Are southern pine forests becoming too warm for the southern pine beetle?

Highlights

• Decreased abundance of D. frontalis is not a result of high temperature mortality.

• There was no change in the maximum high temperature or the occurrence of heat waves.

• Other factors that may influence the abundance of D. frontalis must be considered.

• It is unclear if similar high temperature trends are occurring elsewhere.

• The methods used here can be applied to species in other regions as well.

Abstract

Climate change is altering the geographic distribution of many species, including numerous forest insect pests, with the potential for severe impacts to forest ecosystems. The southern pine beetle (Dendroctonus frontalis), one of the world's most persistently destructive forest pests, has extended its distribution towards the north, while impacts within its historic range in the southeastern United States have diminished, indicating that a range shift may be occurring. This pattern could be explained by the hypothesis of climatic envelopes if summers in the southern part of the range are warming about as much as winters in the northern part of their range. Here, we tested whether the southeastern United States is becoming too hot for D. frontalis even as the northeastern United States has become climatically permissive. We measured effects of heat on insect survival, and characterized the thermal environment of D. frontalis, including estimating the frequency of potentially lethal heat waves and testing whether the intensity or duration of heat waves has increased over the past 80 years. We found that temperatures warm enough to kill D. frontalis larvae, pupae, or callow adults have been rare or absent and that there has been no change in the duration or severity of heat waves in the southern pine forest region over the last 80 years. Therefore, alternative explanations for the reduced activity of D. frontalis within its historic range must be considered. The broad conclusions are not restricted to D. frontalis because they flow from an asymmetry in climate change: at least for now in eastern North America, the coldest winter temperatures have been becoming less extreme without summer heat becoming more extreme. This creates broad potential for northward range expansions without expectation of corresponding contractions in southern distribution limits.

Introduction

The southern pine beetle (Dendroctonus frontalis Zimmermann; Coleoptera: Curculionidae) is one of the most aggressive tree-killing insects in the world (Vega and Hofstetter, 2015). Within its historic distribution in the southeastern United States, losses to the forest products industry exceeded $1 billion for 28 years ending in 2004 (Pye et al., 2011) and this, along with the loss of property value and recreation space, are only some of the effects of D. frontalis on forests and people (Coulson and Klepzig, 2011; Coulson and Meeker, 2011).

In most years, in most forests, D. frontalis is a rare and relatively benign component of the forest insect fauna that primarily harbors in weakened host trees (Thatcher and Pickard, 1964). However, D. frontalis populations also display regional outbreaks that can last from one to several years, kill large expanses of pine trees and transform forests (Clarke et al., 2016). One or two epidemics per decade have been common for the region (Birt, 2011). The population dynamics of D. frontalis are apparently governed by a nonlinear density-dependent feedback system with two locally stable equilibria (endemic and epidemic) separated by a region of positive feedback that produces an unstable equilibrium (escape threshold) (Martinson et al., 2013). When population abundance is near the escape threshold, even modest environmental effects on beetle survival and reproduction can produce a state change (e.g., endemic to epidemic). Factors that are candidates to produce such effects with D. frontalis include forest structure and silvicultural practices (Nowak et al., 2015; Asaro et al., 2017; Aoki et al., 2018), predation (Reeve, 1997; Weed et al., 2017), environmental effects on tree physiology (Lorio, 1986; Reeve et al., 1995; Lombardero et al., 2000a), active suppression by forest managers (Clarke and Billings, 2003; Billings, 2011), interactions involving mites and antagonistic fungi (Hofstetter et al., 2006a), the coldest night of the winter (Ungerer et al., 1999; Tran et al., 2007), and other features of weather (Reeve, 2018).

Over the past 25 years, impacts from this insect have been declining throughout the southern pine region, despite a simultaneous increase in pine habitat (Clarke et al., 2016; Asaro et al., 2017). A number of hypotheses have been put forth to explain this pattern, including changes to forest structure and host resistance from improved management and tree breeding programs, as well as climatological impacts from anthropogenic climate change (see Asaro et al., 2017). While a climatological explanation is plausible, there has not been a discernible link between climate and D. frontalis outbreaks in the South. However, most studies of climatological factors have focused on the positive influence of extreme climate events promoting outbreaks via their effect on host trees (Friedenberg et al., 2008; Asaro et al., 2017). For example, drought and/or high temperatures could weaken hosts and lead to outbreaks (McNichol et al., 2019). Alternatively, regional changes in climate could directly impact survival of D. frontalis, and in that manner alter the occurrence of local outbreaks (Ayres and Lombardero, 2000). Such influence of climate on population has already been seen at the northern extent of their distribution, where recent increases in minimum winter air temperatures have permitted an expansion of D. frontalis outbreaks into New Jersey, Long Island, and Connecticut (Lesk et al., 2017; Dodds et al., 2018), even as impacts within their historic southern range were declining. One possible explanation for the increase in outbreaks at the northern edge of their range and declining impacts in the South is that southern forests have become too warm for D. frontalis, even as northern forests have only recently become warm enough. This would be as expected under the hypothesis of climatic envelopes (Hijmans and Graham, 2006; Green et al., 2008), which predicts approximately symmetrical changes at the warmer and cooler distribution limits given approximately uniform climate warming. The envelope model, in various forms, underlies many forms of species distribution models (Pecchi et al., 2019), but has been difficult to test (Hao et al., 2019). One alternative hypothesis is that warmer is better for many ectotherms at both the warm and cool edges of their distributions (Frazier et al., 2006).

General circulation models predict a continuing rise in mean surface temperatures across most of the northern hemisphere (IPCC, 2013). This can result in less severe minimum winter temperatures (Easterling et al., 1997; Vose et al., 2005; Tran et al., 2007; Daly et al., 2012; Weed et al., 2013) as well as more intense and longer lasting heat waves (Easterling et al., 2000; Meehl and Tebaldi, 2004; Lyon et al., 2019). Some changes in thermal extremes can affect the distribution and abundance of biological populations (Pörtner, 2002; Haynes et al., 2014; Buckley and Huey, 2016; Lloret and Kitzberger, 2018).

Insects die differently from heat than from cold. Many insects, including D. frontalis, are freeze-intolerant (Duman et al., 1991) and have a discrete lower lethal temperature that corresponds to their supercooling point (the temperature at which fluids crystallize and cell membranes rupture). In D. frontalis, brief exposure to temperatures below the supercooling point ensure death, whereas sustained exposure to temperatures slightly above the supercooling point have little effect on survival (Lombardero et al., 2000b). Thus, the effects of cold on overwinter survival of wild populations can be realistically modeled based on the coldest night of the winter (Tran et al., 2007); duration of exposure matters relatively little (Lombardero et al., 2000b). However, when insects die from heat, mechanisms include destabilization of membranes and proteins, increase in reactive oxygen species, as well as disruption of oxygen delivery, water budgets, and energy balance (McCue and De Los Santos, 2013). Depending on the mechanism, the duration of heat exposure required to produce mortality might vary from hours to days, and exact cause of death is likely to depend upon the intensity and duration of heat exposure (Rezende et al., 2014).

We tested the hypothesis that the historic range in the southeastern United States is becoming too hot for D. frontalis by (1) experimentally measuring effects of heat exposure on insect survival, (2) characterizing the thermal environment of D. frontalis within the inner bark of host trees, (3) estimating the historical frequency of potentially lethal heat waves in the historic range of D. frontalis, and (4) testing whether the intensity or duration of heat waves has been increasing during recent decades.