might cause starvation. Overall, it appears that H. halys is dependent on built structures and other buffered microhabitats to successfully overwinter in Ontario. Laboratory-reared diapausing H. halys have lower supercooling points than their non-diapausing counterparts, but LT₅₀ is not enhanced by diapause induction. Diapausing H. halys survive desiccating conditions for 3–4 times longer than those not in diapause, through decreases in both respiratory and cuticular water loss. Diapausing H. halys do not appear to accumulate any more lipid or carbohydrate than those not in diapause, but do have lower metabolic rates, and are slightly more likely to exhibit discontinuous gas exchange.

Introduction

The distribution and performance of ectothermic insects is determined to a large extent by the effects of environmental conditions, especially temperature and water availability (Harrison et al., 2012). In northern temperate environments, low temperatures, desiccation, and energy balance during winter are important determinants of insect survival (Danks, 2000, Leather et al., 1993, Sinclair, 2015). Thus, winter conditions limit the geographic distribution of many insects (e.g. Halbritter et al., 2018), and the success of some invasive insects can be attributed to their overwintering biology (e.g. the emerald ash borer; Cuddington et al., 2018). Identifying the limits to overwintering survival for invasive species is therefore essential for understanding their potential distribution and spread, and for informing management strategies.

At sub-zero temperatures, insects risk their body fluids freezing. A majority of insects are chill-susceptible (killed by cold unrelated to freezing; reviewed by Overgaard and Macmillan, 2017), but some are either freeze-avoidant or freeze-tolerant (Lee, 2010). Freeze-avoidant insects survive low temperatures provided they do not freeze; they increase cold tolerance by depressing the temperature of ice formation (the supercooling point, SCP), usually by accumulating low molecular weight cryoprotectants and sometimes ice-binding proteins, and by removing ice nucleators (Tattersall et al., 2012). By contrast, freeze-tolerant insects can withstand internal ice formation; they often also accumulate similar cryoprotectants to freeze avoiders, but control the initiation of ice formation (reviewed by Toxopeus and Sinclair, 2018). Cold tolerance usually varies seasonally, with insects acquiring increased cold tolerance (and even changing cold tolerance strategy) in the winter. By definition, the lower lethal temperature (LLT) of freeze-avoidant insects is described by the SCP, but needs to be calculated through iterative exposure to low temperatures for chill-susceptible and freeze-tolerant insects (Sinclair et al., 2015). Prolonged exposure to low, but not extreme, temperatures is the most likely scenario for cold exposure in nature (Sømme, 1996), and can be explored by determining a lethal time during exposure to low temperature (Lt).

Winter in temperate environments is dry. Below freezing, liquid water is unavailable, and overwintering insects seldom eat or drink (Danks, 2000, Ring and Danks, 1994, Sinclair et al., 2013). Some insects harness dehydration to improve cold tolerance (e.g. Holmstrup et al., 2002, Kawarasaki et al., 2014). Most insects conserve water by reducing water loss rates; however, some tolerate desiccating conditions by increasing initial water content and/or surviving the loss of more water by tolerating the loss of more water before dying (i.e. decreased water content at death) (Gibbs and Matzkin, 2001, Hadley, 1994, Sømme, 1995, Wharton, 1985). Water loss rates can be reduced by decreasing both metabolic rate (and therefore respiratory water loss, RWL) and cuticular water loss (CWL) (Chown, 2002). Although consuming glycogen can yield large amounts of hydrogen-bound water, and glycogen accumulation is the primary way to increase bulk water stores (Gibbs, 2002), most overwintering insects are thought to fuel their metabolism from energy-dense lipids (Hahn and Denlinger, 2011), which yield metabolic, but not bound, water (Gibbs and Matzkin, 2001, Hadley, 1994, Hahn and Denlinger, 2007, Sinclair, 2015, Sømme, 1995). Thus, increasing initial water storage should be accompanied by higher glycogen stores, and liberating bound water from that glycogen should be accompanied by a shift in fuel source from lipid to carbohydrate. The mechanisms by which insects increase the amount of water loss they can survive are not well-understood outside the extreme example of anhydrobiosis (Chown and Nicolson, 2004, Harrison et al., 2012).

Insects exchange gases with the environment via a tracheal system, and can modulate RWL by opening and closing the spiracles at the entrance to this system (Lighton et al., 1993). In some cases, this opening-and-closing can lead to cyclic gas exchange, including complete occlusion of the spiracles, sometimes for long periods of time, referred to as discontinuous gas exchange (DGE; Matthews, 2018). Although the role of DGE in regulating RWL is controversial (Chown, 2011, Chown et al., 2006, Contreras et al., 2014, White et al., 2007), understanding DGE may be important in insect pest control (Karise and Mänd, 2015). Some fumigants, such as phosphine, cannot penetrate the cuticle, so their absorption requires at least partial opening of the spiracles (Bond et al., 1969, Chaudhry and Price, 1992, Tonapi, 1971). Knowing whether gas exchange patterns are temperature- (e.g. Gudowska et al., 2017, Terblanche et al., 2010) or diapause- (e.g. Stalhandske et al., 2015) dependent is thus useful when determining phytosanitary regimes for fumigation of pest insects.

Because they generally do not feed, overwintering insects depend on energy stores to fuel metabolism and pre-feeding spring activities (Sinclair, 2015). Both carbohydrates and lipids can fuel overwintering, but the bulk of winter metabolism for most species is fuelled by lipids (Sinclair, 2015, Sinclair and Marshall, 2018). This energy consumption is not evenly spread throughout the winter. Because the temperature-metabolic rate relationship is exponential, energy consumption will be highest during extreme warm periods in autumn or spring (Sinclair, 2015). Furthermore, warm overwintering microhabitats (that protect against lethal cold exposure) likely carry an energetic cost. Insects may address these costs of overwintering by 1) accumulating additional resources prior to winter and 2) reducing energetic demand by decreasing the thermal sensitivity of metabolic rate and/or overall suppression of metabolic rate (reviewed by Sinclair, 2015).

Many temperate insects overwinter in diapause, a state of programmed developmental and metabolic arrest (Koštál, 2006). Diapause may be obligate (initiated regardless of environmental cues), or facultative (initiated only in response to specific cues; Hahn and Denlinger, 2011, Koštál, 2006, Tauber et al., 1986). Diapausing insects often have enhanced tolerance of environmental stressors and suppressed metabolic rate (Danks, 2002). In some insects, initiating the diapause program is sufficient to increase environmental stress tolerance (e.g. Denlinger and Armbruster, 2014), while other species increase stress tolerance in response to environmental conditions after they have already entered diapause (e.g. Boiteau and Coleman, 2012).

Halyomorpha halys (Stål, 1855) (Hemiptera: Pentatomidae), the brown marmorated stink bug, is native to subtropical and temperate Southeast Asia, and is an economically-important invasive species in North America, Europe, and South America (Gariepy et al., 2015, Gariepy et al., 2014a, Gariepy et al., 2014b, Kriticos et al., 2017). Halyomorpha halys is polyphagous, attacking a broad range of fruit and crop plants as a nymph and adult, causing substantial economic loss (Haye et al., 2015). Halyomorpha halys overwinter as adults in a reproductive diapause that is induced by short daylength (Musolin et al., 2019, Nielsen et al., 2017, Sibayan, 2018). In North America, adult H. halys aggregate in the autumn and overwinter in protected microhabitats, including natural habitats (e.g. standing dead trees) and residential structures, where they are a nuisance pest (Inkley, 2012, Lee et al., 2014).

Halyomorpha halys can clearly survive winter in places where they have been introduced. In laboratory studies, Lowenstein and Walton (2018) report 50–65% mortality in individuals overwintering in fluctuating conditions with minimum temperatures representing California, Oregon, and Washington States (USA), and Cira et al. (2016) report 100% survival after long periods (weeks) at + 4.3 °C. These conditions are warmer than experienced in more continental habitats in North America, where temperatures below freezing can persist for weeks or months. H. halys overwintered outdoors and in unheated sheds in Minnesota have 100% mortality (Cira et al., 2018, Cira et al., 2016), and there was near-complete overwinter mortality in central Washington where temperatures dropped below −15 °C (Sibayan, 2018). Thus, although diapausing H. halys can withstand long periods close to freezing over winter (especially if the temperatures fluctuate; Lowenstein and Walton, 2018), extended low temperature exposure appears to cause cold-related mortality. In these habitats, snow cover can stabilise the temperature and protect against extremes (Williams et al., 2015b), but in many regions, H. halys appears to overwinter primarily in protected habitats (Inkley, 2012).

Previous work on Halyomorpha halys cold tolerance has relied on measuring supercooling points. Supercooling points in winter are in the region of −17 °C in Minnesota, USA, and −15 °C in Virginia, USA, but higher in laboratory-reared Minnesota animals in which diapause had not been induced with shorter photoperiods (Cira et al., 2016). Field-collected animals in Washington State, USA, had SCPs of −12 to −14 °C, although these were measured at a fairly fast cooling rate (2 °C/min; Sibayan, 2018) which would normally result in a lower-than-expected SCP (Salt, 1966). Although SCPs were used to predict field mortality by Cira et al. (2016), SCPs are only a valid estimate of cold tolerance for freeze-avoidant insects (Bale, 1987, Baust and Rojas, 1985). Halyomorpha halys appear to be killed by temperatures well above the SCP (Cira et al., 2018, Cira et al., 2016, Lowenstein and Walton, 2018). However, the cold tolerance strategy has not been formally delimited (see Sinclair et al., 2015 for a description and rationale). Furthermore, Cira et al. (2018) estimated lethal temperatures using fast cooling rates and no exposure (‘soak’) at the actual test temperature, so their values likely do not represent the responses to cold in nature. Nevertheless, previous work (Cira et al., 2018, Cira et al., 2016, Lowenstein and Walton, 2018, Sibayan, 2018) remains strongly suggestive that low temperature mortality is a significant contributor to overwinter mortality in H. halys. Thus, although some argument can be made that changes in SCP reflect physiological changes even in chill-susceptible species (cf. Coleman et al., 2014), there are no measures of H. halys cold tolerance that reflect best practices, particularly if H. halys is chill-susceptible, and there is only circumstantial evidence that the reported overwinter mortality is caused specifically by low temperatures rather than the other stresses of overwintering.

In the protected, above-freezing, habitats they select to overwinter, H. halys do not eat or drink. Both dehydration and starvation can cause overwinter mortality in insects (Williams et al., 2015b). Provision of water had no effect on overwinter mortality at 10 °C (Chambers et al., 2019), suggesting that H. halys either 1) effectively conserve water over winter, 2) survive extensive dehydration, or 3) do not drink water to improve survival, even when it is provided. Thus, it is unclear if water loss is a cause of mortality in overwintering H. halys. By contrast, H. halys appear to consume resources over winter: post-winter survival is limited in individuals without access to food (Funayama, 2012), males have less overall energy stores than females, and glycogen content declines over winter (Skillman et al., 2018). Because glycogen declined, but lipid did not, Skillman et al. (2018) concluded that glycogen was the main source of overwinter energy, but their measures of lipid content depended on the vanillin assay, which is compromised by changes in fatty acid saturation, which is expected a priori in overwintering insects (Williams et al., 2011). Although overwintering H. halys appear to have reduced metabolic rates at 15 °C but not lower temperatures (suggesting reduced thermal sensitivity; Sibayan, 2018), it is unclear whether diapause induction changes energy consumption or water balance in H. halys, or if these processes are a response to changing environmental conditions during winter. Thus, overwinter energetics may be a key determinant of mortality in H. halys, but the roles of water balance and diapause are less clear.

Diapause physiology is specifically relevant to H. halys dispersal and invasion. Halyomorpha halys aggregate in built structures (including shipping containers) to overwinter (Hoebeke and Carter, 2003, Inkley, 2012). Thus, these aggregated, diapausing, individuals are the individuals most likely to disperse to new locations, and H. halys have been intercepted in shipping containers and luggage (Hoebeke and Carter, 2003, Meurisse et al., 2019, Ormsby, 2018). Thus, the physiological tolerances of diapausing individuals are directly relevant to management and control of dispersing H. halys. Furthermore, although diapause may have been induced in dispersing individuals, they may not have been exposed to other cues that could lead to acclimatisation in nature. Thus, understanding the physiology of individuals for which diapause has been induced by photoperiod alone is important for understanding the potential for accidental introduction of this species across climatic zones. In particular, during these dispersal events, we expect that dehydration and starvation tolerance likely determine survival more than low temperature exposure (Nixon et al., 2019), and we speculate that discontinuous gas exchange could reduce the efficacy of fumigants that cannot pass through the cuticle.

Here, we fill several gaps in our understanding of H. halys overwintering physiology in North America. We measure supercooling point at slower cooling rates to allow comparison with previous work, we estimate lethal limits using best practices, and (crucially) we formally determine the cold tolerance strategy to allow interpretation of cold tolerance data. Because the role of lipid consumption over winter remains unclear, we wanted to clarify the relative roles of carbohydrates and lipids in fuelling overwinter metabolism by using robust methods to measure the latter. Although providing water did not improve overwinter survival (Chambers et al., 2019), there are several explanatory hypotheses that still allow desiccation to be an important overwintering stress, so we explored water balance in overwintering H. halys to determine how they are able to tolerate this long dry period. We also explore the extent to which environmental physiology is changed solely by photoperiod-induced diapause (without any temperature or other environmental cues). This is relevant to the biology of hitchhiking dispersers, as well as verifying that experiments conducted on individuals in the laboratory are broadly representative of the biology of individuals in nature.

Our first objective was to determine seasonal patterns of mortality, cold tolerance, water balance, and energetics in H. halys overwintering in protected and natural habitats near its current Northern range edge in southwestern Ontario, Canada. Second, we determined whether photoperiod-induced diapause (without any additional environmental conditioning) affects cold tolerance, water balance, energetics, and metabolism and gas exchange of lab-reared H. halys; this is important for understanding how the diapause program drives environmental tolerances, and also to support future studies using lab-reared animals to develop control methods for diapausing H. halys overwintering in structures or during dispersal.