Figure 3 - uploaded by Goggy Davidowitz
Content may be subject to copyright.
Artificial advancement of reproduction alters learning across days but not within days. In Experiment 2, treatment butterflies were treated with methoprene (JH mimic) upon emergence as adults, advancing their reproduction, whereas control butterflies were treated only with the solvent acetone. Behavior was measured as the change in host-finding efficiency (the proportion of landings on hosts) within the first day of host search or between the 2 days of host search. Shown are least-square means (and standard errors) from ANOVAs controlling for host color and nonhost complexity.
Source publication
The evolution of learning has long been hypothesized to be limited by fitness trade-offs such as delays in reproduction. We explored the relationship between host learning and reproduction in the cabbage white butterfly, Pieris rapae. The cabbage white female is innately biased to search for common green hosts but can learn to search for rare red h...
Contexts in source publication
Context 1
... Family- level measures of behavior were treated as dependent factors in tests of ovary characteristics at emergence. Body size (hind wing area, see Snell-Rood et al. 2009) of a family was included in each analysis: families exhibited significant genetic variation in body size (e.g., Experiment 1: N 1⁄4 123, F 11,111 1⁄4 5.66, P , 0.0001). ANOVA was used to test for effects of hormone treatment in Experiment 2 and testing environment in Experiment 3. All proportional measures of behavior were arc-sin square-root transformed for statistical analyses, although un- corrected values are presented in figures. Butterflies showed improvements in host-finding efficiency both within and between days, although this improvement was more pronounced in the red host relative to the green host environment (Figure 1; statistics presented in Table 2 of Snell-Rood et al. 2009). Host-finding efficiency was initially higher in the green host environment relative to the red host environment, but by the second day of host searching was comparable between host environments. Host-finding efficiency was consis- tently higher in the simple nonhost environment relative to the complex nonhost environment, but performance improved over time in both environments (Figure 1). The effect of nonhost complexity—a function of both nonhost density and diversity—on host-finding efficiency (Figure 1) was greater than in another experiment where these environments varied in only nonhost diversity (Snell-Rood and Papaj 2009). Butterfly families that showed a greater between-day increase in host-finding efficiency in the red host environment were significantly more likely to emerge with smaller eggs and marginally significantly more likely to emerge with fewer mature eggs (Table 1 and Figure 2). The relationship between measures of ovary maturity and proxies of learning were specific to between-day changes in host-finding efficiency in the red host environment. There were no family-level correlations between egg number or size at emergence and total hosts located, changes in host-finding efficiency within days or changes in host-finding efficiency in the green host environment (Table 1 and Figure 2). In Experiment 2, changes in host-finding efficiency of un- treated (control) butterflies improved over time in all host and nonhost environments (Supplementary Tables 1 and 2; Figure 1), similar to changes in performance observed in Experiment 1. Initially, host-finding efficiency was marginally higher in the green host environment but rapidly improved in the red host environment to a comparable level. Furthermore, host-finding efficiency was higher in the simple relative to the complex nonhost environment, although this difference was most pronounced on the second day of host searching (Supplementary Table 1; Figure 1). The effect of nonhost complexity—a function of both nonhost density and diversity—on host-finding efficiency (Supplementary Table 1; Figure 1) was greater than in another experiment where simple and complex environments varied in only nonhost diversity (Snell-Rood and Papaj 2009). The use of both green and red nonhosts also allowed us to quantify color choice in this experiment (proportion of nonhost landings on green nonhosts). Butterflies chose a higher proportion of green nonhosts than red nonhosts in the green host but not the red host environment (Supplementary Table 1 and Figure 1). At day 2 of adulthood (when butterflies underwent behavioral testing), butterflies treated with methoprene at emergence had significantly more eggs in their ovaries relative to control females that were treated only with the solvent, acetone ( N 1⁄4 13, F 1⁄4 17.6, P 1⁄4 0.002; ANOVA controlling for body size; Supplementary Figure 2). We also compared egg production after behavioral testing, reasoning that if methoprene advanced reproduction, egg production should decrease at an earlier age in methoprene-treated butterflies relative to controls. In an ANOVA controlling for body size (hind wing area), family, host color, and nonhost complexity during learning, control and treatment butterflies did not differ in egg production just after host learning (day 5: N 1⁄4 86, P 1⁄4 0.53, Supplementary Figure 3). However, shortly thereafter, egg production dropped significantly in methoprene-treated butterflies (day 6: N 1⁄4 86, P 1⁄4 0.009; day 7: N 1⁄4 77, P 1⁄4 0.002), before dropping to comparable levels in control butterflies (day 8: N 1⁄4 69, P 1⁄4 0.47). Taken together, these results suggest that our hormone treatment significantly advanced the timing of reproduction. Hormone treatment had significant effects on measures of host learning. Methoprene-treated butterflies were less likely to improve host-finding efficiency across the 2 days of host search relative to controls (Figure 3, Table 2). However, hormone treatment did not affect measures of learning within the first day of host search or changes in color choice (between or within days; Figure 3 and Table 2). We also tested whether hormone treatment had effects on overall measures of fitness (oviposition frequency) during testing. Because methoprene-treated butterflies had more total landings during testing relative to control butterflies (Table 3; mean [standard error]: methoprene 1⁄4 22.0 [1.8], control 1⁄4 37.7 [1.3]), all measures of fitness during testing were corrected for overall landings. Hormone-treated butterflies had a significantly lower ‘‘independent oviposition frequency,’’ a function of ovipositions on host plants separated by landings on other plants (Table 3 and Figure 4), analogous to the fitness measure used in Experiment 1 (see Table 1). However, hormone-treated butterflies were more likely to repeat ovipositions on the same host plant, without landing on different host or nonhost plants in between (Table 3 and Figure 4). When all ovipositions were tallied (independent, repeat, and batch), total oviposition frequency (ovipositions/total landings) was not significantly different between hormone-treated and control butterflies (Table 3). In Experiment 3, we tested for effects of the learning experience itself on lifetime fecundity. Butterflies (control individuals from Experiment 2) were held for 4 days following testing and all eggs laid on both red and green hosts were counted. Butterflies that had searched in the red host environment, relative to those in the green host environment, had significantly lower lifetime fecundity (Table 4 and Figure 5). Butterflies that had searched for hosts in the complex nonhost environment, relative to those that had searched in the simple non-host environment, also had significantly lower lifetime fecundity (Table 4 and Figure 5). There was also a relationship between the amount an individual butterfly learned and lifetime fitness. Controlling for host color, nonhost complexity, body size, and family, there was a negative relationship between an individual’s within-day change in host-finding efficiency and their lifetime fitness (Figure 6; N 1⁄4 25, F 1⁄4 4.50, P 1⁄4 0.05). However, there was no relationship between an individual’s between-day change in host-finding efficiency and their lifetime fitness ( N 1⁄4 16, F 1⁄4 0.05, P 1⁄4 0.82). Learning and cognition have long been hypothesized to be a driving force in the evolution of life-history traits, including the timing of reproduction (Mayr 1974; Johnston 1982; Dukas 1998; Kaplan et al. 2000; Kaplan and Robson 2002; Ricklefs 2004). Our family-level study supports this hypothesis by providing an explicit link between learning performance per se and the timing of reproduction. Butterfly families that emerged with relatively less well-developed ovaries—as measured by both total number of mature eggs and size of oocytes at emergence—showed relatively more improvement in host- finding ability across successive days of host experience (Figure 2 and Table 1). This link was specific to the red host—and not the green host—environment, where color learning has been shown to be used in locating hosts (Supplementary Figure 1; Snell-Rood and Papaj 2009). The family-level relationship between learning ability and ovarian maturation is consistent with the notion of a fitness trade-off between these 2 traits. Johnston (1982) proposed such a trade-off over 25 years ago, offering several nonmutu- ally exclusive explanations by which the evolution of learning might result in delays in reproduction. One explanation focused on allocation of resources in development. Johnston reasoned that learning may incur costs which cause resources to be shunted away from reproduction and into processes related to learning. A growing body of research suggests that learning does indeed have costs (Mery and Kawecki 2003, 2005). In P. rapae , learning entails costs in terms of brain size: Learning ability of a family is associated with greater neural investment at emergence (Snell-Rood et al. 2009). Although direct evidence is lacking, it is conceivable that the additional neural tissue associated with better host learning performance in cabbage whites causes delays in ovarian maturation. Interestingly, we found no correlation between ovary status at emergence and measures of changes in performance within the first day of host searching. Our measures of learning represent a composite of memory processes (short-term, medium- term, and long-term memory; see Margulies et al. 2005), but the between-day changes in performance are at least some- what underlain by long-term memory while within-day changes are not. Thus, our results linking changes in performance across days, but not within days, are reminiscent of the idea that long-term memory, which involves protein synthesis and morphological changes at the synapse level (Lamprecht and LeDoux 2004), is costlier than short-term memory (Mery and Kawecki 2005). In this experiment, the primary fitness trade-off associated with family-level learning ability was a delay in ...
Context 2
... (host color, nonhost complexity, see Experiment 2) but also changes in host-finding efficiency of each individual. Family-level correlations between behavior and life-history traits were used to measure reproductive delay as a cost of learning (Experiment 1). For these analyses, estimates of a family’s learning ability and host search behavior were taken from previous analyses (Snell-Rood et al. 2009). Briefly, mixed-model analyses of variance (ANOVAs) were performed in JMP 7.0, where ‘‘family’’ and ‘‘family by host color’’ were treated as random effects, whereas ‘‘host color’’ and ‘‘nonhost complexity’’ were treated as fixed effects. The least-square means from the family by host color effect were used to estimate a family’s behavior in each host environment. Family- level measures of behavior were treated as dependent factors in tests of ovary characteristics at emergence. Body size (hind wing area, see Snell-Rood et al. 2009) of a family was included in each analysis: families exhibited significant genetic variation in body size (e.g., Experiment 1: N 1⁄4 123, F 11,111 1⁄4 5.66, P , 0.0001). ANOVA was used to test for effects of hormone treatment in Experiment 2 and testing environment in Experiment 3. All proportional measures of behavior were arc-sin square-root transformed for statistical analyses, although un- corrected values are presented in figures. Butterflies showed improvements in host-finding efficiency both within and between days, although this improvement was more pronounced in the red host relative to the green host environment (Figure 1; statistics presented in Table 2 of Snell-Rood et al. 2009). Host-finding efficiency was initially higher in the green host environment relative to the red host environment, but by the second day of host searching was comparable between host environments. Host-finding efficiency was consis- tently higher in the simple nonhost environment relative to the complex nonhost environment, but performance improved over time in both environments (Figure 1). The effect of nonhost complexity—a function of both nonhost density and diversity—on host-finding efficiency (Figure 1) was greater than in another experiment where these environments varied in only nonhost diversity (Snell-Rood and Papaj 2009). Butterfly families that showed a greater between-day increase in host-finding efficiency in the red host environment were significantly more likely to emerge with smaller eggs and marginally significantly more likely to emerge with fewer mature eggs (Table 1 and Figure 2). The relationship between measures of ovary maturity and proxies of learning were specific to between-day changes in host-finding efficiency in the red host environment. There were no family-level correlations between egg number or size at emergence and total hosts located, changes in host-finding efficiency within days or changes in host-finding efficiency in the green host environment (Table 1 and Figure 2). In Experiment 2, changes in host-finding efficiency of un- treated (control) butterflies improved over time in all host and nonhost environments (Supplementary Tables 1 and 2; Figure 1), similar to changes in performance observed in Experiment 1. Initially, host-finding efficiency was marginally higher in the green host environment but rapidly improved in the red host environment to a comparable level. Furthermore, host-finding efficiency was higher in the simple relative to the complex nonhost environment, although this difference was most pronounced on the second day of host searching (Supplementary Table 1; Figure 1). The effect of nonhost complexity—a function of both nonhost density and diversity—on host-finding efficiency (Supplementary Table 1; Figure 1) was greater than in another experiment where simple and complex environments varied in only nonhost diversity (Snell-Rood and Papaj 2009). The use of both green and red nonhosts also allowed us to quantify color choice in this experiment (proportion of nonhost landings on green nonhosts). Butterflies chose a higher proportion of green nonhosts than red nonhosts in the green host but not the red host environment (Supplementary Table 1 and Figure 1). At day 2 of adulthood (when butterflies underwent behavioral testing), butterflies treated with methoprene at emergence had significantly more eggs in their ovaries relative to control females that were treated only with the solvent, acetone ( N 1⁄4 13, F 1⁄4 17.6, P 1⁄4 0.002; ANOVA controlling for body size; Supplementary Figure 2). We also compared egg production after behavioral testing, reasoning that if methoprene advanced reproduction, egg production should decrease at an earlier age in methoprene-treated butterflies relative to controls. In an ANOVA controlling for body size (hind wing area), family, host color, and nonhost complexity during learning, control and treatment butterflies did not differ in egg production just after host learning (day 5: N 1⁄4 86, P 1⁄4 0.53, Supplementary Figure 3). However, shortly thereafter, egg production dropped significantly in methoprene-treated butterflies (day 6: N 1⁄4 86, P 1⁄4 0.009; day 7: N 1⁄4 77, P 1⁄4 0.002), before dropping to comparable levels in control butterflies (day 8: N 1⁄4 69, P 1⁄4 0.47). Taken together, these results suggest that our hormone treatment significantly advanced the timing of reproduction. Hormone treatment had significant effects on measures of host learning. Methoprene-treated butterflies were less likely to improve host-finding efficiency across the 2 days of host search relative to controls (Figure 3, Table 2). However, hormone treatment did not affect measures of learning within the first day of host search or changes in color choice (between or within days; Figure 3 and Table 2). We also tested whether hormone treatment had effects on overall measures of fitness (oviposition frequency) during testing. Because methoprene-treated butterflies had more total landings during testing relative to control butterflies (Table 3; mean [standard error]: methoprene 1⁄4 22.0 [1.8], control 1⁄4 37.7 [1.3]), all measures of fitness during testing were corrected for overall landings. Hormone-treated butterflies had a significantly lower ‘‘independent oviposition frequency,’’ a function of ovipositions on host plants separated by landings on other plants (Table 3 and Figure 4), analogous to the fitness measure used in Experiment 1 (see Table 1). However, hormone-treated butterflies were more likely to repeat ovipositions on the same host plant, without landing on different host or nonhost plants in between (Table 3 and Figure 4). When all ovipositions were tallied (independent, repeat, and batch), total oviposition frequency (ovipositions/total landings) was not significantly different between hormone-treated and control butterflies (Table 3). In Experiment 3, we tested for effects of the learning experience itself on lifetime fecundity. Butterflies (control individuals from Experiment 2) were held for 4 days following testing and all eggs laid on both red and green hosts were counted. Butterflies that had searched in the red host environment, relative to those in the green host environment, had significantly lower lifetime fecundity (Table 4 and Figure 5). Butterflies that had searched for hosts in the complex nonhost environment, relative to those that had searched in the simple non-host environment, also had significantly lower lifetime fecundity (Table 4 and Figure 5). There was also a relationship between the amount an individual butterfly learned and lifetime fitness. Controlling for host color, nonhost complexity, body size, and family, there was a negative relationship between an individual’s within-day change in host-finding efficiency and their lifetime fitness (Figure 6; N 1⁄4 25, F 1⁄4 4.50, P 1⁄4 0.05). However, there was no relationship between an individual’s between-day change in host-finding efficiency and their lifetime fitness ( N 1⁄4 16, F 1⁄4 0.05, P 1⁄4 0.82). Learning and cognition have long been hypothesized to be a driving force in the evolution of life-history traits, including the timing of reproduction (Mayr 1974; Johnston 1982; Dukas 1998; Kaplan et al. 2000; Kaplan and Robson 2002; Ricklefs 2004). Our family-level study supports this hypothesis by providing an explicit link between learning performance per se and the timing of reproduction. Butterfly families that emerged with relatively less well-developed ovaries—as measured by both total number of mature eggs and size of oocytes at emergence—showed relatively more improvement in host- finding ability across successive days of host experience (Figure 2 and Table 1). This link was specific to the red host—and not the green host—environment, where color learning has been shown to be used in locating hosts (Supplementary Figure 1; Snell-Rood and Papaj 2009). The family-level relationship between learning ability and ovarian maturation is consistent with the notion of a fitness trade-off between these 2 traits. Johnston (1982) proposed such a trade-off over 25 years ago, offering several nonmutu- ally exclusive explanations by which the evolution of learning might result in delays in reproduction. One explanation focused on allocation of resources in development. Johnston reasoned that learning may incur costs which cause resources to be shunted away from reproduction and into processes related to learning. A growing body of research suggests that learning does indeed have costs (Mery and Kawecki 2003, 2005). In P. rapae , learning entails costs in terms of brain size: Learning ability of a family is associated with greater neural investment at emergence (Snell-Rood et al. 2009). Although direct evidence is lacking, it is conceivable that the additional neural tissue associated with better host learning performance in cabbage whites causes delays in ovarian maturation. Interestingly, we found no correlation between ovary status at emergence and ...
Context 3
... within the first day of host searching. Our measures of learning represent a composite of memory processes (short-term, medium- term, and long-term memory; see Margulies et al. 2005), but the between-day changes in performance are at least some- what underlain by long-term memory while within-day changes are not. Thus, our results linking changes in performance across days, but not within days, are reminiscent of the idea that long-term memory, which involves protein synthesis and morphological changes at the synapse level (Lamprecht and LeDoux 2004), is costlier than short-term memory (Mery and Kawecki 2005). In this experiment, the primary fitness trade-off associated with family-level learning ability was a delay in reproduction. Just how costly is a delay in reproduction? Because lepidop- teran species mature a large proportion of their eggs during adulthood—many using adult-acquired resources—it is un- clear just how long into adulthood our observed reproductive delay is relevant (O’Brien et al. 2002; Jervis et al. 2005). Regardless, in this butterfly system, where adult life span in the field is typically less than 14 days (Suzuki 1978), delaying reproduction by even a day or 2 likely represents a significant cost. Individuals that delay reproduction should have a lower chance of surviving to maximal reproduction. Additionally, delaying reproduction may lead to missed opportunities due to egg limitation upon emergence (Rosenheim et al. 2000; Jervis et al. 2001)—for instance, butterflies may encounter host plants during their first day of adulthood, but ovary im- maturity may preclude the use of these plants. Although reproductive timing was correlated with changes in performance across days, we found no correlation between ovary maturity at emergence and our measure of overall fitness (total host landings across both days of host searching, corrected for total landings, Table 1). This was a surprising result given that a separate experiment showed that the change in host-finding efficiency between days of host searching was a significant contributor to overall fitness in the red host environment (Snell-Rood and Papaj 2009). There were 3 main differences in the host search environment between these 2 experiments, including 1) the inclusion of both red and green nonhosts versus only green nonhosts, 2) the ratio of hosts to nonhosts, and 3) the degree of color difference between green and red nonhosts. Taken together, these differences suggest that host searching in the present experiment (Experiment 1) occurred under less challenging search conditions than those in Snell-Rood and Papaj (2009) such that the relative benefits of learning (higher fitness in the red host environment) were less pronounced. Although future work will be needed to support this assertion, this discussion under- scores the importance of considering the host environment(s) in which selection is occurring. The costs of delayed reproduction may sometimes be offset by the benefits of learning (e.g., in the red host environment), whereas in other conditions (e.g., green host environments), the global cost of reproductive delay is experienced, but with few offsetting benefits of learning (Figure 1; Snell-Rood and Papaj 2009). Thus, the competitive outcome between learning and nonlearning genotypes will depend on a range of factors such as host distribution over time and space and butterfly life span. We used hormonal manipulations to artificially advance reproduction and provide a separate test of the link between reproductive timing and learning ability. Previous studies on butterflies, and in particular those in the genus Pieris , suggested that topical treatment with methoprene, a JH analog, could be used to advance reproduction in P. rapae (Karlinsky 1963; Benz 1970, 1972; Herman 1973, 1975; Herman and Bennett 1975; Herman et al. 1981). We found that methoprene application did indeed advance egg maturation. Individuals treated with hormone at emergence had, on average, twice the egg load of control females 2 days later (Supplementary Figure 2). Furthermore, when hormone-treated and control butterflies were held for 4 days after behavioral testing, egg production of hormone-treated butterflies dropped off earlier than that of control butterflies (Supplementary Figure 3). These results suggest that our level of hormone application significantly advanced the timing of egg development in P. rapae such that egg production peaked earlier in adulthood in treatment relative to control individuals. Because these manipulations occurred prior to, and independent of the host experience, we feel it simulates a manipulation of an environment-independent, constitutive (or global) reproductive trade-off. Hormone application was accompanied by significant changes in learning performance. Specifically, hormone-treated butterflies, relative to controls, were less likely to improve host-finding efficiency across the 2 days of host searching (Figure 3 and Table 2). However, there was no difference between control and treatment butterflies in changes in host-finding efficiency within the first day of host searching (Table 2). The specificity of this effect to between-day changes in performance paralleled the result of the family-level correlations, where ovary maturity at emergence was related to between-day but not within-day changes in performance (Table 1). The parallel result linking between-day changes in behavior and reproductive delays emerged between experiments despite the differences in host and nonhost species used in greenhouse assays (see MATERIALS AND METHODS), suggesting that this result applies to learning more generally, not certain plant species. Treatment and control butterflies also differed in their patterns of oviposition across different hosts. Over all landings, control butterflies had significantly more independent host ovipositions relative to treatment butterflies (Table 3 and Figure 4). In contrast, treatment individuals had significantly more repeat ovipositions—an independent oviposition event followed by leaving the host and then immediately returning to (and ovipositing on) the same host without landing on another plant in between (Table 3 and Figure 4). Given that oviposition on separate plants should decrease larval compe- tition and contribute to the ‘‘risk-spreading’’ strategy of Pieris (Root and Kareiva 1984; Kivela and Valimaki 2008), these ‘‘repeat ovipositions’’ may be costly. However, overall, total number of ovipositions during testing (corrected for total landings made) did not differ between control and treatment butterflies (Table 3). Taken together, these behavioral observations suggest that artificial advancement of reproduction affects learning performance directly. The comparable total fitness between methoprene-treated and control butterflies suggests the 2 groups had comparable host-seeking motivation. However, methoprene-treated individuals had a selective reduction in between-day learning ability. This reduction may have re- duced independent oviposition frequency, which butterflies compensated for by repeating ovipositions on a given host plant. The fact that these changes in fitness are more pronounced in the red host environment (Figures 3 and 4), where learning is more pronounced (Figure 1; Snell-Rood and Papaj 2009), further implicates the effect of hormone treatment on learning. These results are consistent with the idea that advancement of reproduction trades off against investment in processes or structures necessary for learning, such as neural tissue. Indeed, in a separate analysis, 2 days after emergence (at the time when host searching began in our experiments), hormone-treated animals, relative to controls, tended to have smaller absolute ( N 1⁄4 7 individuals; F 1,5 1⁄4 17.6, P 1⁄4 0.008) and relative mushroom body volume ( N 1⁄4 6 individuals; F 1,3 1⁄4 7.95, P 1⁄4 0.06; Snell-Rood EC and Gronenberg W, unpublished data; see methods in Snell-Rood et al. 2009). Future work on more individuals will be required to validate this observation. Our results suggest that, due to energetic tradeoffs, hormone application has indirect (negative) effects on learning stemming from its direct (positive) effects on reproduction. However, it is important to note that JH regulates many aspects of insect physiology (Wyatt and Davey 1996; Flatt et al. 2005). It is possible that hormonal application also has direct effects on learning and even that our results reflect the JH-mediated coordination of a suite of traits (Flatt et al. 2005). JH has known effects on neural development and learning in other systems such as crickets, where it stimulates neural growth (Cayre et al. 1994). In adult honeybees, increased JH titers are associated with the development of foraging and learning in adult workers (Robinson and Vargo 1997), although in this case, the association was not due to direct effects of JH on neural growth (Fahrbach et al. 1998). Our results echo a re- current theme across these studies that JH has diverse roles in regulating life history and behavior in insects (Nijhout 1994). Indeed, JH affects short-term memory and not long-term memory in honeybees (Maleszka and Helliwell 2001), which differs from our results. Further studies integrating hormone titer measurements and manipulations (Zera et al. 2007) may have implications for understanding the mechanistic basis of the observed genetic variation in learning and reproduction. In the meantime, our results must be interpreted cautiously, knowing that hormonal manipulations have complex effects on suites of traits. Our results suggest that learning ability comes with constitutive fitness trade-offs in the form of reproductive delays. These trade-offs are expressed regardless of the environment in which learning occurs and thus represent global costs of learning, which can explain the maintenance of genetic variation in learning and ...
Context 4
... ‘‘nonhost complexity’’ were treated as fixed effects. The least-square means from the family by host color effect were used to estimate a family’s behavior in each host environment. Family- level measures of behavior were treated as dependent factors in tests of ovary characteristics at emergence. Body size (hind wing area, see Snell-Rood et al. 2009) of a family was included in each analysis: families exhibited significant genetic variation in body size (e.g., Experiment 1: N 1⁄4 123, F 11,111 1⁄4 5.66, P , 0.0001). ANOVA was used to test for effects of hormone treatment in Experiment 2 and testing environment in Experiment 3. All proportional measures of behavior were arc-sin square-root transformed for statistical analyses, although un- corrected values are presented in figures. Butterflies showed improvements in host-finding efficiency both within and between days, although this improvement was more pronounced in the red host relative to the green host environment (Figure 1; statistics presented in Table 2 of Snell-Rood et al. 2009). Host-finding efficiency was initially higher in the green host environment relative to the red host environment, but by the second day of host searching was comparable between host environments. Host-finding efficiency was consis- tently higher in the simple nonhost environment relative to the complex nonhost environment, but performance improved over time in both environments (Figure 1). The effect of nonhost complexity—a function of both nonhost density and diversity—on host-finding efficiency (Figure 1) was greater than in another experiment where these environments varied in only nonhost diversity (Snell-Rood and Papaj 2009). Butterfly families that showed a greater between-day increase in host-finding efficiency in the red host environment were significantly more likely to emerge with smaller eggs and marginally significantly more likely to emerge with fewer mature eggs (Table 1 and Figure 2). The relationship between measures of ovary maturity and proxies of learning were specific to between-day changes in host-finding efficiency in the red host environment. There were no family-level correlations between egg number or size at emergence and total hosts located, changes in host-finding efficiency within days or changes in host-finding efficiency in the green host environment (Table 1 and Figure 2). In Experiment 2, changes in host-finding efficiency of un- treated (control) butterflies improved over time in all host and nonhost environments (Supplementary Tables 1 and 2; Figure 1), similar to changes in performance observed in Experiment 1. Initially, host-finding efficiency was marginally higher in the green host environment but rapidly improved in the red host environment to a comparable level. Furthermore, host-finding efficiency was higher in the simple relative to the complex nonhost environment, although this difference was most pronounced on the second day of host searching (Supplementary Table 1; Figure 1). The effect of nonhost complexity—a function of both nonhost density and diversity—on host-finding efficiency (Supplementary Table 1; Figure 1) was greater than in another experiment where simple and complex environments varied in only nonhost diversity (Snell-Rood and Papaj 2009). The use of both green and red nonhosts also allowed us to quantify color choice in this experiment (proportion of nonhost landings on green nonhosts). Butterflies chose a higher proportion of green nonhosts than red nonhosts in the green host but not the red host environment (Supplementary Table 1 and Figure 1). At day 2 of adulthood (when butterflies underwent behavioral testing), butterflies treated with methoprene at emergence had significantly more eggs in their ovaries relative to control females that were treated only with the solvent, acetone ( N 1⁄4 13, F 1⁄4 17.6, P 1⁄4 0.002; ANOVA controlling for body size; Supplementary Figure 2). We also compared egg production after behavioral testing, reasoning that if methoprene advanced reproduction, egg production should decrease at an earlier age in methoprene-treated butterflies relative to controls. In an ANOVA controlling for body size (hind wing area), family, host color, and nonhost complexity during learning, control and treatment butterflies did not differ in egg production just after host learning (day 5: N 1⁄4 86, P 1⁄4 0.53, Supplementary Figure 3). However, shortly thereafter, egg production dropped significantly in methoprene-treated butterflies (day 6: N 1⁄4 86, P 1⁄4 0.009; day 7: N 1⁄4 77, P 1⁄4 0.002), before dropping to comparable levels in control butterflies (day 8: N 1⁄4 69, P 1⁄4 0.47). Taken together, these results suggest that our hormone treatment significantly advanced the timing of reproduction. Hormone treatment had significant effects on measures of host learning. Methoprene-treated butterflies were less likely to improve host-finding efficiency across the 2 days of host search relative to controls (Figure 3, Table 2). However, hormone treatment did not affect measures of learning within the first day of host search or changes in color choice (between or within days; Figure 3 and Table 2). We also tested whether hormone treatment had effects on overall measures of fitness (oviposition frequency) during testing. Because methoprene-treated butterflies had more total landings during testing relative to control butterflies (Table 3; mean [standard error]: methoprene 1⁄4 22.0 [1.8], control 1⁄4 37.7 [1.3]), all measures of fitness during testing were corrected for overall landings. Hormone-treated butterflies had a significantly lower ‘‘independent oviposition frequency,’’ a function of ovipositions on host plants separated by landings on other plants (Table 3 and Figure 4), analogous to the fitness measure used in Experiment 1 (see Table 1). However, hormone-treated butterflies were more likely to repeat ovipositions on the same host plant, without landing on different host or nonhost plants in between (Table 3 and Figure 4). When all ovipositions were tallied (independent, repeat, and batch), total oviposition frequency (ovipositions/total landings) was not significantly different between hormone-treated and control butterflies (Table 3). In Experiment 3, we tested for effects of the learning experience itself on lifetime fecundity. Butterflies (control individuals from Experiment 2) were held for 4 days following testing and all eggs laid on both red and green hosts were counted. Butterflies that had searched in the red host environment, relative to those in the green host environment, had significantly lower lifetime fecundity (Table 4 and Figure 5). Butterflies that had searched for hosts in the complex nonhost environment, relative to those that had searched in the simple non-host environment, also had significantly lower lifetime fecundity (Table 4 and Figure 5). There was also a relationship between the amount an individual butterfly learned and lifetime fitness. Controlling for host color, nonhost complexity, body size, and family, there was a negative relationship between an individual’s within-day change in host-finding efficiency and their lifetime fitness (Figure 6; N 1⁄4 25, F 1⁄4 4.50, P 1⁄4 0.05). However, there was no relationship between an individual’s between-day change in host-finding efficiency and their lifetime fitness ( N 1⁄4 16, F 1⁄4 0.05, P 1⁄4 0.82). Learning and cognition have long been hypothesized to be a driving force in the evolution of life-history traits, including the timing of reproduction (Mayr 1974; Johnston 1982; Dukas 1998; Kaplan et al. 2000; Kaplan and Robson 2002; Ricklefs 2004). Our family-level study supports this hypothesis by providing an explicit link between learning performance per se and the timing of reproduction. Butterfly families that emerged with relatively less well-developed ovaries—as measured by both total number of mature eggs and size of oocytes at emergence—showed relatively more improvement in host- finding ability across successive days of host experience (Figure 2 and Table 1). This link was specific to the red host—and not the green host—environment, where color learning has been shown to be used in locating hosts (Supplementary Figure 1; Snell-Rood and Papaj 2009). The family-level relationship between learning ability and ovarian maturation is consistent with the notion of a fitness trade-off between these 2 traits. Johnston (1982) proposed such a trade-off over 25 years ago, offering several nonmutu- ally exclusive explanations by which the evolution of learning might result in delays in reproduction. One explanation focused on allocation of resources in development. Johnston reasoned that learning may incur costs which cause resources to be shunted away from reproduction and into processes related to learning. A growing body of research suggests that learning does indeed have costs (Mery and Kawecki 2003, 2005). In P. rapae , learning entails costs in terms of brain size: Learning ability of a family is associated with greater neural investment at emergence (Snell-Rood et al. 2009). Although direct evidence is lacking, it is conceivable that the additional neural tissue associated with better host learning performance in cabbage whites causes delays in ovarian maturation. Interestingly, we found no correlation between ovary status at emergence and measures of changes in performance within the first day of host searching. Our measures of learning represent a composite of memory processes (short-term, medium- term, and long-term memory; see Margulies et al. 2005), but the between-day changes in performance are at least some- what underlain by long-term memory while within-day changes are not. Thus, our results linking changes in performance across days, but not within days, are reminiscent of the idea that long-term memory, which involves protein synthesis and morphological changes at the synapse level (Lamprecht and LeDoux 2004), is ...
Context 5
... which cause resources to be shunted away from reproduction and into processes related to learning. A growing body of research suggests that learning does indeed have costs (Mery and Kawecki 2003, 2005). In P. rapae , learning entails costs in terms of brain size: Learning ability of a family is associated with greater neural investment at emergence (Snell-Rood et al. 2009). Although direct evidence is lacking, it is conceivable that the additional neural tissue associated with better host learning performance in cabbage whites causes delays in ovarian maturation. Interestingly, we found no correlation between ovary status at emergence and measures of changes in performance within the first day of host searching. Our measures of learning represent a composite of memory processes (short-term, medium- term, and long-term memory; see Margulies et al. 2005), but the between-day changes in performance are at least some- what underlain by long-term memory while within-day changes are not. Thus, our results linking changes in performance across days, but not within days, are reminiscent of the idea that long-term memory, which involves protein synthesis and morphological changes at the synapse level (Lamprecht and LeDoux 2004), is costlier than short-term memory (Mery and Kawecki 2005). In this experiment, the primary fitness trade-off associated with family-level learning ability was a delay in reproduction. Just how costly is a delay in reproduction? Because lepidop- teran species mature a large proportion of their eggs during adulthood—many using adult-acquired resources—it is un- clear just how long into adulthood our observed reproductive delay is relevant (O’Brien et al. 2002; Jervis et al. 2005). Regardless, in this butterfly system, where adult life span in the field is typically less than 14 days (Suzuki 1978), delaying reproduction by even a day or 2 likely represents a significant cost. Individuals that delay reproduction should have a lower chance of surviving to maximal reproduction. Additionally, delaying reproduction may lead to missed opportunities due to egg limitation upon emergence (Rosenheim et al. 2000; Jervis et al. 2001)—for instance, butterflies may encounter host plants during their first day of adulthood, but ovary im- maturity may preclude the use of these plants. Although reproductive timing was correlated with changes in performance across days, we found no correlation between ovary maturity at emergence and our measure of overall fitness (total host landings across both days of host searching, corrected for total landings, Table 1). This was a surprising result given that a separate experiment showed that the change in host-finding efficiency between days of host searching was a significant contributor to overall fitness in the red host environment (Snell-Rood and Papaj 2009). There were 3 main differences in the host search environment between these 2 experiments, including 1) the inclusion of both red and green nonhosts versus only green nonhosts, 2) the ratio of hosts to nonhosts, and 3) the degree of color difference between green and red nonhosts. Taken together, these differences suggest that host searching in the present experiment (Experiment 1) occurred under less challenging search conditions than those in Snell-Rood and Papaj (2009) such that the relative benefits of learning (higher fitness in the red host environment) were less pronounced. Although future work will be needed to support this assertion, this discussion under- scores the importance of considering the host environment(s) in which selection is occurring. The costs of delayed reproduction may sometimes be offset by the benefits of learning (e.g., in the red host environment), whereas in other conditions (e.g., green host environments), the global cost of reproductive delay is experienced, but with few offsetting benefits of learning (Figure 1; Snell-Rood and Papaj 2009). Thus, the competitive outcome between learning and nonlearning genotypes will depend on a range of factors such as host distribution over time and space and butterfly life span. We used hormonal manipulations to artificially advance reproduction and provide a separate test of the link between reproductive timing and learning ability. Previous studies on butterflies, and in particular those in the genus Pieris , suggested that topical treatment with methoprene, a JH analog, could be used to advance reproduction in P. rapae (Karlinsky 1963; Benz 1970, 1972; Herman 1973, 1975; Herman and Bennett 1975; Herman et al. 1981). We found that methoprene application did indeed advance egg maturation. Individuals treated with hormone at emergence had, on average, twice the egg load of control females 2 days later (Supplementary Figure 2). Furthermore, when hormone-treated and control butterflies were held for 4 days after behavioral testing, egg production of hormone-treated butterflies dropped off earlier than that of control butterflies (Supplementary Figure 3). These results suggest that our level of hormone application significantly advanced the timing of egg development in P. rapae such that egg production peaked earlier in adulthood in treatment relative to control individuals. Because these manipulations occurred prior to, and independent of the host experience, we feel it simulates a manipulation of an environment-independent, constitutive (or global) reproductive trade-off. Hormone application was accompanied by significant changes in learning performance. Specifically, hormone-treated butterflies, relative to controls, were less likely to improve host-finding efficiency across the 2 days of host searching (Figure 3 and Table 2). However, there was no difference between control and treatment butterflies in changes in host-finding efficiency within the first day of host searching (Table 2). The specificity of this effect to between-day changes in performance paralleled the result of the family-level correlations, where ovary maturity at emergence was related to between-day but not within-day changes in performance (Table 1). The parallel result linking between-day changes in behavior and reproductive delays emerged between experiments despite the differences in host and nonhost species used in greenhouse assays (see MATERIALS AND METHODS), suggesting that this result applies to learning more generally, not certain plant species. Treatment and control butterflies also differed in their patterns of oviposition across different hosts. Over all landings, control butterflies had significantly more independent host ovipositions relative to treatment butterflies (Table 3 and Figure 4). In contrast, treatment individuals had significantly more repeat ovipositions—an independent oviposition event followed by leaving the host and then immediately returning to (and ovipositing on) the same host without landing on another plant in between (Table 3 and Figure 4). Given that oviposition on separate plants should decrease larval compe- tition and contribute to the ‘‘risk-spreading’’ strategy of Pieris (Root and Kareiva 1984; Kivela and Valimaki 2008), these ‘‘repeat ovipositions’’ may be costly. However, overall, total number of ovipositions during testing (corrected for total landings made) did not differ between control and treatment butterflies (Table 3). Taken together, these behavioral observations suggest that artificial advancement of reproduction affects learning performance directly. The comparable total fitness between methoprene-treated and control butterflies suggests the 2 groups had comparable host-seeking motivation. However, methoprene-treated individuals had a selective reduction in between-day learning ability. This reduction may have re- duced independent oviposition frequency, which butterflies compensated for by repeating ovipositions on a given host plant. The fact that these changes in fitness are more pronounced in the red host environment (Figures 3 and 4), where learning is more pronounced (Figure 1; Snell-Rood and Papaj 2009), further implicates the effect of hormone treatment on learning. These results are consistent with the idea that advancement of reproduction trades off against investment in processes or structures necessary for learning, such as neural tissue. Indeed, in a separate analysis, 2 days after emergence (at the time when host searching began in our experiments), hormone-treated animals, relative to controls, tended to have smaller absolute ( N 1⁄4 7 individuals; F 1,5 1⁄4 17.6, P 1⁄4 0.008) and relative mushroom body volume ( N 1⁄4 6 individuals; F 1,3 1⁄4 7.95, P 1⁄4 0.06; Snell-Rood EC and Gronenberg W, unpublished data; see methods in Snell-Rood et al. 2009). Future work on more individuals will be required to validate this observation. Our results suggest that, due to energetic tradeoffs, hormone application has indirect (negative) effects on learning stemming from its direct (positive) effects on reproduction. However, it is important to note that JH regulates many aspects of insect physiology (Wyatt and Davey 1996; Flatt et al. 2005). It is possible that hormonal application also has direct effects on learning and even that our results reflect the JH-mediated coordination of a suite of traits (Flatt et al. 2005). JH has known effects on neural development and learning in other systems such as crickets, where it stimulates neural growth (Cayre et al. 1994). In adult honeybees, increased JH titers are associated with the development of foraging and learning in adult workers (Robinson and Vargo 1997), although in this case, the association was not due to direct effects of JH on neural growth (Fahrbach et al. 1998). Our results echo a re- current theme across these studies that JH has diverse roles in regulating life history and behavior in insects (Nijhout 1994). Indeed, JH affects short-term memory and not long-term memory in honeybees (Maleszka and Helliwell 2001), which differs from our results. Further ...
Citations
... new hosts or non-host environment). Females foraging in complex environments showed higher flight muscle development and, after gaining experience in this environment, increased offspring investment (higher egg size and lipid reserves) compared to females foraging in the control environment (Snell-Rood et al. 2011). The observed decrease in lifetime fecundity associated with an increase in cognition and learning capacities will lead to complex evolutionary life-history trade-offs in thermally costly environments (Dunlap et al. 2009;Snell-Rood et al. 2013). ...
Climate change alters many environmental parameters with strong consequences for ecological interactions, from species interactions to community dynamics. Temperature is crucial in determining ecosystem dynamics, especially for those involving ectothermic species such as plants or insects. Phenotypic plasticity, the capacity of one genotype to produce different phenotypes in response to environmental conditions, is a common mechanism by which individuals adapt to changing environments and is observed in multiple traits. The capacity of genotypes to adapt to novel temperature conditions plays a crucial role in structuring ecosystem dynamics and species persistence in adverse conditions. It is well recognised that temperature in natural ecosystems fluctuates over multiple time scales (e.g., hour, day, season, year). These fluctuations can follow predictable patterns or be unpredictable, with different consequences for phenotypic plasticity and ecosystem dynamics. Among trophic interactions, host–parasitoid interactions represent a special case because of the intimate symbiosis of the parasitoid larvae with their host. Understanding how and to what extent phenotypic plasticity structures species’ ecological niches is of utmost importance in the context of rapid climate change. With a particular focus on host–parasitoid interactions, this review discusses the literature on the role of phenotypic plasticity in fluctuating environments, highlighting the role of temporal dynamics. While we discuss literature on phenotypic plasticity at large, this review emphasises the fundamental effects of extreme temperatures in driving biochemical rates underlying phenotypic plasticity.
... Hence, in older (dominant) females the cost of maintaining a high reproductive output, even in the presence of competitors, might be traded-off against the maintenance of the energetically costly nervous system [13]. For example, experiments in the fruit fly and the cabbage white butterfly (Pieris rapae) have revealed a trade-off between learning performance and competitive ability [77] or female fecundity [78], respectively. In line with this explanation, when analysing long-term reproductive success in dominant babblers, we found that higher cognitive performance was associated with a lower number of fledglings produced per year. ...
Identifying the causes and fitness consequences of intraspecific variation in cognitive performance is fundamental to understand how cognition evolves. Selection may act on different cognitive traits separately or jointly as part of the general cognitive performance (GCP) of the individual. To date, few studies have examined simultaneously whether individual cognitive performance covaries across different cognitive tasks, the relative importance of individual and social attributes in determining cognitive variation, and its fitness consequences in the wild. Here, we tested 38 wild southern pied babblers (Turdoides bicolor) on a cognitive test battery targeting associative learning, reversal learning and inhibitory control. We found that a single factor explained 59.5% of the variation in individual cognitive performance across tasks, suggestive of a general cognitive factor. GCP varied by age and sex; declining with age in females but not males. Older females also tended to produce a higher average number of fledglings per year compared to younger females. Analysing over 10 years of breeding data, we found that individuals with lower general cognitive performance produced more fledglings per year. Collectively, our findings support the existence of a trade-off between cognitive performance and reproductive success in a wild bird.
... However, there are also costs associated with learning, potentially leading to trade-offs between learning and other traits, which may affect the net fitness benefits of learning. For example, faster learners were less competitive as larvae in Drosophila melanogaster (Mery & Kawecki, 2003), and more efficient learners were less fecund in cabbage white butterflies, Pieris rapae (Snell-Rood et al., 2011). Furthermore, performance on different learning tasks may be subject to trade-offs: individuals that make fast and strong initial associations may have more difficulty in reversal learning, which involves additional processes like proactive interference and inhibition. ...
Growing evidence suggests that individual variation in learning is ubiquitous, but why this is the case and what the consequences are is still a subject of much debate and research. One key set of explanations for variation in learning behaviour is that it relates to variation in animal personality traits. If personality traits affect how an individual interacts with its environment or processes information, this could directly affect performance in learning tasks. While this idea is generally well supported, there are inconsistent results on the relationships between specific personality traits and performance on different learning tasks, highlighting the need to measure multiple personality traits and to quantify different aspects of learning in the same individuals. We examined the relationship between three putative personality traits – aggression, latency to emerge from a shelter and time to contact a novel object – and learning speed in both initial and reversal olfactory learning in the house cricket, Acheta domesticus. Crickets were assayed for each personality trait, then tested for their speed to associate an odour with a water reward. Both aggression and latency to emerge were significantly repeatable, but only latency to emerge was related to learning speed, with individuals that took longer to emerge from the shelter requiring fewer trials to reach the learning criterion for both the initial and reversal learning experiments. We also identified sex differences in learning speed in the different experiments. Thus, our results provide some support for a relationship between personality and learning in an invertebrate.
... Second, empirical evidence suggests that the differential allocation of metabolic resources due to learning ultimately results in life-history (Burger, Kolss, Pont, & Kawecki, 2008;Snell-Rood, Davidowitz, & Papaj, 2011) and fitness trade-offs (Mery & Kawecki, 2003; see also Zwoinska, Lind, Cortazar-Chinarro, Ramsden, & Maklakov, 2016). Cognitive abilities, therefore, appear to evolve at some costs (Buchanan et al., 2013) i.e. the metabolic costs of building the neural substrates governing them (DeVoogd, 2004;Laughlin, de Ruyter van Steveninck, & Anderson, 1998;Lefebvre & Sol, 2008;Snell-Rood, Papaj, & Gronenberg, 2009) and the energetic and time costs related to the cognitive processes themselves, for example requiring sampling 85 Learned components of courtship of the environment (Mery & Kawecki, 2004). ...
Research into learning of courtship behavior remains largely confined to birdsong and vocal learning studies. Yet, visually communicated aspects of courtship displays are widespread and prominent, and also deserve consideration. Postural displays, choreographies, and construction of display arenas are all visual signal components mediated by motor activity of the displayer. The goal of this review is to present growing evidence for learning of courtship motor patterns other than song. We tackle two main challenges: we first highlight criteria that can be used to determine whether visual courtship components are learned, and if so, we then assess the type of learning involved. In line with the vocal learning literature, we suggest applying a distinction between usage learning and production learning of motor patterns: usage learning refers to a change in the context in which pre-existing display patterns are used, whereas production learning involves modification in trait structure, i.e. the acquisition of novel display patterns from a model. The effects of imitation, social feedback, and practice are described in detail, drawing on multiple examples from birdsong research. Our goals are to illustrate the learning processes which may affect motor development of courtship signals, to formulate testable predictions for each learning category and related mechanisms, and to suggest possible lines for future research. Although most of the evidence we review here is indirect and not yet conclusive about learning, recent technological advances now provide novel tools to quantify courtship motor patterns, and thus have the potential to produce more direct insights into whether, and how, courtship displays are learned.
... There are also operating costs associated with learning the unique features of many individuals, including the time, energy and resources required to collect, store and recall information. Long-term memory formation also reduces immunity, survival and fecundity [126,127]. ...
Animal groups are often organized hierarchically, with dominant individuals gaining priority access to resources and reproduction over subordinate individuals. Initial dominance hierarchy formation may be influenced by multiple interacting factors, including an animal's individual attributes, conventions and self-organizing social dynamics. After establishment, hierarchies are typically maintained over the long-term because individuals save time, energy and reduce the risk of injury by recognizing and abiding by established dominance relationships. A separate set of behaviours are used to maintain dominance relationships within groups, including behaviours that stabilize ranks (punishment, threats, behavioural asymmetry), as well as signals that provide information about dominance rank (individual identity signals, signals of dominance). In this review, we describe the behaviours used to establish and maintain dominance hierarchies across different taxa and types of societies. We also review opportunities for future research including: testing how self-organizing behavioural dynamics interact with other factors to mediate dominance hierarchy formation, measuring the long-term stability of social hierarchies and the factors that disrupt hierarchy stability, incorporating phenotypic plasticity into our understanding of the behavioural dynamics of hierarchies and considering how cognition coevolves with the behaviours used to establish and maintain hierarchies.
This article is part of the theme issue ‘The centennial of the pecking order: current state and future prospects for the study of dominance hierarchies’.
... The presence of a lineage-specific male response to prior experience and a lineage-specific presence of assortative courtship suggest that H. m. malleti and H. m. rosina may experience different mating-related selective pressures. Maintaining the capacity to learn can be energetically costly and is often associated with fitness trade-offs, such as reduced fecundity (Kotrschal et al., 2013;Snell-Rood et al., 2011), reduced life span (Burger et al., 2008;Kotrschal et al., 2019) or extended development time (Kolss & Kawecki, 2008). Although H. m. malleti and H. m. rosina are conspecifics, they do not co-occur in nature (Brower, 1996;Rosser et al., 2012). ...
Many animals have the ability to learn, and some taxa have shown learned mate preference, which may be important for speciation. The butterfly Heliconius melpomene is a model system for several areas of research, including hybridization, mate selection and speciation, partially due to its widespread diversity of wing patterns. It remains unclear whether social experience shapes realized mating preferences in this species. Here we test whether previous experience with a female influences male mate preference for two H. melpomene subspecies, H. m. malleti and H. m. rosina. We conducted no-choice assays to determine whether male courtship (versus no courtship) and latency to court differed between naïve males and males with previous exposure to a sexually mature, virgin female. To test whether assortative courtship preference is learned in H. melpomene, males were either paired with a female who shared their phenotype or one who did not. Naïve H. m. malleti males courted assortatively, while naïve H. m. rosina males did not. When data were pooled across subspecies, experienced males reduced their courting relative to naïve males, suggesting that social experience with a female sans copulation may be perceived as a negative experience. This effect was likely driven by experienced H. m. malleti males, who reduced their courting relative to naïve males when analysed independently, while experienced H. m. rosina males did not. Our results suggest that social experience can influence male mating behaviour in H. melpomene and has the potential to contribute to the high rate of diversification observed in Heliconius butterflies.
... The shorter foraging careers of faster visual learners (Evans et al., 2017) was thought to have resulted from the energetic cost associated with enhanced cognitive performance, which can negatively impact other energetically demanding processes (Mery and Kawecki, 2003;Mery and Kawecki, 2004;Snell-Rood et al., 2011;Jaumann et al., 2013). Another study provides evidence of a "trade-off " in the opposite direction-increased foraging time lowered olfactory learning performance (reversal learning) among honey bees (Cabirol et al., 2018), further support for an inverse relationship between learning and foraging duration. ...
Individual animals allowed the opportunity to learn generally outperform those prevented from learning, yet, within a species the capacity for learning varies markedly. The evolutionary processes that maintain this variation in learning ability are not yet well understood. Several studies demonstrate links between fitness traits and visual learning, but the selection pressures operating on cognitive traits are likely influenced by multiple sensory modalities. In addition to vision, most animals will use a combination of hearing, olfaction (smell), gustation (taste), and touch to gain information about their environment. Some animals demonstrate individual preference for, or enhanced learning performance using certain senses in relation to particular aspects of their behaviour (e.g., foraging), whereas conspecific individuals may show different preferences. By assessing fitness traits in relation to different sensory modalities we will strengthen our understanding of factors driving observed variation in learning ability. We assessed the relationship between the olfactory learning ability of bumble bees (Bombus terrestris) and their foraging performance in their natural environment. We found that bees which failed to learn this odour-reward association had shorter foraging careers; foraging for fewer days and thus provisioning their colonies with fewer resources. This was not due to a reduced propensity to forage, but may have been due to a reduced ability to return to their colony. When comparing among only individuals that did learn, we found that the rate at which floral resources were collected was similar, regardless of how they performed in the olfactory learning task. Our results demonstrate that an ability to learn olfactory cues can have a positive impact of the foraging performance of B. terrestris in a natural environment, but echo findings of earlier studies on visual learning, which suggest that enhanced learning is not necessarily beneficial for bee foragers provisioning their colony.
... A third possible mechanism by which chronic stress may accelerate brain development is that young individuals process threat as an overall signal of lack of protection and support -that is, they receive cues that the environment requires maturity -and this triggers adaptive top-down processes that cause development to proceed more quickly. This was recently termed the 'developmental support hypothesis' (see reF. 93 ), and aligns with much evolutionary life-history research, including cross-species findings that parental investment is associated with slower maturation [93][94][95] . Understanding which, if any, of these mechanisms affect the pace of brain development is essential for determining when and how it might be possible to intervene. ...
Childhood socio-economic status (SES), a measure of the availability of material and social resources, is one of the strongest predictors of lifelong well-being. Here we review evidence that experiences associated with childhood SES affect not only the outcome but also the pace of brain development. We argue that higher childhood SES is associated with protracted structural brain development and a prolonged trajectory of functional network segregation, ultimately leading to more efficient cortical networks in adulthood. We hypothesize that greater exposure to chronic stress accelerates brain maturation, whereas greater access to novel positive experiences decelerates maturation. We discuss the impact of variation in the pace of brain development on plasticity and learning. We provide a generative theoretical framework to catalyse future basic science and translational research on environmental influences on brain development.
... The associated neural machinery and process of forming long-term memories are metabolically costly (Laughlin et al. 1998;Mery and Kawecki 2005), and large brains result in exponential increases in developmental time (Workman et al. 2013). Thus, the evolution of learning results in tradeoffs in juvenile competitive ability (Mery and Kawecki 2003) and adult reproduction, often delaying reproduction and resulting in greater investment in fewer offspring (Barrickman et al. 2008;Snell-Rood et al. 2011;Kotrschal et al. 2013). In many cases, there are even direct tissue tradeoffs between plasticity machinery (such as brain size) and other costly tissues, such as gut or flight muscle (Isler and van Schaik 2006;Liao et al. 2016). ...
... Therefore, investment in learning and the maintenance of learning skills can result in trade-offs with investments in other biological functions. Previous studies have confirmed this expectation and showed that in some species, the ability to learn is associated with delayed reproduction , 2004Snell-Rood et al. 2011). This phenomenon occurs because resources that might be invested in reproductive tissue are redirected to development and maintenance of energetically expensive neural tissue, which is required for learning and memory (Laughlin et al. 1998). ...
... Studies that have evaluated trade-offs between learning ability and other life functions are not limited to fruit flies. Negative correlations between learning ability and trade-offs were found in butterflies (Pieris rapae ) (Snell-Rood et al. 2011), antlion (Myrmeleon bore ) larvae ) and predatory mites (Phytoseiidae) (Christiansen et al. 2016). ...
... All previous studies that focused on trade-offs between learning skills and the reproductive potential of individuals focused on solitary species (e.g., Mery and Kawecki 2004;Mery and Kawecki 2005;Snell-Rood et al. 2011). However, the honeybee is a eusocial species, where individuals live together and the whole society is involved in the raising of the queen's offspring. ...
Learning ability, which allows individuals to adjust their behaviour to changing environmental conditions, has a considerable positive impact on individual fitness. However, in addition to benefits, learning also incurs a cost, which means that investment in learning and maintaining learned skills can lead to trade-offs impacting other biological functions. Here, we tested whether a trade-off exists between learning skills and reproductive potential in honeybee workers. For this purpose, we compared learning ability between two groups of workers that differed in reproductive potential—normal and rebel workers. The results showed that workers with high reproductive potential (rebels), measured according to the number of ovarioles in the ovary, learned faster than normal workers with low reproductive potential. Moreover, by performing separate regression analyses within the rebel and non-rebel worker groups, we found that the reproductive potential of workers was positively correlated with their learning ability. The results show that in honeybees, there is no trade-off in resource allocation between two costly biological functions, learning and reproduction.
