Introduction
Nocturnal insect attraction to light sources is a phenomenon that has long been studied and so there is a large body of literature regarding this behaviour (e.g. Taylor & French, Reference Taylor and French1974; Southwood & Henderson, Reference Southwood and Henderson2000; Fayle et al., Reference Fayle, Sharp and Majerus2007). The available data are useful for detecting and monitoring pests and disease vectors (Hendricks et al., Reference Hendricks, Lingren and Hollingsworth1975; Raimondo et al., Reference Raimondo, Strazanac and Butler2004; Cowan & Gries, Reference Cowan and Gries2009; Baker et al., Reference Baker, Tann and Fitt2011), for practical nature conservation via short- or long-term studies (Conrad et al., Reference Conrad, Warren, Fox, Parsons and Woiwod2006; Kadlec et al., Reference Kadlec, Kotela, Novak, Konvicka and Jarosik2009), and for making recommendations to eliminate light pollution (Eisenbeis & Hänel, Reference Eisenbeis, Hänel, McDonnell, Hahs and Breuste2009).
Numerous types of light sources – differing both in intensity of emitted light and in spectral composition – have been used during field research (Eguchi et al., Reference Eguchi, Watanabe, Hariyama and Yamamoto1982; Kelber et al., Reference Kelber, Balkenius and Warrant2002). Based on these differences, it can be stated that insects are attracted to specific parts of the electromagnetic spectrum, and mainly to the ultraviolet (UV) range (Barghini & de Medeiros, Reference Barghini and de Medeiros2012). For this reason, those light sources emitting a higher proportion within the UV spectrum both attract greater numbers of individuals (Rydell, Reference Rydell1992) and are attractive to more taxa (Van Langevelde et al., Reference Van Langevelde, Ettema, Donners, WallisDeVries and Groenendijk2011). Moreover, previous studies using light within precisely defined spectral ranges have demonstrated there to be sex-related differences in attraction to light. Females react much less to emitted lights (Altermatt et al., Reference Altermatt, Baumeyer and Ebert2009; Cowan & Gries, Reference Cowan and Gries2009; Baker et al., Reference Baker, Tann and Fitt2011; Van Geffen et al., Reference Van Geffen, Van Eck, De Boer, Van Grunsven, Salis, Berendse and Veenendaal2015). In species wherein both sexes are fully winged, such differences could be explained by behavioural response in different parts of the reproduction cycle (Cowan & Gries, Reference Cowan and Gries2009).
Lepidoptera is an order with a high incidence of species having flightless females (Hackman, Reference Hackman1966; Heppner, Reference Heppner1991). Such females are typified by abdominal expansion and with continuous loss of wings and senses responsible for orientation in space during flight (Lau et al., Reference Lau, Gross and Meyer-Rochow2007; Meyer-Rochow & Lau, Reference Meyer-Rochow and Lau2008). Hackman (Reference Hackman1966) divided flightless females into two groups according to their mobility: the ‘heavy egg-filled type’ with very restricted movement and the ‘mobile type’ with well-developed legs. While ‘heavy’ females often remain in their cocoons, attract males by pheromones and lay eggs in the nearest surroundings, ‘mobile’ females are able to crawl short distances and lay eggs away from the place of their hatching (Hackman, Reference Hackman1966). Mobile females, despite the eyes reduction, must be able to orient in space. Their eyes retain some degree of visual sensitivity, even though they are smaller, with lower numbers of facets, and with slower adaptation to light changes relative to males (Meyer-Rochow & Lau, Reference Meyer-Rochow and Lau2008). A research with lights of different spectral composition points to the fact that such females are still able to detect the moment of sunset when experimentally increased presence of light with shorter wavelengths inhibits their activity (Van Geffen et al., Reference Van Geffen, Van Eck, De Boer, Van Grunsven, Salis, Berendse and Veenendaal2015).
Among geometrids, and especially in the subfamilies Ennominae and Larentiinae (Leraut, Reference Leraut2009; Wahlberg et al., Reference Wahlberg, Snäll, Viidalepp, Ruohomäki and Tammaru2010), there occurs a group of species with ‘mobile type’ females. Broadly polyphagous, these moths species colonize temperate woodlands in stable conditions, finding there sufficient food plant sources for larval development (Hackman, Reference Hackman1966), and frequently they occur also as defoliator pests in cyclical outbreaks (e.g. Alford, Reference Alford2000; Raymond et al., Reference Raymond, Vanbergen, Watt, Hartley, Cory and Hails2002; Tenow et al., Reference Tenow, Nilssen, Bylund and Hogstad2007). For this reason, changes in their population development are carefully monitored, primarily using diverse methods based on pheromone attraction of males (Hand et al., Reference Hand, Ellis and Stoakley1987; Szöcs et al., Reference Szöcs, Tóth, Francke, Schmidt, Philipp, König, Mori, Hansson and Löfstedt1993; Alford, Reference Alford2000). In the present study, by means of experiments in a simulator arena, we investigated (i) possible differences in visual ability within a selection of discrete ranges of precisely defined short vectors of the electromagnetic spectrum by males (as ordinary flying insects) and flightless females of nine geometrid species, mostly forest defoliators, (ii) whether or not possible differences vary among species and sexes, and (iii) if such differences in visual perceptions could be explained by the rate of eyes reduction in contrast to continual wing loss. Thus, in a case of selectivity of specific narrow part of light spectrum by studied species, we could supplement the knowledge about their eco-physiology and discuss possible improvements of the existing monitoring system.
Materials and methods
Study species
Experiments were carried out on nine species of geometrid moths with flightless females (table 1) belonging to the subfamilies Ennominae – Agriopis aurantiaria (Hübner, 1799), A. leucophaearia (Denis & Schiffermüller, 1775), A. marginaria (Fabricius, 1776), Alsophila aceraria (Denis & Schiffermüller, 1775), A. aescularia (Denis & Schiffermüller, 1775), Erannis defoliaria (Clerck, 1759), Theria rupicapraria (Denis & Schiffermüller, 1775) – and Larentiinae – Operophtera brumata (Linnaeus, 1758), O. fagata (Scharfenberg, 1805) (Leraut, Reference Leraut2009; Wahlberg et al., Reference Wahlberg, Snäll, Viidalepp, Ruohomäki and Tammaru2010). These crepuscular and nocturnal species are commonly distributed in forest habitats throughout Central Europe, where they often constitute a majority of late autumn and early spring lepidopteran communities (Leraut, Reference Leraut2009). With the exception of T. rupicapraria (an oligophagous species feeding on shrubs of the family Rosaceae), all the species are widely polyphagous on deciduous trees and shrubs, and most of them are among the heavy defoliators of forest stands (Alford, Reference Alford2000). Adults are active from late September to December (A. aurantiaria, A. aceraria, E. defoliaria, O. brumata, O. fagata) or from January to April (other species) (Leraut, Reference Leraut2009). Females are wholly apterous or with much-reduced wings (brachypterous). After hatching, they crawl to the nearest tree or shrub to wait for males with which to mate; they then crawl up into the tree or shrub, where they lay eggs (Hackman, Reference Hackman1966; Van Dongen et al., Reference Van Dongen, Matthysen, Sprengers and Dhondt1998; Meyer-Rochow & Lau, Reference Meyer-Rochow and Lau2008; Leraut, Reference Leraut2009).
Morphometric characteristics of studied species and total numbers of individuals and of individual measurements carried out in the experiment.
No. indiv., number of individuals per sex used in experiment; No. measur., total number of individual measurements per sex; M, males; F, females; BRA, brachypterous females; APT, apterous females.
Moths sampling
Individuals were collected always one night before an experiment during the time when adults are active (i.e. October–November 2013, March–April 2014) in an older deciduous forest stand dominated by oaks and beech near Kromeriz, Czech Republic (49°08′28″N, 17°15′32″ E, altitude: 550 m a.s.l.). Moths were searched by torchlight from twilight to midnight by walking through stands. Males were captured by insect net directly during flight or were found sitting on trees and shrubs; all observed females were captured during crawling or sitting on tree trunks and shrubs. Individuals were held separately by sex and species in plastic boxes under shady conditions and natural temperatures until the next night.
Experimental design and moth characteristics
The sensitivity of moths’ eyes to various light spectra was tested in a cylindrical experimental arena (3 m in diameter, 2 m high). The walls were made from black light-impermeable polyester textile. The interior was thus protected against unwanted light penetration from the surrounding environment (Eisenbeis & Hänel, Reference Eisenbeis, Hänel, McDonnell, Hahs and Breuste2009). The arena was installed under shelter in outdoor conditions to avoid possible behavioural changes due to unnatural temperatures and the experiments were realized under temperatures when adults are active (up to −2°C; Lorentzen, Reference Lorentzen1974). In the centre of the arena there was installed a transparent box, in which the tested moths were held for 15 min in order to adapt to internal conditions and become orientated.
To test attraction, an electromagnetic stimulus having its main portion in the visible light range and the remainder within the UV region (360–660 nm) was selected. The test spectrum was divided into two repetitions, each of which was subdivided into 28 spectral vectors (fig. 1). Each spectral vector was emitted from a monochromatic light-emitting diode (LED) with a specific wavelength of radiated light. To avoid possible influences caused by differences in radiant flux of one wavelength of LED versus others (Cowan & Gries, Reference Cowan and Gries2009), the radiative flux of each LED was calibrated to 2.5 mW. LEDs were installed around the inside perimeter of the arena in order to maintain an equal distance (17 cm) between each succeeding LED, and separate vectors were linked to one another by their progressively increasing wavelengths (fig. 1). This placement of lights allowed moths gradually to choose the most attractive narrow parts of radiation. To minimalize the effects of reflected light on selectiveness of LEDs, together with small radiative flux, the light was emitted from sources in narrow viewing angle (maximally 30°) orientated directly to the starting place in the centre of arena.
Schematic representation of the experimental light arena. Dots indicate positions of monochromatic LEDs around the walls (separated by approximately 17 cm). Arrows show the direction of increasing light-wavelengths in 10 nm increments. The inner square indicates the position of the starting box.
Moths were tested always in the night following their collection and during the time of their natural activity (twilight to midnight). Each night, experiments were made first with males and then with females. After completing the experiments, the entirety of the inner walls and starting box was cleaned with 70% ethanol. During each experiment, up to four individuals of the same sex and species were tested together. Each was marked using a permanent marker to ensure its consistent identification. After assimilation in the starting box, moths were released and the precise spectrum of the LED reached was recorded. When tested moths reached a certain diode, the measurement was recorded as positive reaction to the light source and the exact spectrum of LED was noted. If there was no reaction within 15 min after release (i.e. a moth remained at the starting point), the measurement was recorded as negative. Four consecutive measurements were made for each individual.
After the experiment, the moths were euthanized by chloroform and eye diameter (EYE SIZE) and dry body mass (MASS) were measured. Each individual was dried in an oven at 80°C for 12 h and weighed on an analytical scale (Van Langevelde et al., Reference Van Langevelde, Ettema, Donners, WallisDeVries and Groenendijk2011). Moths were then decapitated and EYE SIZE was measured by cellSens Entry 1.6 (Olympus, Tokyo, Japan) using DP73 camera attached to Olympus SZX16 stereomicroscope.
Data analysis
We used generalized linear mixed-effects models (GLMM) for hypothesis testing. All models were analysed using gamma distribution of errors. All analyses were run in the program R (R Development Core Team 2012) using the lme4 package (Bates et al., Reference Bates, Maechler, Bolker, Walker, Christensen, Singmann and Dai2014).
A first analysis focused on the differences between species and/or sexes in their reaction to lights of various wavelengths. The spectral length (LENGTH) for the LED reached by the moths in each individual experiment was the dependent variable. In view of the fact that multiple measurements were made for a given individual, which were possibly able to change the selection of the lights after the first experience with the conditions of the arena, the identity of the exact individual (INDIVIDUAL) and trial run (TRIAL RUN) within the measurements of the same individual were specified as a random effects. Thus, the model distinguish the intra-individual effects. SPECIES and SEX (male, female) were added as the fixed effects of main interest.
A second analysis endeavoured to explain EYE SIZE as a possible response to continuous wing reduction. In these models, precise eye size was used as the dependent variable, whereas species was taken as a random effect. Wings type (WING) was specified as FULL (fully developed wings in males), BRACHYPTERY (at least small remnants in females), or APTERY (without any wings in females) and used as a fixed effect. In addition, as the eye size depended strongly on discrete body size due to allometry (Shingleton et al., Reference Shingleton, Mirth and Bates2008), MASS was added as another fixed effect in the first position of the model. A similar model with SPECIES as random factor was fitted to test the dependence of MASS on WING type. Tukey's HSD tests were used to examine the differences between levels of significant factorial predictors.
The most parsimonious models were selected by comparing AIC values of model fits. This procedure compared models by balancing their complexity and goodness of fit (Akaike, Reference Akaike1974). Null models with random effects were fitted first, and then these were mutually compared with the more complex models with fixed effects and their interactions. Those models with the lowest AIC and with ΔAIC ≤ 2 were selected as best fitting to the data. We used a χ2 test as a probability function to express significance of the models.
To distinguish whether or not some precise wavelengths are preferred by individuals with different eye size and wing type, a canonical correspondence analysis (CCA) was run in Canoco for Windows 4.5 (Ter Braak & Smilauer, Reference Ter Braak and Smilauer2002). As response variables (referred to in the software as ‘species data’), the light wavelength preferences (measured as the number of observations with positive reactions to each light per each individual) were used. The predictors (‘environmental data’ in the software) used were relative eye size (expressed as RATIO between EYE SIZE and MASS) and wing type (FULL, BRACHYPTERY or APTERY). The full model was then simplified by forward selection procedure and only significant predictors were selected. A Monte Carlo permutation test (999 runs, full model) was used to express the significance of final ordinations containing only significant predictors. All models were fitted with and without affinity of individuals to SPECIES to check the possible influence at species level.
Results
Overall, the experiment evaluated a sample of 204 individuals (112 males, 92 females [73 brachypterous, 19 apterous]) of nine geometrid species collected under natural conditions (table 1). Sexes differed in their reaction during experiments. Only five individual measurements of males from a total of 448 were negative, while in the case of females 144 measurements (from 368) were negative. Overall, 443 of the individual measurements of males and 224 of those for females were positive (table 1).
In analysing attractiveness of lights with different wavelengths, SEX was the only significant predictor and no effect of SPECIES was detected (table 2). All other tested models had ΔAIC > 2 and did not improve significance over the previous model. Males and females differed strongly in their reactions to light sources (fig. 2). While males rather preferred lights with shorter wavelengths (mean ± SD: 402 ± 51 nm), females varied much more in their light selection (mean ± SD: 478 ± 104 nm) and it was impossible to detect any preferred spectral area for them.
Comparison of the preference for lights with distinct spectral composition by males and females in nine studied geometrid species. n – number of individual measurements for each species. Box-plot attributes: minimum value, 25th percentile, median, 75th percentile, maximum value and outliers (empty circles) are shown.
Results of general linear mixed-effects models showing the relationships between response variable (LENGTH) and tested predictors. Random effects in models were INDIVIDUAL and TRIAL RUN within the measurements of each individual.
AIC, Akaike information criterion; NULL, null model only with random effect.
***P < 0.001, ns – P > 0.05.
Differences in individual EYE SIZE were best explained by a model, which contained the variables MASS, WING, and their interaction (table 3), but the interaction had only weak effect on the response of main effects (ΔAIC between the model with and without interaction is only about 3; table 3). EYE SIZE differed significantly between individuals with different wing type (Tukey's HSD test in GLMM: APTERY–BRACHYPTERY: z = 5.20, P < 0.001; APTERY–FULL: z = 16.06, P < 0.001, BRACHYPTERY–FULL: z = 19.05, P < 0.001; table 1, fig. 3A), with EYE SIZE changing from smallest in apterous females (mean ± SD: 0.5836 ± 0.0733 mm) through brachypterous females (mean ± SD: 0.6336 ± 0.0667 mm) to fully winged males (mean ± SD: 0.8366 ± 0.1164 mm). This effect was significant also after withdrawing the effects of individual MASS (table 3). Individuals with different WING type differed strongly in their MASS, but even across species this effect related only to sex and not species (table 1, fig. 3B). While fully winged males were lighter than females (Tukey's HSD test: APTERY–FULL: z = −5.75, P < 0.0001; BRACHYPTERY–FULL: z = −4.90, P < 0.0001), apterous and brachypterous females did not differ significantly by MASS (Tukey's HSD test: APTERY–BRACHYPTERY: z = −0.58, P > 0.05).
Comparison of the individual morphometric attributes (A – eye size, B – dry body mass) of nine studied geometrid moth species with distinct reduction of wings. n – number of individuals for each wing type. Box-plot attributes: see fig. 1 for description.
Results of general linear mixed-effects models showing the relationships between response variable (EYE SIZE) and tested predictors. Random effect in models was SPECIES.
AIC, Akaike information criterion; NULL, null model only with random effect.
***P < 0.001, *P < 0.05.
CCA ordination pointed to differences in preferred spectral lengths by individuals with different WING types (test of significance of first axes: eigenvalue = 0.248, F = 5.579, P = 0.001. All canonical axes (trace = 0.321, F = 3.636, P = 0.001) explained the 3.49% variability in species data (fig. 4). RATIO was observed to be insignificant during forward selection (F = 0.872, P = 0.635). Thus, the fully winged males flew more often to lights with shorter wavelengths (360–510 nm) than did females. Brachypterous females tended slightly more toward shorter wavelengths than did apterous females. When filtering out the effects of affinity to SPECIES in the covariate model, the explained variability of species data by WING types decreased to 2.64%. Nevertheless, this was still highly significant (test of significance of first axes: eigenvalue = 0.169, F = 3.769, P = 0.001; all canonical axes: trace = 0.233, F = 2.613, P = 0.001).
Canonical correspondence analysis diagram of relationship between lights with specific wavelengths reached by moths with different wing types (triangles). Monte-Carlo test with 999 permutations: test of significance of first axes: eigenvalue = 0.248, F = 5.579, P = 0.001; all canonical axes: trace = 0.321, F = 3.636, P = 0.001.
Discussion
Fully winged males and flightless females differed in their reactions and preferences to lights with varying spectral composition. While males flew more often to shorter wavelengths (under 500 nm), females varied more widely in their choice. This is in accordance with the presumption that males react as flying insects and have better abilities for orientation in space (Meyer-Rochow & Lau, Reference Meyer-Rochow and Lau2008), while wingless females are more disadvantaged by their reduction of wings and perceptions. Even though the geometrids’ wingless females in the studied species may in several cases have developed independently (Snäll et al., Reference Snäll, Tammaru, Wahlberg, Viidalepp, Ruohomäki, Savontaus and Huoponen2007; Wahlberg et al., Reference Wahlberg, Snäll, Viidalepp, Ruohomäki and Tammaru2010), the move-to-lights differences between sexes were similar for all tested species.
These similarities imply by way of the so-called ‘winter moth syndrome’ (Hackman, Reference Hackman1966; Barbosa et al., Reference Barbosa, Krischik and Lance1989) a rather convergent adaptation to the environmental conditions of temperate European climate that is typical for late autumn or early spring (cold weather, insufficient nutrients for adults) with decreased adult predation by vertebrates (Buse et al., Reference Buse, Dury, Woodburn, Perrins and Good1999; Snäll et al., Reference Snäll, Tammaru, Wahlberg, Viidalepp, Ruohomäki, Savontaus and Huoponen2007; Wahlberg et al., Reference Wahlberg, Snäll, Viidalepp, Ruohomäki and Tammaru2010). In such conditions, males comprise the dispersed sex, which is flying in space and trying to find a female climbing on tree trunks or shrubs with which to mate (e.g. Van Dongen et al., Reference Van Dongen, Matthysen, Sprengers and Dhondt1998; Van Dongen et al., Reference Van Dongen, Sprengers, Löfstedt and Matthysen1999). Females with reduced wings and wing muscles (Hackman, Reference Hackman1966; Heppner, Reference Heppner1991) should allocate more energy to reproduction. While polymorph females in Lepidoptera are rather rare (e.g. Heppner, Reference Heppner1991; Lau et al., Reference Lau, Gross and Meyer-Rochow2007), this relationship has been more studied in other insect orders with macropterous and brachypterous females (Denno et al., Reference Denno, Olmstead and McCloud1989; Zera & Denno, Reference Zera and Denno1997). For example, brachypterous females of the planthopper Prokelisia dolus have been shown to have greater overall fecundity than fully winged migrating forms (Denno et al., Reference Denno, Olmstead and McCloud1989). In resolving a trade-off between flying and overall fecundity, therefore, flightless and pheromone-producing females (Svensson, Reference Svensson1996) reduced their sensory inputs, which would be necessary during active flight (Rydell et al., Reference Rydell, Skals, Surlykke and Svensson1997; Meyer-Rochow & Lau, Reference Meyer-Rochow and Lau2008). These eventually became superfluous and so, in comparison with males, they increased their body size via hypertrophy of their reproductive organs (Hackman, Reference Hackman1966). The results of our experiments correspond with this knowledge, as the females tested were significantly heavier than males, had reduced eyes, and their selection of lights spectra was poorer. The females’ eyes were significantly smaller than those of the males, and according to Meyer-Rochow & Lau (Reference Meyer-Rochow and Lau2008), the eyes of flightless females also have decreased sensitivity and poorer resolution.
On the other hand, these differences in light preferences between the sexes may also be explained in part by the theory of behavioural and physiological adaptations to changes in light conditions during twilight and night (Johnsen et al., Reference Johnsen, Kelber, Warrant, Sweeney, Widder, Lee and Hernández-Andrés2006). Directly after sunset, the light conditions are dominated by spectra with shorter wavelengths near 450 nm and these change to longer wavelengths during the night (Johnsen et al., Reference Johnsen, Kelber, Warrant, Sweeney, Widder, Lee and Hernández-Andrés2006). Males of majority of the studied species are active during twilight, and, despite their lesser flying ability (Heinrich & Mommsen, Reference Heinrich and Mommsen1985; Van Dongen et al., Reference Van Dongen, Matthysen and Dhondt1996, Leggett et al., Reference Leggett, Jones, Burke, Hails, Sait and Boots2011), their eyes are fully adapted to orientation in space under changing light conditions, albeit still with short wavelength dominance (Meyer-Rochow & Lau, Reference Meyer-Rochow and Lau2008). In general, flying moths’ eyes contain blue receptors sensitive to shorter wavelengths (Eguchi et al., Reference Eguchi, Watanabe, Hariyama and Yamamoto1982; Cutler et al., Reference Cutler, Bennett, Stevenson and White1995; Briscoe & Chittka, Reference Briscoe and Chittka2001) and the fliers are better orientated in such light type (Cowan & Gries, Reference Cowan and Gries2009). Thus, a male preference for shorter wavelengths during the experiments should be explained by better orientation in space within which lights have such composition.
Females, meanwhile, must be active from twilight, when they are crawling on trees and attracting males (Van Dongen et al., Reference Van Dongen, Matthysen, Sprengers and Dhondt1998), into the dark night when laying eggs. They, too, need a certain orientation in space, albeit not with such high resolution (Meyer-Rochow & Lau, Reference Meyer-Rochow and Lau2008). This implies that during orientation, insects’ eyes should be able to shift their functioning to longer wavelengths (Johnsen et al., Reference Johnsen, Kelber, Warrant, Sweeney, Widder, Lee and Hernández-Andrés2006). Thus, the females should be able to perform different behavioural activities under different light conditions. Moreover, females are able to change their affinity to different parts of the light spectrum according to their particular motivations. In the Indian meal moth (Plodia interpunctella) it has been observed, for example, that unmated females preferred slightly different wavelengths than did mated females or males (Cowan & Gries, Reference Cowan and Gries2009). The studied females during our experiments could have been in varying stages of the reproduction cycle (mated versus non-mated), and, inasmuch as such difference was not recognized, that could have been the reason why the females crawled toward such a wide range of the light spectrum. Moreover, it could have been that the unfertilized females were not motivated to react to any part of the light spectra whatsoever or that females were suppressed in activity and disoriented by light in general, and that accounts for the large proportion of negative reactions among females (Cowan & Gries, Reference Cowan and Gries2009; Van Geffen et al., Reference Van Geffen, Van Eck, De Boer, Van Grunsven, Salis, Berendse and Veenendaal2015).
Whereas males preferred a relatively distinct range of shorter light wavelengths – in principle across all tested species – the use of a distinct narrow spectrum of light for attracting moths could markedly enhance the effectiveness of established methods used in forest pest monitoring systems, which mainly involve species-specific pheromone attraction of males (Hand et al., Reference Hand, Ellis and Stoakley1987; Szöcs et al., Reference Szöcs, Tóth, Francke, Schmidt, Philipp, König, Mori, Hansson and Löfstedt1993; Alford, Reference Alford2000). A combining of these two methods has been tested, for example, on the forest pest Operophtera brumata and it was found out that more males were caught to pheromone traps illuminated by light with shorter wavelengths (Van Geffen et al., Reference Van Geffen, Van Eck, De Boer, Van Grunsven, Salis, Berendse and Veenendaal2015). A similar mixed method (chemical and light attractant) has been tested with success also on stored-product beetle pests Tribolium castaneum (Duehl et al., Reference Duehl, Cohnstaedt, Arbogast and Teal2011) and Cylas formicarius (McQuate, Reference McQuate2014), with increasing of the probability of adults’ detection by mixed pheromone and light traps, or on mosquitoes (Chen et al., Reference Chen, Wang, Teng, Chen, Chang, Lu, Lin, Jian and Wu2011). Studies on Lepidoptera without mixed treatment have shown that light-trapping can be used with similar efficiency as other monitoring methods (Raimondo et al., Reference Raimondo, Strazanac and Butler2004; Baker et al., Reference Baker, Tann and Fitt2011). Whereas the selected spectral range is attractive also to other groups of Lepidoptera (e.g. Van Langevelde et al., Reference Van Langevelde, Ettema, Donners, WallisDeVries and Groenendijk2011), it should be expected – by using of combined pheromone-light traps – higher presence of other species in trapped samples. On the other hand, during autumn, winter, and spring only low diversity of moths occurs in forest stands of the temperate zone, and a substantial part of the trapped insects will consist, too, of such other monitored forest pests as within the genera Conistra, Orthosia or Pannolis (Alford, Reference Alford2000).
Our results show there to be differences in move-to-light reactions between the sexes and continual changes in relative eye size and body mass and thus continual repression of redundant senses. In view of the fact that females (and males) should be able to spend only energetic resources allocated during the larval stage (Van Dongen et al., Reference Van Dongen, Sprengers, Löfstedt and Matthysen1999; Jervis et al., Reference Jervis, Boggs and Ferns2005), brachypterous females should be more penalized in their overall fecundity than apterous females (Rhainds et al., Reference Rhainds, Leather and Sadof2008). We should note that we recognize that our results can be biased by unequal data, and especially for females.
A number of questions arise for further research in this area, including (i) how these findings might need to be adjusted after including observations for heavy apterous females (e.g. E. defoliara and the genera Lycia and Phigalia), (ii) are we able to improve efficiency of existing monitoring systems in both sexes, especially females, using light-trapping methods, (iii) whether or not the two types of females differ in their fecundity, and (iv) whether or not they differ in their nocturnal activity under residual light with long wavelengths. As implied by the multivariate analysis, the two types differ very slightly in response to spectral range. One explanation could lie in different timing of activity during night-time, but this should be more species specific, as, for example, wingless of Phigalia pilosaria females are active directly at twilight while the females of E. defoliaria are more active later in the night (Pikner M., personal observation). If such differences would be proven, then the ‘mobile-type’ of flightless females (Hackman, Reference Hackman1966) should be divided into two different groups of females having different vision abilities and reproduction success.
Acknowledgements
The authors would like to thank Jan Ruzicka for help with measuring the moths’ characteristics and Zdenek Fric for useful discussion about the data analysis. Funding support was provided by the Internal Grant Agency of the Faculty of Environmental Sciences, Czech University of Life Sciences Prague (20144259).
