| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Behavioral Ecology Vol. 15 No. 1: 63-70
© 2004 International Society for Behavioral Ecology
The relative availabilities of complementary resources affect the feeding preferences of ant colonies
Department of Biology, University of Utah, Salt Lake City UT 84112, USA
Address correspondence to A. Kay, who is now at the Department of Ecology, Evolution and Behavior Ecology, University of Minnesota, St. Paul, MN 55108, USA. E-mail: kayxx011{at}umn.edu.
Received 2 February 2002; revised 1 December 2002; accepted 3 February 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Theory predicts that consumers selecting among complementary resources will show stronger preferences for items that become relatively less available. I tested this hypothesis in a field study that compared the preferences of ant colonies given simultaneous access to experimental foods differing in carbohydrate and protein content. In the first part of the study, I examined the effect of nutrient supplementation on colony-level preference in the ant Dorymyrmex smithi. Colonies that had received a protein solution for 24 h consumed proportionally more carbohydrates than control colonies that had been given access to water, suggesting that colonies preferred nutrients when they became relatively rare. In the second part of the study, I compared colony-level preference among eight species of ants that differ in their relative access to carbohydrates and protein in the field. I found that species with relatively easy access to carbohydrates preferred protein, whereas species with greater access to protein preferred carbohydrates. These results suggest that the benefits of a nutritionally mixed diet coupled with differences in the relative availability of nutrients may explain variation in feeding decisions both within and among ant species.
Key words: ants, complementary resources, foraging theory, preference.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
When confronted with alternative feeding opportunities, foraging animals must often make decisions that ultimately affect their reproductive success. When foragers encounter alternatives simultaneously, differences in the characteristics of preferred and neglected items can reveal the factors that influence their choices (Howard, 1987
; Voelkl et al., 1999). Identifying these factors is often useful for understanding not only patterns of patch and habitat use (Brodbeck et al., 1990; Rank, 1992) but also for understanding higher-level ecological processes such as competitive interactions (Kronfeld and Dayar, 1999; Savolainen, 1991) and the impact of consumers on prey communities (Crist and MacMahon, 1992; Hjälten, 1997).
In this study, I examined the preferences of ant colonies for foods that differ in carbohydrate and protein content. Investigations of foraging decisions are frequently based on models that use a single resource currency as a surrogate for reproductive success (Schoener, 1971). One limitation of these models is that they cannot be readily applied to situations where interactions among multiple demands influence success (Stephens and Krebs, 1986). For example, because different macronutrients synergistically improve nutrition (Evans and Pierce, 1995; Raubenheimer and Simpson, 1997), consumers may choose foods that provide them with an appropriate balance of nutrients rather than an abundance of any single resource (Dearing and Schall, 1992; Pennings et al., 1993; Sanchez-Vasquez et al., 1999).
One approach for analyzing such decisions uses a conceptual framework derived from microeconomic theory (Covich, 1972; Rapport, 1971). In this approach, different combinations of resources are assumed to provide a consumer with equal fitness benefits. When resources synergistically improve performance (i.e., they are complementary; Tilman, 1982), combinations yielding equal benefits are represented by an isocline that is convex to the origin (Figure 1). The combinations that a consumer can select are constrained by the availabilities of the resources. This constraint is represented by a curve, which shows the possible combinations of the two resources that a consumer can acquire during a certain time. Factors that make resources more available, such as an increase in resource abundance or greater foraging skill, shift the constraint curve farther from the origin. The slope of the constraint curve reflects the relative availability of the resources. Together, the isoclines and constraint curve specify optimal consumer choice: to maximize fitness, a consumer should select the resource combination on the highest isocline that intersects the constraint curve. For complementary resources, this combination is also the point of tangency between an isocline and the constraint curve (see Stephens and Krebs, 1986, for a more thorough review of this approach).
|
This approach is useful for predicting preference. Preference reflects the relative value of items that are presented as alternatives. In the model, the relative value of resources is equal to the inverse of the marginal rate of substitution (MRS), the amount of one resource that a consumer can substitute for one unit of another without changing fitness. Because the MRS equals the inverse of the relative value of the resources, it should also be inversely related to a consumer's relative preference for those resources. For example, when the MRS of protein for carbohydrates decreases, a consumer is willing to exchange less protein for a unit of carbohydrates, and thus it should show a stronger preference for protein when choosing between the two resources. Because the MRS is given by the negative of the isocline's slope, the relative value of items can be represented graphically by the slope of a vector perpendicular to the isocline (Figure 1).
The model makes several predictions about preference among foods containing different ratios of complementary resources. First, preference for a food should decrease when the nutrients in that food become relatively more available (Figure 1). In the first part of this study, I tested this hypothesis by investigating whether colonies supplemented with protein show a stronger preference for carbohydrates when they encounter protein and carbohydrates simultaneously. Although similar predictions have been tested in other systems (Cottam, 1985; Greenstone, 1979; Pennings et al., 1993), this study is the first to test the effect of resource supplementation on ant colony preferences. Second, the preferences of consumers in the field should be inversely related to the relative availability of resources in their environment. I tested this prediction by comparing preferences among eight ant species, four of which collect both invertebrates and carbohydrate-rich exudates from plants and homopterans and four that forage for invertebrates or seeds but are not known to collect exudates. Because carbohydrate:protein availability ratios are higher for exudate feeders (Kay, 2002), these species should have a stronger preference for proteinaceous foods. This study is the first to investigate interspecific differences in preference among complementary resources.
I assayed the preferences of ant colonies using simultaneous feeding experiments, which likely simulate situations faced by ants under natural conditions. Although individual workers may be no more likely to encounter resources simultaneously than are other terrestrial animals, ant colonies may frequently discover resource patches at the same time. In most ants, numerous workers concurrently search the surrounding environment for food (Hölldobler and Wilson, 1990) and recruit nest mates to valuable finds. If food items from different sites are brought back to the nest at about the same time, the colony in essence discovers each site simultaneously. Colonies can then exhibit a preference among sites through differential recruitment. Simultaneous feeding experiments on ants have frequently been used to test predictions from optimal foraging theory (Holder Bailey and Polis, 1987; Taylor, 1977) and to identify the feeding habits of colonies (Bristow and Yanity, 1999; Folgarait et al., 1996; Lanza et al., 1993).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Study organisms
I conducted this study near Portal, Arizona, USA, during July–October 1998. For the supplementation experiment, I used colonies of Dorymyrmex smithi located in oak woodland habitat (elevation 1700 m). For the interspecific comparison of preference, I studied four species (D. smithi, Myrmecocystus mendax, Pheidole diversipilosa, and Pogonomyrmex occidentalis) in oak woodland habitat and four others (Aphaenogaster cockerelli, Forelius sp. 1, Formica perpilosa, and Pogonomyrmex barbatus) in desert thornscrub near the town of Portal (elevation 1400). Forelius sp. 1 is apparently an undescribed species, distinguishable from Forelius maccooki by the lack of erect setae on the antennal scapes (S. Cover, personal communication). The diets of these species fall into two categories: they either contain exudates from plants and homopterans or they do not (Table 1). The species that do not collect exudates are members of the subfamily Myrmecinae, whereas those species that do collect exudates are members of two closely related subfamilies, the Dolichoderinae (D. smithi and F. sp. 1) and the Formicinae (F. perpilosa and M. mendax).
|
General methods
I used one general approach for measuring colony-level preference. However, because the species in this study have a wide range of foraging characteristics, I had to modify the details of the approach in several cases.
The general approach involved measuring the preferences of colonies among three foods containing either sucrose, casein hydrolysate enzymatic (a soluble source of amino acids), and a 1:1 (mass:mass) mixture of sucrose and casein. I used experimental foods because it allowed me to manipulate precisely the nutritional content of foods while controlling other attributes that could affect preference such as nutrient density and texture. Nutrient content in all foods was 15% (mass/mass).
In all cases, I began a trial by placing foods approximately 25 cm from a colony's entrance. To ensure that patterns of food exploitation reflected choice at the level of the colony, I spatially separated food types. Because ants recruit nest mates to locations, the spatial segregation of food types allowed differential recruitment to affect the pattern of food collection. I also switched the position of foods after the baits were first discovered but before substantial recruitment had begun. When species have effective recruitment systems, a food discovered first may be heavily exploited before the other foods are found. Differences in food collection could thus be the result of differences in discovery time rather than colony-level selection. After I switched the positions of foods, recruitment trails led workers to a food that had not been visited previously. At the same time, visitors that were already on the discovered baits began returning to the nest and initiated new recruitment trails. Thus, switching the position of foods ensured that recruitment to all foods began at about the same time.
Colonies had access to foods for 50 min during each trial. I began timing trials when at least three workers had sampled and left each food. To keep workers from visiting less attractive baits simply because preferred baits were crowded, I added more of all three food types whenever workers filled more than three-quarters of the feeding sites at any one of the baits. I placed these additions contiguous to the original baits. I added extra food to only one F. sp. 1 colony and three A. cockerelli colonies. To ensure that only workers from the test colony visited the baits, I guarded all racks during the experiment by aspirating any heterospecific ant or conspecific neighbor that approached a bait. I abandoned a trial if a foreign colony began recruiting to the racks.
I provided colonies with foods in one of two forms. For six species, A. cockerelli, P. occidentalis, D. smithi, F. sp. 1, F. perpilosa, and M. mendax, I provided colonies with solutions in 0.65 ml Eppendorf tubes supported in wooden racks. Each rack could hold 16 tubes. Workers climbed onto the racks and drank while standing on the edge of a tube. This approach provided colonies with easy access to the resource while limiting the exposed surface area of the liquid and thus reducing evaporation. In two species, P. diversipilosa and P. barbatus, workers would not climb onto the racks to collect the solutions. For these species, I converted the solutions into agar-based gels. Converting solutions into gels likely did not affect the results of the preference tests: like colonies choosing among solutions, colonies with access to gels still chose among foods that differed only in carbohydrate and protein content. I provided pieces of the gels to colonies on separate notecards wrapped in aluminum foil.
For D. smithi, M. mendax, F. sp. 1, and F. perpilosa, I determined preference by comparing the mass loss of each food during the experiment after correcting for loss due to evaporation. To correct for evaporation, I placed racks containing tubes filled with 15% sucrose solution in a microclimate that was similar to that in which experimental racks were located. In preliminary work, I found that 15% sucrose evaporated at the same rate as the solutions with casein. I created additional control racks whenever I added extra food to heavily visited trials and used only tubes from those racks to correct for evaporation from the extra food. I placed control racks in a shallow plastic tub whose outer surface was coated with fluon to prevent ant visitation. At the end of a trial, I cleared foragers from experimental tubes by prodding them with a toothpick. When a tube was free of foragers, I removed it from the rack, sealed it, and placed it on ice until I could return it to the lab for weighing. Ants occasionally fell into the tubes. If an ant was floating on the surface, I removed it with forceps. I did not retrieve ants if they had sunk to the bottom of the tube because removing them would have resulted in the loss of a substantial amount of solution. I estimated that the mass of sunken ants was always less than 1/30 of the mass of liquid consumed from the tubes.
I used periodic censusing rather than mass loss comparisons to determine preference in P. diversipilosa, P. occidentalis, A. cockerelli, and P. barbatus. Using mass loss to assess preference was essential for D. smithi and F. sp. 1 because workers in these species are small (<2 mg), fast-moving, and forage in large numbers; they are thus extremely difficult to census accurately. However, comparing mass loss was ineffective for measuring preference in P. diversipilosa, P. occidentalis, A. cockerelli, and P. barbatus because workers in these species collect liquid very slowly (A. Kay, unpublished manuscript). As a result, mass loss in experimental foods often did not differ significantly from that in control foods. (In fact, experimental Eppendorf tubes that were heavily visited often weighed more after a trial than did control tubes, indicating that workers acted like a plug that prevented evaporation.) For these species, I used relative visitation to baits to compare food collection. Because the three food types were presented to a colony in the same form, differences in the number of ants at each bait were likely similar to differences in the amounts of food harvested. To assay relative visitation, I counted the number of foragers collecting each food type every 10 min for 50 min. I began each census 10 min after all foods had been discovered. For P. diversipilosa, a species with a completely dimorphic worker caste, I counted both the number of minors and the number of majors at baits. Because minors and majors collect liquid resources at a rate proportional to their mass (Kay, unpublished manuscript), I converted majors into minor equivalents by dividing the number of majors by the ratio of major:minor body mass (=3.16).
Experiments
In the first experiment, I investigated how increasing a colony's access to protein affected its preference. I provided six experimental colonies of D. smithi with a 6% casein solution during daylight hours (ca. 0500–1900 h) throughout a 24-h period; I also provided six control colonies with tubes filled with water during the same period. A 6% casein solution is a highly attractive resource for D. smithi: workers almost always accept drops of this solution when they encounter them (Kay, 2002), and colonies rapidly recruited to the solution after it was discovered. To deliver the liquid to a colony, I put the casein solution or the water in a rack of Eppendorf tubes and placed the rack within 0.5 m of the nest entrance. Workers quickly discovered these supplements. I provided replacement tubes with fresh solution every 2–3 h. I set out enough tubes to ensure that colonies had unlimited access to the solution during the supplementation period. At the end of the 24-h period, I removed the rack of tubes and began the preference test after a 30 min delay. I supplemented and then tested the preference of one experimental colony and one control colony per day over the course of 6 days. Each day, I conducted preference tests sequentially. I randomly determined the treatment of the first trial on the first day and then alternated the treatment order on subsequent days.
In the second comparison, I tested for differences in preference among species. I measured preference for four colonies of P. occidentalis and six colonies of each of the other seven species. I conducted all preference assays during the peak period of diurnal foraging for each species.
Statistical analyses
I tested for differences in preference using a one-way multivariate analysis of variance (MANOVA). To control for differences in total consumption or total visitation among subjects, I converted mass loss to proportion of total food collected and counts to proportion of total visitation. Because the components must sum to unity, converting values to proportions produces a linear constraint, which causes the sample covariance matrix to be singular. To create an invertible covariance matrix required for analysis, I simply eliminated one of the components (Lockwood, 1998).
Within groups, I tested for preferences among food types by determining if the proportional values for each food type differed significantly from a null model of no choice (i.e., one-third for each of the three food types). Using MANOVA, I compared the matrix of sample means to a D matrix containing values of one-third.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the supplementation experiment, Dorymyrmex smithi colonies showed a stronger preference for carbohydrates after receiving protein. Preference differed significantly between casein-supplemented colonies and water-supplemented colonies (Figure 2; F2,9 = 12.568, p =.002). Water-supplemented colonies collected significantly less of the sucrose solution than predicted by the null model of no choice (F1,5 = 10.631, p =.022), indicating a preference for casein. Water-supplemented colonies also collected significantly more of the mixed solution than expected (F1,5 = 30.937, p =.003). In contrast, casein-supplemented colonies showed a preference for sugar: they collected significantly more of the pure sucrose solution (F1,5 = 155.885, p <.005) and significantly less of the pure casein solution (F1,5 = 26.096, p =.004) than predicted by the no-choice model.
|
The interspecific comparison revealed substantial differences in preference among species (Figure 3; F14,74 = 10.536, p <.0005). Four species (Pogonomyrmex barbatus, Pogonomyrmex occidentalis, Pheidole diversipilosa, and Aphaenogaster cockerelli) showed a strong preference for sucrose: ant visitation patterns in these species were biased toward the pure sucrose food and away from the pure casein food (Table 2). In contrast, Formica perpilosa collected more of the pure casein and less of the pure sugar food than predicted by the no choice model (Table 2). Three species (D. smithi, Forelius sp. 1, and Myrmecocystus mendax) preferred the mixture (Table 2), although the amount of the mixture collected in F. sp. 1 was only marginally greater than one-third of the total amount collected (F1,5 = 5.691, p =.063). Preference also differed among species within the same habitat, both in the desert (F6,38 = 12.258, p <.0005) and in the oak woodlands (F6,34 = 7.079, p <.0005).
|
|
Diet category had a significant effect on preference (Figure 4; F2,5 = 12.980, p =.010). As a group, the seed harvesters and generalist scavengers showed a strong preference for sucrose. On average, 63.6% of these ants counted on the baits were collecting the pure sucrose food, and only 9.4% were feeding from the pure casein food. Both of these percentages are significantly different from the no-choice model (sucrose: F1,3 = 21.283, p =.019; casein: F1,3 = 76.523, p =.003). In contrast, species that feed on carbohydrate-rich exudates collected less of the sugar solution than predicted by the no-choice model (F1,3 = 10.598, p =.047).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Supplementation experiment
Results of the supplementation experiment can be explained in terms of nutritional benefits and foraging costs. First, casein-supplemented colonies may have collected more sucrose to obtain the nutritional benefits of a mixed diet. The balance of carbohydrates and protein is a key determinant of diet quality for insects (Dadd, 1985
), and a variety of insects have demonstrated the ability to select nutritionally favorable mixtures from foods that differ in carbohydrate and protein content (Waldbauer and Friedman, 1991). Colonies may have selected more sucrose because they needed sucrose to couple with the casein that they had acquired during the supplementation period. At the same time, casein-supplemented colonies may have collected proportionally more sucrose because of differences in foraging opportunity costs. After the supplementation period, D. smithi workers perceived casein to be relatively more available (Kay, 2002). When a forager simultaneously encounters two resources that contain different nutrients but are otherwise identical, the item containing the rarer substance is more valuable as a forage item because fewer opportunities exist for collecting it elsewhere in the environment. These explanations correspond to the factors that interact to determine optimal diet choice in the microeconomic model. The nutritional interpretation refers to the slope of the isocline, which in Figure 1 becomes steeper at the optimal solution after supplementation. A steeper isocline indicates that protein is more valuable to consumer nutrition relative to carbohydrates. The foraging interpretation refers to the slope of the constraint curve, which also becomes steeper at the optimal solution after supplementation. A steeper constraint curve indicates that protein is more available relative to carbohydrates. These explanations are each consistent with the model because the slopes of these curves are equal at the optimal solution.
How ant colonies are able to balance nutrient intake is not known. Solitary insects may use nutrient-specific feedbacks to maintain nutrient balance (Simpson and Raubenheimer, 1996): one mechanism involves changes in the sensitivity of taste receptors in response to an animal's recent nutritional history (Simpson et al., 1991). An analogous mechanism operating at the level of the colony could have allowed D. smithi to collect proportionally more sugar after casein supplementation. After supplementation, D. smithi workers still eagerly accepted each experimental solution: drops of even 15% casein (the least consumed option in the preference test) were accepted by more than 95% of workers in these colonies (Kay, 2002). Thus, differences in the acceptability of solutions to workers cannot explain the shift in colony-level preference of casein-supplemented colonies. Instead, this shift likely resulted from differential recruitment to baits. For example, workers from casein-supplemented colonies may have released more recruitment pheromone after feeding on sugar or less pheromone after feeding on casein. Elucidating the mechanisms that social insects use to balance resource intake will be important for understanding the overall allocation of colony foraging labor (Seeley, 1995) and may provide an insightful contrast to investigations of nutrient balancing in solitary organisms.
Interspecific comparison
Results of the second part of this study suggest that the microeconomic model may also explain some interspecific differences in foraging behavior. Preference among the sucrose, casein, and mixed foods differed significantly among species, indicating that in general the feeding decisions of ants do not maximize any single resource currency. Differences in preference were instead consistent with the model, as species generally preferred the nutrient to which they had less access.
The link between nutrient access and preference may instead result from other characteristics that species in each diet category share due to their common ancestry. Species within each category are more closely related to one another than they are to species in the other category (species that do not collect exudates are in the subfamily Myrmecinae, while species that do collect exudates are in two closely related subfamilies, the Dolichoderinae and the Formicinae), and there is no evidence to suggest that exudate feeding evolved more than once in these species. As a result, the comparison of preference between these two groups must be interpreted cautiously. It demonstrates only that the species that feed on exudates and have greater access to carbohydrates relative to protein select higher protein:carbohydrate ratios. To test whether the evolution of exudate feeding results in a stronger preference for protein over carbohydrate, other phylogenetically independent contrasts must be made.
The comparisons among species within the same habitat suggest that nutrient availability must be quantified before differences in food selection among consumers can be understood. Researchers often infer the availability of resources from a measure of their abundance in the habitat (Hutto, 1990; Johnson, 1980). However, resource availability is also influenced by other extrinsic factors such as the intensity of competition and the threat of predation and intrinsic features of a consumer such as its ability to locate, capture, and transport prey items (Wainwright, 1994; Wiens, 1984). Differences in preference both among species in the oak woodlands and among those in the desert suggest that the distinction between abundance and availability is important. In both habitats, carbohydrates are more available and protein is relatively rarer for exudate feeders in part because workers in these species can drink much faster than can workers in the other species in this study (A. Kay, unpublished manuscript). Somewhat paradoxically, specializations that allow colonies to exploit carbohydrates effectively may increase the value of (and their preference for) other nutrients such as protein. The within-habitat differences suggest that assessing the impact of nutrition on food choice may hinge on the accuracy with which resource availability is quantified.
Factors other than nutrient availability may also have influenced patterns of bait visitation in this study. First, species-specific foraging characteristics may have affected the precision of colony-level selection. For example, Forelius sp. 1 uses an effective mass recruitment system: as in other Forelius species (Hölldobler, 1982), chemical trails laid down by successful foragers can rapidly recruit hundreds of workers from the nest to a food source. This system allowed F. sp. 1 to rapidly locate each bait, but the large recruitment response occasionally resulted in heavy exploitation of the bait that was located first, regardless of its content. Although I reduced the impact of this effect by changing the position of baits after the initial discovery, this initial surge of recruits likely affected the final results. In contrast, weak recruitment signals may have caused imprecise colony-level selection in Aphaenogaster cockerelli. A. cockerelli scouts use short-lived chemical trails, vibrational signals, and aerially released secretions that can effectively recruit nearby workers but are not able to channel large numbers of workers to distant food finds (Hölldobler et al., 1978; Markl and Hölldobler, 1978). Because of the relative weakness of these signals, differential recruitment may have had less of a role in generating visitation patterns in this species.
Implications
Results from this study have several implications for the analysis of foraging behavior. First, a forager's choice among nutritionally diverse food items will depend on the availability of items elsewhere in the environment. Thus, researchers will not be able to determine if foraging decisions are optimal without quantifying the nutritional context within which the decisions are made. In contrast, energetic models predict invariant preference among simultaneously encountered items given the intrinsic characteristics of the encounter—that is, the relative profitability of the items and the encounter rate (Stephens and Krebs, 1986). In addition, this study suggests that variation in the relative availabilities of nutrients in a forager's environment can generate differences in preference both within and among species.
The results of this study also suggest that, in the wild, a forager's preference among nutritionally diverse foods will not reveal its physiological requirements. In laboratory experiments, a variety of insects can choose a nutritionally optimal mixture from diverse foods (Waldbauer and Friedman, 1991). Thus, diet choice can often serve to assay nutritional needs (Simpson and Raubenheimer, 1995). However, foragers in the wild will prefer foods rich in rare nutrients because these foods are more difficult to acquire elsewhere, and they can be mixed with more common nutrients to improve nutrition. These choices do not reveal the composition of the nutritionally optimal diet for a consumer; they merely suggest what the greatest impediment is to reaching that diet.
|
ACKNOWLEDGEMENTS |
|---|
I thank Fred Adler, Lissy Coley, Denise Dearing, Don Feener, and Tom Kursar for their comments on the manuscript. I also thank the staff at the Southwestern Research Station for all of their help. This work was funded by a grant from the Theodore Roosevelt Memorial Fund of the American Museum of Natural History and a Graduate Research Fellowship from the University of Utah.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Bristow CM, Yanity E, 1999. Seasonal response of workers of the Allegheny mound ant, Formica exsectoides (Hymenoptera: Formicidae) to artificial honeydews of varying nutritional content. Great Lakes Entomol 32:15-27.
Brodbeck BV, Mizell RFI, French WJ, Anderson PC, Aldrich JH, 1990. Amino acids as determinants of host preference for the xylem feeding leafhopper, Homalodisca coagulata (Homoptera: Cicadellidae). Oecologia 83:338-345.
Cole AC, 1968. Pogonomyrmex harvester ants: a study of the genus in North America. Knoxville: University of Tennessee Press.
Cottam DA, 1985. Frequency-dependent grazing by slugs and grasshoppers. J Ecol 73:925-933.
Covich A, 1972. Ecological economies of seed consumption by Peromyscus—a graphical model of resource substitution. Trans Conn Acad Arts Sci 44:71-93.
Crist TO, MacMahon JA, 1992. Harvester ant foraging and shrub-steepe seeds: interactions of seed resources and seed use. Ecology 73:1768-1779.
Dadd RH, 1985. Nutrition: organisms. In: Comprehensive insect physiology, biochemistry and pharmacology. (Kerkut GA, Gilbert LI, eds). Oxford: Pergamon Press; 364–437.
Dearing MD, Schall JJ, 1992. Testing models of optimal diet assembly by the generalist herbivorous lizard Cnemidophorus murinus. Ecology 73:845-858.[CrossRef]
Evans JD, Pierce NE, 1995. Effects of diet quality and queen number on growth in leptothoracine ant colonies (Hymenoptera: Formicidae). J NY Entomol Soc 103:91-99.
Folgarait PJ, Dyer LA, Marquis RJ, Braker HE, 1996. Leaf-cutting ant preferences for five native tropical plantation tree species growing under different light conditions. Entomol Exp Appl 80:521-530.
Greenstone MH, 1979. Spider feeding behavior optimises dietary essential amino acids. Nature 282:501-503.
Hjälten J, 1997. Willow hybrids and herbivory: A test of hypotheses of phytophage response to hybrid plants using the generalist leaf-feeder Lochmaea caprea (Chrysomelidae). Oecologia 109:571-574.[CrossRef]
Holder Bailey K, Polis GA, 1987. Optimal and central-place foraging theory applied to a desert harvester ant, Pogonomyrmex californicus. Oecologia 72:440-448.
Hölldobler B, 1982. Interference strategy of Iridomyrmex pruinosum (Hymenoptera: Formicidae) during foraging. Oecologia 52:208-213.
Hölldobler B, Wilson EO, 1990. The ants. Cambridge, Massachusetts: Harvard University Press.
Hölldobler B, Stanton RC, Markl H, 1978. Recruitment and food-retrieving behavior in Novomessor (Formicidae, Hymenoptera). I: Chemical signals. Behav Ecol Sociobiol 4:163-181.[CrossRef]
Howard JJ, 1987. Leafcutting ant diet selection: the effect of nutrients, water, and secondary chemistry. Ecology 67:503-515.
Hutto RL, 1990. Measuring the availability of food resources. In: Avian foraging: theory, methodology, and applications (Morrison ML, Ralph CJ, Verad J, Jehl JR Jr, eds). Lawrence, Kansas: Allen Press; 20–28.
Johnson DH, 1980. The comparison of usage and availability measurements for evaluating resource preference. Ecology 61:65-71.[CrossRef]
Kay A, 2002. Applying optimal foraging theory to assess nutrient availability ratios for ants. Ecology 83:1935-1944.
Kronfeld SN, Dayar T, 1999. The dietary basis for temporal partitioning: Food habits of coexisting Acomys species. Oecologia 121:123-128.[CrossRef]
Lanza J, Vargo EL, Pulim S, Chang YZ, 1993. Preferences of the fire ants Solenopsis invicta and S. geminata (Hymenoptera: Formicidae) for amino acid and sugar components of extrafloral nectars. Environ Entomol 22:411-417.
Lockwood JRI, 1998. On the statistical analysis of multiple-feeding preference experiments. Oecologia 116:475-481.[CrossRef]
Markl H, Hölldobler B, 1978. Recruitment and food-retrieving behavior in Novomessor (Formicidae, Hymenoptera). II. Vibrational signals. Behav Ecol Sociobiol 4:183-216.[CrossRef]
Pennings SC, Nadeau MT, Paul VJ, 1993. Selectivity and growth of the generalist herbivore Dolabella auricularia feeding upon complementary resources. Ecology 74:879-890.[CrossRef]
Rank NE, 1992. Host plant preference based on salicylate chemistry in a willow leaf beetle (Chrysomela aeneicollis). Oecologia 90:95-101.
Rapport DJ, 1971. An optimization model of food selection. Am Nat 105:575-587.[CrossRef]
Raubenheimer D, Simpson SJ, 1997. Integrative models of nutritional balancing: application to insects and vertebrates. Nutr Res Rev 10:151-179.[CrossRef]
Sanchez-Vazquez FJ, Yamamoto T, Akiyama T, Madrid JA, Tabata M, 1999. Macronutrient self-selection through demand-feeders in rainbow trout. Physiol Behav 66:45-51.[Medline]
Savolainen R, 1991. Interference by wood ant influences size selection and retrieval rate of prey by Formica fusca. Behav Ecol Sociobiol 28:1-8.
Schoener TW, 1971. Theory of foraging strategies. Annu Rev Ecol Syst 2:369-404.[CrossRef]
Schumacher A, Whitford WG, 1974. The foraging ecology of two species of Chihuahuan desert ants: Formica perpilosa and Trachymyrmex smithi neomexicanus (Hymenoptera: Formicidae). Insect Soc 21:317-330.
Seeley TD, 1995. The wisdom of the hive: the social physiology of honey bee colonies. Cambridge, Massachusetts: Harvard University Press.
Simpson SJ, James S, Simmonds MSJ, Blaney WM, 1991. Variation in chemosensitivity and the control of dietary selection behaviour in the locust. Appetite 17:141-154.[CrossRef][ISI][Medline]
Simpson SJ, Raubenheimer D, 1996. Feeding behaviour, sensory physiology and nutrient feedback: a unifying model. Entomol Exp Appl 80:55-64.[CrossRef]
Simpson SJ, Raubenheimer D, 1995. The geometric analysis of feeding and nutrition: a user's guide. J Insect Physiol 41:545-553.[CrossRef]
Snelling RR, 1976. A revision of the honey ants, genus Myrmecocystus (Hymenoptera: Formicidae). Sci Bull Nat Hist Mus LA County 24:1-163.
Stephens DW, Krebs JR, 1986. Foraging theory. Princeton, New Jersey: Princeton University Press.
Taylor F, 1977. Foraging behavior of ants: experiments with two species of myrmecine ants. Behav Ecol Sociobiol 2:147-167.
Tilman D, 1982. Resource competition and community structure. Princeton, New Jersey: Princeton University Press.
Voelkl W, Woodring J, Fisher M, Lorenz MW, Hoffmann KH, 1999. Ant-aphid mutualisms: the impact of honeydew production and honeydew sugar composition on ant preferences. Oecologia 118:483-491.[CrossRef]
Wainwright PC, 1994. Functional morphology as a tool in ecological research. In: Ecological morphology: integrative organismal biology (Wainwright PC, Reilly SM, eds). Chicago: University of Chicago Press; 42–59.
Waldbauer GP, Friedman S, 1991. Self-selection of optimal diets by insects. Annu Rev Entomol 36:43-63.[CrossRef]
Whitford WG, Depree E, Johnson P, 1980. Foraging ecology of two Chihuahuan desert ants: Novomessor cockerelli and Novomessor albisetosus. Insect Soc 27:148-156.
Wiens JA, 1984. Resource systems, populations, and communities. In: A new ecology: novel approaches to interactive systems (Price PW, Slodobchikoff CN, Gard WS, eds). New York: John Wiley and Sons; 397–436.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Dussutour and S. J. Simpson Carbohydrate regulation in relation to colony growth in ants J. Exp. Biol., July 15, 2008; 211(14): 2224 - 2232. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||