Omnivorous animals forage in complex and heterogenous food landscapes and must choose the correct mix of nutrients for growth and fitness. Many insects have evolved the ability to make such choices and will select an array of foods that offer the correct balance of essential nutrients. Researchers have used artificial diets to examine the regulation of macronutrients, especially protein and carbohydrates, in a number of diverse species (1). These studies have demonstrated that insects feed selectively when offered two foods which differ in protein:carbohydrate ratios, causing intake of each nutrient to match a ratio that promotes species-specific fitness (2–9). Insects also regulate the amount of food consumed and adjust their intake depending on the nutrient density of the food. For example, grasshoppers compensate behaviorally for nitrogen dilute diets by increasing the number and duration of feeding bouts (10); bumble bees decrease consumption of pollen as the concentration of protein increases (11). An additional key finding is that, following protein or carbohydrate deprivation, insects subsequently engage in compensatory feeding on the limited substance to restore nutrient balance. The compensatory behavior can occur following days of deprivation or just a single meal that lacks a macronutrient (12–16).

In addition to macronutrient selection, insects can distinguish other substances, including micronutrients and free amino acids. Amino acids are important since they may signal the presence of protein and are necessary for numerous biochemical functions, including neurotransmitter biosynthesis. The ability of insects to perceive free amino acids and self-select food based on this perception has been studied in Drosophila adults and larvae (17–19) and nectar-feeding insects such as bees, ants, and mosquitoes (20–25). These studies demonstrate that an amino acid and sugar mix may be preferred over sugar alone, and that individual amino acids vary in their ability to stimulate feeding. Studies have also demonstrated that diets lacking essential amino acids (EAAs) can cause rejection of the unbalanced food (26) and enhance attraction to and feeding on amino acid mixtures or protein (17, 27). Lack of even a single EAA can induce this compensatory feeding on amino acid mixtures or protein (24, 28). Adaptive changes in amino acid intake can involve changes in search behavior to locate sources of amino acids or changes in sensory neuron responses to favor amino acid feeding (1, 29).

In addition to nutritional state, responses to protein, carbohydrate, and other nutrients vary with developmental stage, sex, and mating status. For example, Cohen et al. (3) found that carbohydrate consumption varied as cockroach larvae developed through successive instars, with intake highest directly following molting. Carbohydrate intake also shifted in mealworm larvae becoming greater in later compared to earlier instars (9). In Drosophila, larvae chose to linger on yeast patches (30) whereas well-nourished adult Drosophila fed preferentially on sucrose. However, compared to males, female Drosophila show a greater preference for feeding on yeast, a source of protein (18). Likewise adult crickets of both sexes prefer carbohydrate over protein (31, 32), but females consume more protein than males (31). The preference for protein was markedly enhanced in Drosophila females after mating, a response attributed to an elevated need for protein to produce eggs (15, 33). In the cricket Gryllus bimaculatus, protein consumption was also greater in mated compared to virgin females (34). Because the dietary switch occurred after the female crickets laid eggs, Tsukamoto et al. suggest that the enhanced protein appetite reflected the need to produce additional eggs for future episodes of mating and egg laying.

The house cricket (Acheta domesticus) is a common model organism for both research and teaching and is also important commercially as pet food and more recently as a source of protein for human consumption (35). Like many cricket species, A. domesticus is omnivorous, consuming foods that vary greatly in nutritional content across space and time. These feeding habits predict that crickets will be tolerant of temporary nutritional imbalances and able to address imbalances by compensatory feeding when needed nutrients become available (14). Compensatory feeding on macronutrients has been shown in several cricket species, but the ability to detect and self-select amino acids has not yet been studied. The quantity of amino acids consumed affects growth, reproduction, and longevity in insects and, as in mammals, certain amino acids are essential and must be acquired from the diet. We designed experiments to examine the ability of house crickets to detect EAA, and investigated how feeding decisions are affected by nutritional state, age, and sex.

Our first experiment addressed the hypothesis that house crickets can discriminate between sucrose and EAA, and that preference for the latter would be enhanced following a meal containing only sucrose. In this experiment, we used a two-choice preference test to compare food intake and feeding behavior of crickets following pre-feeding with sucrose or EAA. Previous studies have demonstrated that appetite for carbohydrates and protein can vary significantly in insects depending on age and sex. To determine if such variation occurs in house crickets, we tested males and females separately and compared responses in last instar nymphs and adult animals. Previous studies have also reported that female reproductive status affects protein appetite, and hence the first experiment included two additional groups: mated females, and mated females allowed to lay eggs. For the latter comparisons, we predicted that amino acid appetite would be enhanced in females compared to males, and in mated compared to virgin females. In our second experiment, we used extended periods of EAA deprivation to determine if crickets would subsequently engage in compensatory feeding on an amino acid mixture. Finally, to further understand cricket motivation for amino acid consumption, we examined growth and survival of juvenile and adult crickets fed a holidic diet lacking all amino acids and protein.

The agar test diets were well accepted by the crickets and only four tanks (two juvenile male, one juvenile female, one adult male) were eliminated due to lack of any food consumption. Figure 2 displays the amount of sucrose and EAA food consumed following pre-feeding with each substance (Experiment 1). In adult virgin females, the intake of EAA was significantly greater than sucrose following both sucrose (Z = -2.073, p = 0.038) and EAA pre-feeding (Z = -2.668, p = 0.008). In juvenile females, the intake of EAA was significantly greater than sucrose following pre-feeding with sucrose (Z = -2.366, p = 0.018). In contrast to these results, mated females significantly preferred sucrose following EAA pre-feeding (Z = -2.395, p = 0.017) and females that produced eggs significantly preferred sucrose following pre-feeding with EAA (Z = -2.803. p = 0.005) and sucrose (Z = -2.075, p = 0.035). Other within group comparisons found insignificant differences in sucrose and EAA intake regardless of preparation. Data on feeding behavior was consistent with the preferences noted above ( Table 1 ). A preliminary analysis of 20 randomly selected trials found that most (M = 77%) feeding behavior occurred during the first hour and hence our subsequent analysis focused on this time period. Adult virgin females spent significantly more time feeding on EAA disks compared to sucrose disks after both EAA and sucrose pre-feeding, and visited EAA disks more times after EAA pre-feeding. Mated females spent significantly more time feeding on sucrose disks compared to EAA disks after EAA pre-feeding and visited sucrose disks more; results were similar for mated females allowed to lay eggs ( Table 1 ). Despite significant differences in feeding time, visits, and food intake, each food was sampled multiple times suggesting that crickets chose to consume more sucrose or EAA after sampling both food types.

To examine EAA intake across groups, we compared the percent of EAA food consumed relative to total food intake in the six test groups ( Figure 3 ). Following sucrose pre-feeding, there was no significant difference among groups in the percent of EAA food consumed (H(5) = 10.61, p = 0.06), though intake tended to be greater in adult and juvenile females. Following EAA pre-feeding, the percent of EAA intake differed significantly among groups (H(5) = 20.281, p = 0.001). Dunn-Bonferroni pairwise comparisons indicated that mated females post-egg laying consumed significantly less EAA food compared to both adult (p = 0.024) and juvenile females (p = 0.005). These data combined with the previous comparisons indicate that females differentiate between sucrose and EAA, and favor EAA consumption over sucrose prior to mating. However, this preference reverses after mating and egg-laying. Under the conditions of Experiment 1, juvenile and adult males did not significantly differ in sucrose and EAA food intake though both groups tended to prefer sucrose.

For Experiment 2, consumption of EAA in adult males differed significantly among the four groups tested after 24 h, 48 h, 72 h, and six days of EAA deprivation (H(3) = 16.9, p = 0.001; Figure 4 ). We used Dunn-Bonferroni pairwise tests to compare nutrient intake among groups and found that, compared to Group 1 (24 h EAA deprivation), EAA consumption increased significantly in Group 3 (p = 0.039) and Group 4 (p = 0.001), but not Group 2 (p = 0.465). A concomitant decrease in sucrose intake occurred across the four different EAA deprivation periods (H(3) = 19.3, p = 0.001). Dunn-Bonferroni pairwise tests indicated that, compared to Group 1, sucrose consumption decreased significantly in Group 2 (p = 0.04), Group 3 (p = 0.001), and Group 4 (p = 0.001). Because the amount of food consumed varied among groups, we also considered the percent of EAA food consumed relative to total food intake in the four groups ( Figure 5 ). This analysis confirmed the finding that EEA consumption differed among groups (H(3) = 23.8, p = 0.001) and, compared to Group 1, was significantly increased in Groups 3 (p = 0.001) and 4 (p = 0.001), but not Group 2 (p = 0.164). These data indicate that male crickets differentiate between sucrose and EAA, and after 72 h of EAA deprivation their preference switches from sucrose to EAA.

In adult females, consumption of EAA (H(3) = 13.9, p = 0.003) and sucrose (H(3) = 14.6, p = 0.002) also differed significantly among the four groups ( Figure 4 ). For EAA consumption, Dunn-Bonferroni tests indicated that, compared to Group 1, EAA consumption increased significantly in Group 2 (p = 0.003) and Group 4 (p = 0.027), but not Group 3 (p = 0.357). For sucrose, compared to Group 1, consumption decreased in Group 3 (p = 0.011) and Group 4 (p = 0.016), but not Group 2 (p = 0.786). A comparison of the percent of EAA food consumed relative to total food intake revealed significant differences among groups (H(3) = 18.3, p = 0.001) and significant increases in EAA intake in Groups 2 (p = 0.032), 3 (p = 0.003), and 4 (p = 0.001), relative to Group 1 ( Figure 5 ). These data support the previous finding that females differentiate between sucrose and EAA and they switch their preference from sucrose to EAA after 48 h of EAA deprivation.

In Experiment 3, the overall survival of juvenile crickets in community tanks differed significantly among the three diets (log-rank test: X² = 61.4, p < 0.001, Figure 6A ). Pairwise log-rank comparisons indicated that survival in the control diet differed significantly from the EAA diet (X² = 26.8, p < 0.001) and the No EAA diet (X² = 63.9, p < 0.001); also, the EAA and No EAA diets differed significantly in juvenile survival (X² = 10.5, p = 0.001). Crickets fed the control diet had a 98% survival rate, an average increase in weight of 320 mg/cricket, and 78% (47 insects) had molted to adulthood by the end of the 30-day period. Crickets fed the artificial diet with EAA or No EAA had survival rates of 60% and 30%, respectively. However, growth was severely restricted in both groups (average weight increase in EAA group: 21 mg/cricket; No EAA group: 14 mg/cricket) and no insects molted to adulthood. These data indicate that, while both artificial diets limited growth, access to EAA significantly enhanced survival. For juvenile crickets held individually, survival was significantly greater in the control group compared to the crickets in the No EAA group (log rank test: X² = 9.8, p < 0.002, Figure 6B ). All juveniles fed the control diet survived and molted to adulthood during the 30-day period, whereas 60% of the crickets fed the No EAA diet survived and growth was restricted (average weight increase over 30 days was 15 mg/cricket). These data confirm the superior performance of juveniles fed control food and suggest that reduced survivorship in the community tanks fed the No EAA diet may have been due in part to greater levels of cannibalism. All adult crickets of both sexes survived for 30 days when fed the control, EAA, or No EAA diet, indicating that the three diets were able to sustain life for this time period.

The present results contribute several novel findings about nutrient selection in the house cricket. First, we found that house crickets can distinguish EAA from sucrose and that selection of these two nutrients is affected by prior feeding experience. We also found that age and reproductive state affected food choice, with juvenile and virgin females showing a greater preference for EAA than juvenile or adult males. Contrary to previous studies, mated females preferred sucrose over EAA both prior to and after egg laying. We also found that crickets of both sexes increased their intake of EAA when exposed to longer periods of deprivation, suggesting that they engage in compensatory feeding on these nutrients. Finally, we found that growth was severely limited in juveniles fed a diet lacking all amino acids. However, many juveniles and all adults survived on this diet for 30 days indicating that they possess considerable resilience to dietary restriction.

Our results support previous studies which have found that insects respond behaviorally to free amino acids when making feeding decisions. These chemicals may serve as attractants or deterrents in nectar feeders, allowing insects to gain needed nutrients while avoiding excessive amino acid intake (20, 22, 23). In Drosophila, adaptive responses to amino acids may likewise allow larvae (39) and adults to acquire an appropriate balance of these nutrients and respond to deprivation with enhanced amino acid preference. House crickets are highly generalized feeders and are likely to occur in areas where intensive farming creates environments with nutritional imbalances (40). Hence, they may encounter foods that vary in the quantity and quality of amino acids or where abundant food sources lack EAA. Dietary challenges may be exacerbated by the tendency of house crickets to aggregate (41), causing social stress and competition for food (42, 43). Hence the ability to detect EAA and compensate for deprivation is likely to be especially adaptive in these insects. The site of amino acid perception has not yet been established in crickets, but in other insects individual amino acids can be detected by taste sensilla on the legs (18, 44) and labellum (45, 46). Recent studies have also identified molecular receptors for amino acids in Drosophila including the odorant binding protein OBP19b (47), the ionotropic receptor Ir76b (18, 39), and several additional ionotropic and gustatory receptors (46). Comparative research is needed to determine if these mechanisms are conserved across other insect species.

In experiment 1, we found that after 48 h EAA deprivation, juvenile and adult virgin females chose to consume more EAA than sucrose; however, juvenile and adult males did not display a preference for EAA under the same conditions. When protein is limited, free EAA are a source of required nutrients and the attraction of females to EAA is consistent with the tendency of female crickets to consume more protein than males (31). Oogenesis in house crickets begins on day three following adult ecdysis and oviducts are full by day 15 (48), the period of time during which we tested virgin females. Hence the need to produce eggs may drive EAA appetite in females just prior to and after the adult ecdysis. The preference for sucrose after mating and egg laying in female house crickets contrasts with findings in Drosophila (15, 33) and two-spotted crickets (34) in which protein ingestion increased after mating. Our results possibly reflect motivation to replenish energy via carbohydrate feeding immediately following periods of heightened physical activity and decreased food intake which occur during mating and egg laying (48). In future studies, it would be useful to examine food preferences throughout adulthood to better understand the associations between protein or EAA ingestion and oogenesis, mating, and oviposition.

In male crickets, sucrose pre-feeding for 72 h and 6 days induced a switch from sucrose to EAA preference (Experiment 2). Virgin females also consumed more sucrose after a 24 h fast, but consistent with the results of Experiment 1, preferred EAA after 48 h deprivation and one sucrose pre-feeding. These results indicate that both sexes perceive and compensate for EAA deficiency. However, under our experimental conditions, house crickets did so only after a relatively long period of deprivation whereas some other species displayed compensatory selection of carbohydrate and protein after dietary restrictions lasting only a few hours (12, 13). The results of the survival study (Experiment 3) are interesting in this context since they indicate that adult crickets displayed resilience in the face of EAA deprivation given that 100% of adults of both sexes survived for 30 days on sucrose alone. This suggests that, as an omnivorous and opportunistic feeder, adult house crickets have evolved to tolerate temporary nutritional imbalances and may not be under strong selective pressure to very rapidly compensate for dips in EAA availability. The situation is different for juvenile stages since we found that growth and survival were significantly impacted by lack of EAA in 4^th instar nymphs. We did not assess dietary choice in early cricket instars, but because optimal growth depends on EAA availability it is possible that a compensatory response to EAA deprivation would occur sooner in juveniles relative to adults.

Many additional discoveries remain to be made about cricket diet selection. Further investigations of feeding behavior at different life stages would contribute additional information on the adjustments insects make to optimize nutritional balance as they mature and reproduce (3). In addition, other major behavioral events such as aggressive interactions and courtship singing may trigger changes in nutrient selection, a possibility that remains to be examined in crickets and other species. Another topic of interest concerns the perception and selection of individual amino acids. Previous work has shown that some insects distinguish among single amino acids; for example, Drosophila larvae are most attracted aspartic acid and avoid leucine (19, 39), and honey bees are more attracted to essential amino acids, especially phenylalanine, compared to nonessential amino acids (23). Such information is not yet available in crickets and would be useful for comparative purposes and as a foundation for future physiological studies. The latter are important since the neural mechanisms used by crickets to sense amino acids, both externally and internally, are not yet elucidated. Finally, amino acid appetite may differ between insects from inbred laboratory colonies and their wild counterparts. For example, in the laboratory locusts regulate nutrient intake in a consistent manner, whereas wild locusts display considerable plasticity in their feeding behavior across generations and seasons (49). Also, gut microbial communities differ between wild crickets and those confined and fed a laboratory diet, a factor that may affect how amino acids are synthesized and degraded (50, 51). Our study examined amino acid appetite in laboratory animals fed a commercial cricket diet and more general conclusions could be drawn by extending the experiments to wild populations.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

AT designed the study, ran experiments, analyzed data, and wrote the original draft of the manuscript. EV, LJ, SH, MP, and MS contributed to data collection and analysis. All authors contributed to the article and approved the submitted version.