The most common use of insect pheromones in Integrated Pest Management (IPM) is to bait traps to detect or delimit populations of the target insect pest (Witzgall et al., 2010). In some instances, capture of adult insects in pheromone-baited traps can predict reproductive capacity of the population (Contarini et al., 2009), subsequent densities of immatures (Evenden et al., 1995) and crop damage (Mori et al., 2014). Despite the extensive interest in and application of pheromone monitoring in IPM, relatively few studies examine how the effectiveness of pheromone-baited traps may vary with population density of the target pest and therefore, the population estimates obtained (McNeil, 1991). As population density increases, the number of individuals releasing pheromone in the managed ecosystem will also increase. If natural pheromone plumes are as or more attractive than those from synthetic lures, high insect densities could cause synthetic pheromone-baited traps to be less efficient at high than low population densities (Unnithan and Saxena, 1991; Thorpe et al., 1993; Delisle et al., 1998), and result in population estimates that are biased downwards at high density (Asaro and Berisford, 2001; Beroza et al., 1974; Jones et al., 2009). This problem would be predicted to be most acute for species that experience cyclical changes in population density, like many lepidopteran defoliators of forests.

Pheromone-baited traps have been developed as a tool to detect invasions and predict population densities for several forest defoliators. Male spruce budworm moths, Choristoneura fumiferana (Clem.) (Lepidoptera: Tortricidae), captured in pheromone-baited traps positioned in Ontario, Canada annually over a 21-year period were strongly correlated with late-instar larval population densities in each subsequent year (Sanders, 1988). As a result, sex pheromone-based monitoring is used operationally for spruce budworm in eastern North America. The number of male gypsy moths, Lymantria dispar L. (Lepidoptera: Erebidae), captured in traps baited with either 1 or 500 μg of the gypsy moth sex pheromone, (+)-disparlure, was correlated with subsequent egg mass density at the same sites (Thorpe et al., 1993). Male moth capture in traps baited with low but not high doses of the sex pheromone of the western hemlock looper, Lambdina fiscellaria lugubrosa (Hulst) (Lepidoptera: Geometridae), was correlated with samples of larval and pupal population density within the same generation and significantly predicted the number of eggs sampled at the same sites in the subsequent generation. This relationship was best, however, at the beginning of the flight season before peak flight (Evenden et al., 1995). In several studies, however, pheromone-baited trap capture of forest defoliators does not accurately reflect population densities and/or pheromone traps are less efficient at high population densities (Delisle et al., 1998). Western spruce budworm, Choristoneura occidentalis (Walsingham) (Lepidoptera: Tortricidae), season-long capture was not correlated to larval numbers sampled in the mid-crown of 50 trees at the same sites (Sweeney et al., 1990). Trap capture of Douglas-fir tussock moths, Orgyia pseudotsugata (McDunnough) (Lepidoptera: Erebidae), in traps baited with various pheromone doses did not predict numbers of egg masses or defoliation in the following generation (Shepherd et al., 1985).

The forest tent caterpillar is a native forest defoliator with a broad geographic range in North America (Stehr and Cook, 1968). Although broadly polyphagous, the primary host of the forest tent caterpillar in its northern range is trembling aspen, Populus tremuloides Michenaux (Salicaceae). Forest tent caterpillars experience cyclical changes in population density (Daniels and Myers, 1995). At high densities, repeated defoliation of aspen can cause reduced tree growth (Hildahl and Reeks, 1960; Churchill et al., 1964) and tree mortality (Candau et al., 2002). The commercial value of trembling aspen has increased in recent years (Hogg et al., 2002; Brandt et al., 2003) and tools are needed to monitor forest tent caterpillar populations and accurately predict incipient outbreaks. Males search for mates in response to a female-produced sex pheromone (Struble, 1970) and orient to synthetic pheromone positioned in non-saturating, high volume traps (Schmidt et al., 2003). Trap capture of male forest tent caterpillar moths in synthetic pheromone-baited traps increases with moth density to an intermediate level but declines at the highest population densities tested (Jones et al., 2009).

There are several potential mechanisms that could be driving the reduced trap capture of male forest tent caterpillars in synthetic pheromone-baited traps at high population densities. Competition for pheromone sources could occur at high population densities due to the high number of female moths releasing pheromone. It is known that male forest tent caterpillars compete for mates at high population densities, as late-arriving males will attempt to separate pairs already in copula (Bieman, 1982; Bieman and Witter, 1983). This competition may be enhanced in established outbreak populations that exhibit a male-biased sex ratio in the adult forest tent caterpillar population (Witter, 1979). A high density of calling females in stands with high population densities could create a natural mating disruption effect. Although mating success of female spruce budworm moths increases with population density, there is evidence for competition for pheromone sources at high population densities that negatively affects the capacity of females to attract a mate (Régnière et al., 2013). Mating disruption using synthetic pheromone sources results in reduced trap capture of male forest tent caterpillars in synthetic pheromone-baited traps positioned in treated aspen stands (Palaniswamy et al., 1983) but the potential for natural mating disruption at high population densities has not been tested in this species.

Another potential impact of high population density on mate finding behavior is the effect of mating on subsequent pheromone response. At high population densities, due to close proximity of females in the area, male moths may quickly secure a mate. Following copulation, there may be physiological changes in male moths that alter their ability to subsequently respond to pheromone. Mating is known to reduce pheromone production in female moths (Raina et al., 1994; Ramaswamy et al., 1996) and pheromone response by male moths (Barrozo et al., 2010). Adult forest tent caterpillar moths have a short 5–10 day lifespan (Fitzgerald, 1995) and even a short post-mating refractory period could significantly impact the likelihood that a mated male moth would orient to pheromone and be captured in a pheromone-baited monitoring trap. The effect of mating on subsequent capture in pheromone-baited traps would be expected to be highest at high population densities.

Moth quality can be reduced at high population densities due to larval competition for nutrition and increased incidence of disease. The size of male moths affects flight behavior in several outbreaking forest lepidopteran species. Large male C. conflictana fly farther than small moths (Elliott and Evenden, 2009) and male Epirrita autumnata (Lepidoptera: Geometridae) from low-density populations fly farther than those from high-density populations despite higher wing loading of individuals at low densities (Ruohomäki, 1992). The forest tent caterpillar is a capital-breeding moth that obtains all nutrients for adult traits through larval feeding (Fitzgerald, 1995). Nutritional stress during the larval stage results in prolonged larval development in this species (Jones and Despland, 2006) and moths that emerge late in the season have smaller absolute wing area than early emerging adults (Jones and Evenden, 2008). Larval feeding by forest tent caterpillars results in both rapid- and delayed-induced responses in aspen that change the quality of the food source and negatively affects pupal mass (Parry et al., 2003). Male flight requirements to find females will likely vary with population density and by extension with availability of larval nutrition whereby less food may be available at higher densities. Recent research (Evenden et al., 2015) showed that flight capacity was not influenced by wing loading of male forest tent caterpillar moths collected from populations at different densities but energy use in flight increased with distance flown. At low population densities, males may have to fly considerable distances to encounter a receptive female (Elkinton and Cardé, 1980; Sanders, 1983). The effect of moth quality on mate location is not known for the forest tent caterpillar but large male E. autumnata are more successful in mate location by flight than small males (Tammaru et al., 1996). The effect of moth size on mate location of E. autumnata is more apparent at low than high population densities where males would need to fly greater distances to successfully locate a mate (Tammaru et al., 1996).

The purpose of this study is to determine how population density affects orientation of male forest tent caterpillar moths to pheromone sources. We test male moth response to calling females at different densities in the field. We subsequently use field cage bioassays to assess the effects of density and sex ratio on male orientation and mating success. Based on previous findings (Jones et al., 2009) in field bioassays, we predict that competition for pheromone plumes occurs at high population densities. Our experiments are also designed to determine if moth quality (size) and mating status influence mate finding and mating behavior in this species.

The results of this study show that density of forest tent caterpillar populations influences the capacity of male moths to orient to the pheromonal cues released by females. In general, female forest tent caterpillars were more successful at attracting a mate in aspen stands with high than low population densities. These data support the presence of a mate-finding Allee effect (Gascoigne et al., 2009) similar to that which has been illustrated in other forest defoliators that undergo cyclical changes in population density (Contarini et al., 2009; Régnière et al., 2013). In gypsy moth populations in Virginia, West Virginia (Sharov et al., 1995) and Wisconsin (Tcheslavskaia et al., 2002; Contarini et al., 2009), the probability of females mating increases exponentially with male moth density, as measured by moth capture in pheromone-baited traps. The relationship between mating success of female spruce budworm moths and population density is non-linear and mating levels plateau before reaching 100%, even at the highest population densities examined (Régnière et al., 2013). In the current study, non-linear models reveal that, for the forest tent caterpillar, an intermediate density promotes the highest levels of orientation to pheromone and subsequent mate finding. Similarly, trap capture of male forest tent caterpillar moths in synthetic pheromone-baited traps is highest at intermediate population densities (Jones et al., 2009). This suggests that competition for the chemical communication channel exists in dense populations of forest tent caterpillars and interferes with pheromone orientation and mate location.

Competition for pheromone sources is implicated in the reduced efficiency of pheromone-baited traps at high population densities in some forest defoliators (Delisle et al., 1998) but not in others (Östrand et al., 2007). Delisle et al. (1998) directly compare the efficiency of pheromone-baited vs. light traps at different densities of the eastern hemlock looper, Lambdina fiscellaria (Lepidoptera: Geometridae). At low population densities, pheromone-baited traps capture more moths than light traps, whereas, light traps are more efficient than pheromone traps at high population densities (Delisle et al., 1998). Although several studies suggest that competition for natural pheromone sources can reduce the efficacy of pheromone-baited traps at high population densities (Campbell et al., 1992; Delisle et al., 1998), few studies directly manipulate moth density to test this assumption. In the current study, we manipulate female and male moth density in field cages and illustrate that increased female density results in higher competition for pheromone sources and reduced orientation capacity of males to sentinel females. This can be considered as a natural mating disruption effect as orientation interference occurs when females are evenly distributed throughout the field cage space and is greater at high than low female densities. Recent application of theoretical research (Miller et al., 2006a) to empirical studies of mating disruption of moths using synthetic pheromone in various managed ecosystems (Miller et al., 2006b) shows that competition for pheromone sources is the main mechanism by which mating disruption interferes with male moth orientation to females in pheromone-treated crops. This is true for pheromone-based disruption of mating of the gypsy moth in which interference of mate location increases with the number of synthetic pheromone dispensers positioned in the study area (Tcheslavskaia et al., 2005). In our study, competitive attraction among female-produced pheromone sources appears to be driving the disruption of orientation of male forest tent caterpillar moths to sentinel females because alteration of male density in the field cages does not affect the number of male moths captured. These data provide a mechanistic understanding for the lower capture of male forest tent caterpillar moths in synthetic pheromone-baited traps at high compared to low population densities (Jones et al., 2009).

Interestingly, competition for pheromone sources at high female densities interfered with male forest tent caterpillar moth orientation to sentinel pheromone sources only when the females were evenly distributed throughout the field cages. When the females were flying freely within the cage, their presence appeared to stimulate rather than interfere with male orientation capacity to the sentinel female. Females released into the cage tended to be clumped in distribution as they alighted on the mesh covering of the cage surface. It is likely that natural mating disruption by competitive attraction does not occur when females are clumped in distribution because much of the caged area would contain clean air and allow for males to orient to the distinct pheromone signal of the sentinel females. Enhanced orientation to the sentinel female in cages containing free-flying females as compared to the control cages without females could be explained by increased activity or sensitivity of males as a result of stimulation by female pheromone. In a few instances, exposure to pheromone appears to sensitize male moths so that subsequent response to pheromone is enhanced (Anderson et al., 2003, 2007; Stelinski et al., 2004). Further research is required to measure the neurophysiological response of male forest tent caterpillars to pheromone exposure to determine if it enhances subsequent upwind orientation to pheromonal cues.

Although moth condition is known to influence mate finding behavior in other forest defoliators (Tammaru et al., 1996), we could find no evidence that moth size or mating status influence subsequent mate finding behavior in the forest tent caterpillar. Some of the male moths captured in female-baited traps in the field cage were marked with fluorescent powder indicating that they had come into contact and potentially mated with the free-flying females in the cage prior to orientation to the calling sentinel female. Male forest tent caterpillar moths are known to mate multiple times (Stehr and Cook, 1968) but this study provides evidence that males can mate more than once in a single scotophase. Mating duration varies with population density in forest tent caterpillar moths with copulations lasting an average of 85 and 187 min at low and high densities, respectively (Bieman, 1980). It is more likely that males would pursue a second mate within the same scotophase at low than high population density as most matings are restricted to a three and a half hour period commencing 2 h before sunset in the northern parts of its range (Shepherd, 1979; Bieman, 1980). This finding suggests that mated male forest tent caterpillar moths are physiologically capable of orientation to pheromone and reduction of male moth trap capture at high population densities (Jones et al., 2009) is not likely due to differential attraction of virgin male moths to pheromone sources.

The quality of nutrition obtained by forest tent caterpillar larvae is known to affect larval development (Jones and Despland, 2006; Noseworthy and Despland, 2006) and the traits of adult moths (Colasurdo et al., 2009). Both rapid- and delayed-induced responses to forest tent caterpillar feeding occur in aspen that change the quality of the food source and negatively affect pupal mass and subsequent fecundity (Parry et al., 2003). Less is known about the effect of nutrient quantity on the development and mating behavior of forest tent caterpillars, despite the implication of larval competition for food in the collapse of population outbreaks (Witter et al., 1972; Witter, 1979; Smith and Goyer, 1986). In the current study, the quantity of food provided to larvae was manipulated and well-fed larvae developed the fastest and had the highest pupal weights. There was, however, no effect of male moth body size on the proportion of released moths captured in sentinel female-baited traps. Flight capacity of male forest tent caterpillar moths is not influenced directly by wing loading but is limited by energy use (Evenden et al., 2015). As the field cage limited the dispersal distance of released moths in the current study, energy use is likely not a factor in male moth orientation to females within the cage. Under field conditions, copulation is more likely to occur between male and female forest tent caterpillar moths of similar size (Miller, 2006) which suggests that short range mate acceptance behavior occurs in this species. In the current study, individual pairings of moths of varying quality did not reveal an effect of moth quality on mating success. A future bioassay in which moths of different quality have a choice of mates might provide further insight into the occurrence of size-based assortative mating in this species.

The cyclical changes in population density of the forest tent caterpillar make it a good system to examine the impact of moth density on mate location using pheromones. The work has further implications to understanding the population dynamics of this species and the use of pheromone-baited traps to accurately monitor populations. Females positioned in traps in the field were more likely to attract male moths at sites with increasing moth density. This finding suggests that there is a mate-finding Allee effect that may limit population establishment at low population densities, as has been reported in other forest defoliators (Contarini et al., 2009; Régnière et al., 2013). Interestingly, mate attraction was greatest at intermediate population densities as the capacity of male moths to orient to females declined at sites with the highest densities. This is similar to the curvilinear relationship reported for male forest tent caterpillar moths captured in synthetic pheromone-baited traps at sites with different population densities (Jones et al., 2009). This suggests that population estimates from pheromone-baited traps are less reliable at high population densities and monitoring efforts should be adjusted to incorporate approaches that are less affected by density such as the use of light traps (Delisle et al., 1998). Closer examination of male orientation to females in field cages in which female density was manipulated revealed that the most likely mechanism for reduced mate finding at high densities was mating disruption by competitive attraction to many natural pheromone sources. Fewer male moths were attracted to sentinel females in cages with females evenly distributed throughout the field cage. Male moths spend time and energy orienting to female-produced plumes and are less likely to orient to sentinel females. This study also provides preliminary evidence for a stimulatory effect of pheromone exposure on male moth mate searching, as males in cages with free-flying females were more likely to orient to sentinel females than in cages without the presence of females. There was no evidence from our studies that lower moth condition as a result of nutrient stress from larval competition influenced mate finding or close range mating behaviors. Further experimentation is necessary to assess any potential effect of moth condition on long and short distance mating behaviors under varying population densities and field conditions.