Introduction
Dispersal and reproduction are critical processes for population maintenance and evolution. Dispersal behaviour is defined as the spread of individuals away from others or from their breeding substrate (Begon et al., Reference Begon, Townsend and Harper2009) and, for most insects, dispersal involves flight (Johnson, Reference Johnson1969). Dispersal may occur at different spatial scales, from searching for food in a local patch to migratory movements at the regional or continental scale (Ims, Reference Ims, Hansson, Fahrig and Merriam1995). Insect flight is known to be energetically the most costly physiological process in the animal kingdom (Candy et al., Reference Candy, Becker and Wegener1997). The metabolic rates of flying insects can be 50–100 times higher than those of resting animals (Beenakkers et al., Reference Beenakkers, Van der Horst and Van Marrewijk1984). Flight requires biosynthesis and storage of large quantities of flight fuel, in the form of blood sugars, glycogen and lipids, or a mixture of these compounds for long-distance flight (Chapman et al., Reference Chapman, Simpson and Douglas2013). In insects there is often a trade-off between the flight capacities and other energy-expensive vital functions, such as reproduction. The construction of a flight apparatus and the production of flight fuel compete for energy with ovarian development and oogenesis (Rankin & Burchsted, Reference Rankin and Burchsted1992; Ronce, Reference Ronce2007). Zera & Denno (Reference Zera and Denno1997) showed that the energetic cost of fuel biosynthesis for flight would probably lead to a decrease in the fitness of dispersing individuals. Dispersal and reproduction are therefore, dependent on the available energy budget which is based on acquisition of nutrients during the larval and/or adult stages (Boggs & Ross, Reference Boggs and Ross1993; Coll & Yuval, Reference Coll and Yuval2004).
Resource allocation in dispersing and reproducing adults can be seen as the expenditure of capital resources accumulated and stored at earlier life stages or as the expenditure of income resources acquired during the adult stage (Stearns, Reference Stearns1992). Jervis et al. (Reference Jervis, Boggs and Ferns2005) proposed two extreme strategies linking allocation strategy and ovarian development in holometabolous insects. ‘Capital breeder’ species invest in reproduction, the energy principally accumulated during the larval stages. Species of this type are pro-ovigenic and adults contain mature eggs at emergence. By contrast, ‘income breeder’ species acquire the energy for reproduction from feeding during the adult stage. The adults have no mature eggs at emergence and are synovigenic. An extreme example of ‘capital breeder’ in the Cerembycidae family are the Prioninae long-horn beetles. Interestingly, the adults do not feed and in this sub-family gametogenesis is completed by the end of the larval stage, after which gonads degenerate. On the opposite, the Lamiinae sub-family is an example of ‘income breeder’. The adults feed and gametes production continues throughout the adult life (Edwards, Reference Edwards1961).
However, many insects, especially long-lived species, have the potential to allocate both larval-derived capital and adult-acquired income resources to reproductive function (Jervis et al., Reference Jervis, Boggs and Ferns2007, Reference Jervis, Ellers and Harvey2008). Long-lived species usually invest more energy resources into building a robust body during metamorphosis, whereas short-lived species would invest more in egg production. In many insects, investment in the thoracic musculature and wings (which control aerial dispersal capability) is thought to occur at the expense of the investment in reproductive tissues and eggs, or vice versa, in accordance with the oogenesis–flight syndrome theory (Dixon et al., Reference Dixon, Horth and Kindlmann1993). This trade-off between dispersal and reproduction has been studied mostly in dimorphic insect species, such as crickets like Gryllus firmus Scudder (Orthoptera: Gryllidae) and several aphid species (Mole & Zera, Reference Mole and Zera1994; Guerra, Reference Guerra2011). Only a few studies have focused on monomorphic insects, such as Cydia pomonella Linnaeus (Lepidoptera: Tortricidae) (Gu et al., Reference Gu, Hughes and Dorn2006) or Chrysoperla sinica Tjeder (Neuroptera: Chrysopidae) (Khuhro et al., Reference Khuhro, Biondi, Desneux, Zhang, Zhang and Chen2014) but with a particular focus on the reproductive side of the trade-off.
We propose here to extend the conceptual framework of ‘capital’ vs. ‘income’ life-history strategies to dispersal with the concepts of ‘capital disperser’ and ‘income disperser’. Our prediction is that, the strategy for allocating energy to dispersal is adapted to the availability of food resources to larval vs. adult stages in insects. The ‘income disperser’ strategy allows species to deal with rare feeding resource for larvae, but abundant for adults. For example adults of the common pine shoot beetle Tomicus piniperda Linnaeus (Coleoptera: Scolytidae) feed on fresh pine shoots, which are very abundant in the pine forests; whereas larvae develop on the decaying trees which are rare and scattered in the forest landscape (Lieutier et al., Reference Lieutier, Day, Battisti, Gregoire and Evans2004). By contrast the ‘capital disperser’ strategy should concern insects with abundant breeding substrate (for larvae), but rare feeding resources for adults. An extreme is the case of insects with non-feeding adults. as Prioninae.
Logically ‘income dispersers’ would have long-lived adults (T. piniperda beetles may live at least 3 years) and ‘capital dispersers’ short-lived adults which is the case of most of the Prioninae (Edwards, Reference Edwards1961). Lastly, the ‘income disperser’ strategy may necessitate a maturation period at the beginning of the adult stage, such as in T. piniperda, but also in the Colorado potato beetle, Leptinotarsa decemlineata Say (Coleoptera: Chrysomelidae) which needs to assimilate lipids before being able to fuel flight (Weeda et al., Reference Weeda, de Kort and Beenakkers1979). Resource availability and accessibility might also drive the trade-off between energy allocation to the dispersal and reproduction. According to Glazier (Reference Glazier1999) favourable conditions allow animals to invest energy in several traits resulting in an apparent absence of trade-off but this has been rarely tested in insects.
This study focused on Monochamus galloprovincialis Olivier (Coleoptera: Cerambycidae) which is a long-lived beetle, with a generation time of 236 ± 47 days and adult longevity of 63 ± 6 days (Naves et al., Reference Naves, Sousa and Quartau2006a; Akbulut et al., Reference Akbulut, Keten and Stamps2008). This beetle is a source of growing concern because of its role as a main vector of the pine wood nematode (PWN), since its introduction into Europe (Sousa et al., Reference Sousa, Bonifácio, Pires, Penas, Mota and Bravo2001). The PWN, Bursaphelenchus xylophilus (Steiner and Buher) Nickle is responsible for pine wilt disease, which caused extensive tree mortality in Asia (Cheng et al., Reference Cheng, Lin, Li and Fang1983) and now constitutes the principal threat to pine forests in Southern Europe. The natural spread of PWN depends on the dispersal ability of its insect vector (David et al., Reference David, Giffard, Piou and Jactel2014). Beetles of the genus Monochamus need dead or decaying pines for larval development (Naves et al., Reference Naves, Sousa and Quartau2006b). This resource is often rare and scattered across the forest landscape, unlike the healthy trees required for adult maturation feeding. Indeed, after emergence, adult M. galloprovincialis beetles spend about 3 weeks feeding on the fresh pine shoots, during which time they mature and begin inoculating healthy trees with nematodes (Evans et al., Reference Evans, McNamara, Braasch, Chadoeuf and Magnusson1996; Naves et al., Reference Naves, Camacho, Sousa and Quartau2007). It is therefore, of great importance to know more about the establishment of flight capacity and the possible trade-offs between flight capacity and investment in reproduction for immature adults of M. galloprovincialis. This could be helpful to improve our predictions on the spread of PWN disease as the number of nematode transmissions mainly depends on the number of times young adult beetles have to fly and feed on the fresh pine shoots.
To summarize, M. galloprovincialis has a long lifespan, feeding resources (fresh pine shoots) can be abundant and well spread; whereas oviposition substrate (decaying trees) is often rare and scattered in the landscape. We can therefore, hypothesize that females are synovigenic (Jervis et al., Reference Jervis, Ellers and Harvey2008), use capital resources to invest in flight musculature and adult income resources for both oogenesis and long flights in search for oviposition substrates. By contrast, males are expected to invest all the income resources to sustain flight in search for females. Indeed spermatogenesis is energetically less costly than oogenesis in insects (Fadamiro et al., Reference Fadamiro, Chen, Onagbola and Graham2005).
Based on these hypotheses, we predicted that: (1) the flight capacity of M. galloprovincialis increases with the ageing of beetles, during the maturation period, due to the accumulation of flight fuel through pine shoot feeding; and (2) flight capacity is sex-dependent, due to competition for fuel allocation between reproduction and dispersal in females. To test these hypotheses, we estimated the flight capacity of beetles with a flight mill apparatus and used the number of mature eggs as a proxy for the investment of females in reproduction. We used the ratio between the dry weights of thorax and abdomen to estimate the trade-off in energy allocation between the dispersal and reproduction during the early stages of adult life.
Materials and methods
Insect material
Adult beetles were obtained from fresh deadwood of the maritime pine (Pinus pinaster Aiton), collected in three pine stands located close together in Cestas, France (44°44′43″N; 0°40′52″W), in January and February 2012. Logs were stored in a climatic chamber until the beetles emerged (temperature 21 ± 1°C and L:D 16 h:8 h). Newly emerged adult insects were collected daily, marked individually with white spots on the elytra, and weighed to the nearest 1 µg with a precision balance (Mettler Toledo, Semi-Micro MS205DU). This initial weight was defined as the weight at emergence (W 0). We recorded the sex of each beetle and the length of its right elytron. Beetles were kept in plastic boxes (35 × 45 × 24 cm3), with a maximum of ten beetles per box. All the beetles in a given box were of the same sex. Boxes were stored in a climatic chamber, under the same controlled conditions as for the pine logs.
For studying the effect of maturation, we used ten males and ten females of five age classes (0, 5, 10, 20 and 30 days old), except for males of 0 days and females of 10 days of age, for which we only used nine individuals. In total, we used 49 males and 49 females of M. galloprovincialis in the experiment. The beetles were supplied with fresh maritime pine shoots without restriction, until testing for flight performance. They were not able to fly in the rearing boxes before testing.
Flight mill recordings
The flight mill design was adapted from that described by Atkins (Reference Atkins1961) and Jactel & Gaillard (Reference Jactel and Gaillard1991). For a detailed description see (David et al., Reference David, Giffard, Piou and Jactel2014). Flight was initiated by a gentle air flow blown by the experimenter infront of the beetle's head. Each beetle was tested for 10 min. This test time was considered a good compromise between flight durations long enough to compare flight performances (The total flight duration during a 2 h session was significantly correlated with the time flown during the first 10 min of each session; see Appendix 1 in Supplementary material) and short enough to avoid total energy consumption. Beetles were considered ‘fliers’ if they were able to complete more than 30 s of cumulated flight over the 10 min test, based on the bimodal shape of the frequency distribution of cumulated flight durations (Appendix 2 in Supplementary material). For each insect, we analysed flight performance as the probability of flying and the total distance flown during the 10 min test. All insects were weighed before the flight recording, and the value obtained was considered to be the weight at age t (W t ). After the test, the insects were frozen at −20°C until biochemical analyses.
Energy reserves and muscle extractions
All the tested beetles were divided into four parts: the pterothorax, the abdomen, the eggs and the remaining parts (i.e. head, pronotum, legs and wings). We counted the total number of mature eggs (with chorion) in the abdomen of each female. All the body parts were then dried individually in a desiccator for 24 h at 60°C and their dry weight was measured with a precision balance (accuracy: ±1 µg). Most insects used lipids or carbohydrates to fuel flight. For example, bees (Suarez et al., Reference Suarez, Darveau, Welch, O'Brien, Roubik and Hochachka2005) and parasitoid wasps (Amat et al., Reference Amat, Besnard, Foray, Pelosse, Bernstein and Desouhant2012) use only carbohydrates. By contrast, most Coleoptera, such as the Colorado potato beetle (Weeda et al., Reference Weeda, de Kort and Beenakkers1979), Phryneta spinator Fabricius (Coleoptera: Cerambycidae) and Ceroplesis thunbergii Fahraeus (Coleoptera: Cerambycidae) (Gäde & Auerswald, Reference Gäde and Auerswald2000) mainly use proline to fuel flight which corresponds to an indirect use of lipids, as the synthesis of proline requires mobilization of fatty acids (Gäde & Auerswald, Reference Gäde and Auerswald2002). To estimate the quantity of reserves (lipids content) allocated to flight activities, we immersed the pterothorax in 2 ml of petroleum ether for 48 h at 40°C. This body part was then desiccated again for 24 h at 60°C and reweighed (Ellers & Van Alphen, Reference Ellers and Van Alphen1997). The difference between the first and the second dry weight values was considered the energy reserve content. Finally, for the quantification of flight muscles, the pterothorax was immersed in 2 ml of 10% KOH solution for 24 h at 40°C, re-desiccated and reweighed (Fischer & Kutsch, Reference Fischer and Kutsch2000). The difference between the thorax dry weight after reserve extraction and the final dry weight was considered the weight of flight muscles.
Statistical analyses
Effect of maturation on flight performances and resource allocation
The fresh weight of the beetle at emergence W 0 and its size (elytron length) were highly correlated (r = 0.91, P < 0.001) and could not, therefore, be used as independent variables in the same model. We used only body weight in subsequent analyses, because this approach was more consistent with our hypothesis of weight gain during the maturation period.
We assessed the effects of weight at emergence (W 0), sex and age on flight performances, i.e., the probability of flying and the distance flown. The probability of flying was assessed with a generalized linear model (GLM), using a binomial family with a logit link. The distance flown was tested only for ‘fliers’, with a linear model. We tested the effects of W 0 and age on the number of mature eggs for females, using a quasi-Poisson GLM with a log link (count data with overdispersion). We analysed the effect of ageing during maturation on resource allocation, by using linear models to assess the effects of age and sex on the weight of entire insect at age t (W t ), on the dry weight of thorax (W th ) and on the dry weight of abdomen for males and the sum of dry weight of abdomen and dry weight of eggs for females (W ab ).
We explored the effect of ageing during maturation on the energy allocation trade-off between dispersal (building of flight apparatus) and reproduction, using the ratio between the dry weights of the thorax and abdomen (W th /W ab ).We used linear models to assess the effects of age and sex on this ratio. In all models, we used the weight at emergence (W 0) as covariable to take differences in the beetle size into account (Raubenheimer, Reference Raubenheimer1995). The inclusion of log-transformed age improved AIC values (Akaike's Information Criterion) in all models. We therefore, included a log-transformed age variable in all subsequent statistical analyses.
Effect of weight allocation on dispersal capacities
We investigated the effect on the distance flown of relative thorax weight (rW th = W th /W 0), proportion of muscles in the thorax (%th mus = Wth mus /W th ), proportion of reserves in the thorax (%th res = Wth res /W th ) and relative abdomen weight (rW ab = W ab /W 0), with Wth mus and Wth res the dry weight of muscles and energy reserves in the thorax. We used relative weights (i.e. dividing by total weight at emergence) rather than absolute weights, to take into account differences in beetle size.
Each of the four explanatory variables was tested in separate linear models, together with their interactions with sex as a factor. All the models were simplified, by the removal of non-significant interaction terms. The significance of each effect term was assessed by comparing the deviances of models fitted with and without each effect term, by F tests for linear models or χ2 tests for binomial and quasi-Poisson GLM (Crawley, Reference Crawley2012). For each model, we checked that the residuals complied with the assumptions of normality and homoscedasticity. All statistical analyses were carried out in the R statistical environment (R development core team, 2013).
Results
Effect of maturation on flight performances and energy allocation
Overall, 45% of the beetles (44 out of 98 individuals) were classified as fliers. The probability of flying significantly increased with the weight at emergence but was not affected by sex or age (table 1). The distance flown on flight mill significantly increased with increasing age (fig. 1), but not with weight at emergence. There was no difference in the distance flown between the sexes (table 1).
Effect of beetle age on distance flown per session by M. galloprovincialis. The line corresponds to the model prediction, and the shaded area to the standard error. Distance flown (m) = 10.71 × Age (days) + 443.63.
Summary of results for linear and generalized linear models assessing the effects of weight at emergence (W 0), sex and age on flight parameters, weights and number of eggs.
Significant χ2 or F values are indicated in bold characters (***P < 0.001, **P < 0.01, *P < 0.05). *χ2 for tests on the probability of flying, F values for all other variables.
After removing the effect of weight at emergence, beetles’ weight significantly increased with age, with a rapid initial gain followed by a levelling out (table 1, fig. 2). The sex of the insect had no effect on weight (table 1).
Effect of beetle age on body weight of M. galloprovincialis. The line corresponds to the model prediction, and the shaded area to the standard error. Weight = 2 × 10−2 × Log(Age + 1) + 0.24.
Thorax weight also increased with age, in a similar manner for both the sexes (table 1, fig. 3). By contrast, females gained significantly more abdominal weight with age than did the males (table 1, fig. 3, respectively: t df = 93 = 8.28; P < 0.001 and t df = 93 = 3.95; P < 0.001). The ratio of thorax weight by abdomen weight significantly decreased with age for both males and females, but significantly more so for females (respectively, t df = 93 = −3.96; P < 0.001 and t df = 93 = −8.44; P < 0.001, table 1, fig. 3).
Effect of beetle age on thorax weight W th (black lines), abdomen weight W ab (grey lines) and their ratio, W th /W ab (dash line) in M. galloprovincialis. (a) males, [W th = 1 × 10−2 × Log(Age + 1) + 2 × 10−2; W ab = 1 × 10−2 × Log(Age + 1) + 1 × 10−2; Ratio = −0.1 × Log(Age + 1) + 4 × 10−2] and (b) females, [W th = 1 × 10−2 × Log(Age + 1) + 2 × 10−2; W ab = 2 × 10−2 × Log(Age + 1) + 1 × 10−2; Ratio = −0.2 × Log(Age + 1) + 3 × 10−2].
In the first 10 days after emergence, the females contained almost no eggs (fig. 4). The number of eggs then exponentially increased with age, and was not dependent on the weight at emergence (table 1, fig. 4).
Effect of age on the number of eggs per female. The line corresponds to the model prediction, and the shaded area to the standard error. Number of eggs = e(0.12 × Age − 1.17).
Effect of energy allocation on dispersal capacities
Relative thorax weight had a significant and positive effect on distance flown by M. galloprovincialis, whereas relative abdomen weight did not (table 2). The distance flown on flight mills significantly increased with increasing proportion of energy reserves in the thorax (table 2, fig. 5), but not with proportion of muscle mass. Sex had no effect on distance flown, in any of the models.
Effect of the proportion of thorax reserves (%th res ) on distance flown per session. The line corresponds to the model prediction, and the shaded area to the standard error. Distance flown = 5366 ×%th res + 117.9.
Summary of results for linear models assessing the effects of beetles characteristics on distance flown per session.
The estimate for sex was calculated with ‘female’ as the reference.
Significant F values are shown in bold characters (***P < 0.001, **P < 0.01, *P < 0.05).
Discussion
There is no ideal method for obtaining an accurate estimate of insect flight performance. Nevertheless, even if flight mills do not exactly reproduce the natural conditions for flying (e.g. no need to lift the body weight, absence of wind), they provide standard conditions to precisely compare dispersal performances between individuals of different sex or physiological status (Hughes & Dorn, Reference Hughes and Dorn2002).
We found that almost 50% of sexually immature M. galloprovincialis beetles flew for longer than 30 s on flight mill and this proportion did not change during the maturation phase (30 days), for either sex. Thus, the flight apparatus (pterothorax structure and muscles) was already operational at beetle's emergence. However, the probability of flying was dependent on the weight at emergence. Flight ability is therefore likely to be built upon the capital resources acquired during the larval stages and allocated during metamorphosis. This is consistent with the hypothesis of (Boggs, Reference Boggs1981) that long-lived insects invest in dispersal prior to reproduction. This ensures that their foraging abilities are already operational at emergence, allowing them to find new food sources and to sustain fuel accumulation for reproductive functions, such as egg production. Furthermore, good mobility at emergence allows insects to cope with environmental variability and a lack of predictability concerning the availability of food resources (Jervis et al., Reference Jervis, Ellers and Harvey2008).
There was a significant and linear increase in the distance flown with age, during maturation. Distance flown was also positively correlated with relative weight of the thorax which increased with age. The energy income from adult feeding is thus, probably partly allocated to dispersal (Jervis et al., Reference Jervis, Ellers and Harvey2008). The thorax essentially contains flight muscles and fuel (Chapman et al., Reference Chapman, Simpson and Douglas2013), so the increase in flight capacity associated with an increase in thorax size probably results from a larger musculature and/or the storage of larger amounts of energy to sustain the flying effort (Marden, Reference Marden2000). However in M. galloprovincialis, we observed that the distance flown was significantly correlated with, the proportion of energy reserves in the thorax but not with the proportion of muscle mass. The accumulation of lipids in the thorax is, therefore, probably the main driver of improved flight performance during maturation.
In M. galloprovincialis, females allocated more weight to the abdomen than males during the maturation period, due to oogenesis. However, abdominal weight had no effect on distance flown, indicating that the investment in reproduction in adults was not made at the expense of dispersal capacity. This lack of trade-off between energy allocation to dispersal and reproduction maybe due to the short duration of the experiment, 10 min of flight being not long enough to spend a lot of fuel and trigger allocation processes. However, the absence of such trade-offs seems to be quite common in wing-monomorphic species compared with wing-dimorphic species (Zera & Denno, Reference Zera and Denno1997). For example, Hanski et al. (Reference Hanski, Saastamoinen and Ovaskainen2006) showed that the change in the dispersal capability of Glanville butterflies (Melitaea cinxia Linnaeus) did not come at the expense of fecundity. According to Glazier (Reference Glazier1999) favourable conditions may allow animals to invest energy in several traits resulting in an apparent absence of trade-off. Absence of such trade-off should be a good strategy for ‘income breeders’ when food resources are not limited, which is the case of M. galloprovincialis feeding on the fresh pine shoots during the maturation period. It provides females with good flying ability which is essential for reproduction because: (i) virgin females need to disperse to find nutrients and sustain oogenesis; (ii) females must fly to meet and mate; (iii) gravid females must fly long distances to find a suitable host (decaying tree) for oviposition. When food resources are limited, Pelosse et al. (Reference Pelosse, Jervis, Bernstein and Desouhant2011) showed with the ‘income breeder’ parasitoid Venturia canescens (Gravenhorst) that investing energy in both reproduction and flight resulted in a reduction of adult longevity. Hanski et al. (Reference Hanski, Saastamoinen and Ovaskainen2006) also suggest a physiological trade-off between metabolic performance and lifespan.
In M. galloprovincialis, we found no mature eggs in females in their first 10 days, confirming that this species is synovigenic (Jervis et al., Reference Jervis, Boggs and Ferns2005). Thus most of the energy invested in reproduction comes from adult feeding, indicating that these beetles being income breeders (Jervis et al., Reference Jervis, Ellers and Harvey2008).
Many insects have the potential to use energy from both larval-derived capital and adult-acquired income resources but this requires the ability to manage an array of nutrients, some of which being acquired at different times in the life cycle and involved in competing physiological processes (Jervis et al., Reference Jervis, Ellers and Harvey2008). This flexibility explains that ‘income’ vs. ‘capital’ strategies may occur within the same insect species. In this study, immature beetles of both sexes were able to fly even without feeding, demonstrating that the flight apparatus was already operational at emergence. Thus, M. galloprovincialis can be categorized as a capital disperser in terms of flight ability. However, we found that the beetles flew longer when they had more time to feed on the pine shoots, which also resulted in higher lipid content in the thorax. M. galloprovincialis would thus also be an income disperser in terms of flight performance. Such a complex strategy of energy allocation has also been demonstrated by Casas et al. (Reference Casas, Pincebourde, Mandon, Vannier, Poujol and Giron2005) who showed that Eupelmus vuilletti (Craw) is both a capital breeder for lipids and an income breeder for sugars. In conclusion, this study confirms that both the larval development and the maturation period of young adults are key stages for the reproductive success of M. galloprovincialis. Investing capital resources in the construction of flying capacity at emergence allows insects to forage and invest income resources in reproduction via egg production and long-distance dispersal to find suitable oviposition substrates.
Our results also highlight the importance of maturation feeding for sustaining important dispersal capacities in M. galloprovincialis. These findings call into question the relevance of the clear cut zones (500 m in radius) recommended for slowing the spread of PWN disease in Europe. Indeed, cutting the forest around infested trees will not prevent the beetles from flying (as they can fly without prior feeding), but will instead force them to fly further than they would otherwise do (to find nearby fresh pine shoots), thus disseminating the PWN over a larger area.
The ‘income disperser’ strategy seems to be a good strategy for long-lived adult insects, which emerge in environments with large and predictable feeding resources while oviposition substrates are less abundant or more scattered. On the opposite, the ‘capital disperser’ strategy seems to be more appropriate for short-lived adult insects, often also ‘capital breeder’, living in environments where feeding resources are rare or absent but oviposition substrates are largely abundant. We therefore suggest that the balance of feeding substrate abundance or predictability between adult and larval stages is a main predictor of species adopting income vs. capital strategy for both dispersal and reproduction processes, providing a useful framework for predicting physiological trade-offs and population dynamics in flying insects.
Supplementary material
The supplementary material for this article can be found at http://www.journals.cambridge.org/10.1017/S0007485315000553.
Acknowledgements
We thank Christophe Chipeaux, Fabrice Vetillard and Elorri Segura for their valuable help with the flight mill device and laboratory analyses. We also thank Bastien Castagneyrol for his helpful comments on early versions of the manuscript. G.D. was supported by a grant from the French Ministry of Agriculture, Food and Forestry. This study was financed through a grant from the European Union Seventh Framework Programme FP7 via the Project REPHRAME (grant agreement 265483).The authors declare that they have no conflict of interest.
