Seed preference in the Central American Agouti, Dasyprocta punctata Kristen Coogan Department of Wildlife and Conservation Biology, University of Rhode Island ABSTRACT Scatter hoarding animals in the tropics typically depend on a cached supply of reserv es when food is scarce. The Optimal Foraging Theory seeks to explain the behaviors these organisms exhibit upon prey selection. It is important for these scatter hoarding animals to choose the most beneficial food item that will increase their energy int ake. Central American agouties Dasyprocta puncata were presented with two differently sized seed species, the water apple Syzygium malaccense and the peach palm Bactris gasipaes . A total of 400 flagged seeds were used, 200 of each species in groups of 20 seeds per species, over a 14 day period in Monteverde, Costa Rica. The seeds were collected and observed to determine seed fate. There was a significant difference in seed weight between the two species T test, p = <0.0001 and a significant diff erence between which species D. punctata chose to cache more Ã°c 2 , p = <0.05, although there was no significant difference between which seed species D. punctata chose to eat Ã°c 2 , p=>0.05. Although D. punctata chose to cache more of the larger seed spec ies, size and weight alone may not be the only measure for seed selection, therefore D. punctata may exhibit other selection processes in order to obtain its energy intake. RESUMEN Los animales dispersores y acaparadores tropicales dependen tÃpicamente d el suministro oculto de reservas cuando el alimento es escaso. La TeorÃa de Forrajeo "ptimo procura explicar las conductas que estos organismos exhiben en la selecciÃ³n de presa. Es importante para estos animales dispersores y acaparadores escoger el artÃc ulo de alimento mÃ¡s beneficioso que aumentarÃ¡ su toma de energÃa. Se presentaron dos especies de semillas de diferentes tamaÃ±os, la manzana de agua Syzygium malaccense y el pejibaye Bactris gasipaes a la guatusa centroamericana Dasyprocta punctata . Se utilizÃ³ un total de 400 semillas marcadas; 200 de cada especie en grupos de 20 semillas por especie en un perÃodo de 14 dÃas en Monteverde, Costa Rica. Las semillas se colectaron y fueron observadas para determinar el destino de cada una. Se encontrÃ³ un a diferencia significativa en el peso de las semillas entre las dos especies prueba de t, p = <0.0001 y una diferencia significativa entre cuÃ¡l especie D. punctata enterrÃ³ mÃ¡s frecuentemente Ã°c 2 , P = <0.05. Sin embargo, no hubo diferencias significati vas entre las especies que D. punctata prefiriÃ³ consumir Ã°c 2 , P = >0.05. Aunque D. punctata escogiera enterrar mÃ¡s semillas de la especie mÃ¡s grande, el tamaÃ±o y el peso por si solos no pueden ser los Ãºnicos criterios para la selecciÃ³n de la semilla; por lo tanto, puede que D. punctata utilice otros procesos de selecciÃ³n para obtener su energÃa. INTRODUCTION The goal of every organism in their lifetime is to survive and reproduce. Natural selection determines what phenotypes will be represented in future generations based on adaptation to selective pressures in the surrounding environments Osborne 2000.
Organisms must evaluate resource availability, foraging techniques, and survival probability in order to guarantee their survival and fitness. The Op timal Foraging Theory states that an organism will choose a feeding method that will maximize their net rate of energy intake. The net energy is measured as the amount of energy in the food minus the energy expended in finding, investing, and digesting it Osborne 2000. The Optimal Foraging theory seeks to explain certain behaviors exhibited by organisms upon their diet selection. Selection of prey is made by evaluating both the positive and negative outcomes, and selection may be influenced by prey den sity, prey size, and nutrient content. Organisms will generally discriminate against food items which will not optimize their survival or fitness and choose those that will be most beneficial to them. Many animals hoard food from local abundance in orde r to conserve it for future use. In the case of a scatter hoarding organism, small amounts of food are dispersed among many spatially spaced caches Jansen et al. 2002. These individuals need to evaluate their food quality, in order to receive the highe st net energy gain for the present time, as well as in the future when retrieving their cached reserves. It is beneficial for organisms which exhibit the scatter hoarding strategy to manage a small supply of large seeds rather than a large supply of small , less nutritious seeds Jansen et al. 2002. Seed fate is also important in determining the value of the food source. Scatter hoarding animals typically eat few seeds while caching others. The seeds they store are presumably more valuable than the seed s they are consuming in order for them to be more beneficial in times of scarcity Hallwachs 1994. The Central American Agouti, Dasyprocta punctata a scatter hoarding organism. This species is a ground dwelling rodent, mostly active throughout the d ay, which can withstand disturbed habitats Wainwright 2002. The diet of D. punctata consists mainly of fruits and seeds that are high in nutrients. Unlike the equally common seminivore frugivore related species in the area, Agouti paca, D. punctata is unable to store fat. Their solution to seasonal scarcity is to bury seeds in times of plenty and dig them up later Smythe 1978. Agoutis are therefore classified as a scatter hoarding rodent that usually caches a surplus of seeds in a dispersed pattern d ue to the uncertainty of retrieval Smythe 1978. Seeds that have not been recovered have a possibility of germination, thereby also making D. punctata an important disperser for many plants. The aim of this study was to examine what preferences D. punct ata exhibited for seed size. Smaller seeds were expected to be eaten due to their less nutritional value, while larger seeds were expected to be cached in order to benefit the agouti in times of scarcity. METHODS This experiment took place from April 2 3 rd through May 7 th , 2006 at the Santuario EcolÃ³gico located in Cerro Plano, Costa Rica, during the dry season Fig. 1. The Santuario EcolÃ³gico is a secondary, lowland forest that was previously used for banana production and agouties are often observed foraging in a patch of open understory, approximately six meters by two meters in size. Two morphologically different tree seed species were used in choice experiments to determine whether D. punctata commonly took larger seeds over smaller seeds. The tw o species chosen for this study were Syzygium malaccense and Bactris gasipaes . The
value of the seeds picked were placed under three categories based on their fate. The seed fate categories included being eaten, cached, or un recovered. S. malaccense an d B. gasipaes were purchased from a vegetable and fruit stand. The fruit was split with a knife; seeds were removed and then washed and dried. In order to locate the seeds in the field, each seed needed to be marked with flagging tape. Holes were carefu lly drilled in the center of every seed, and a meter of test nylon fishing line was threaded through the seed and tied. Twelve centimeters of flagging tape were then attached to the fishing line. Seeds were weighed in grams and then each tape was recorde d with the seed number and weight. D. punctata bury the seeds not the flagging tape, so cached seeds were easily found. Previous studies have shown that marking seeds in this manner does not alter disperser behavior; however, this cannot be discounted com pletely Jansen et al. 2002. Twenty seeds of each species were placed in two separate piles at the site, side by side Fig. 2. Seeds were placed at 9 a.m. and checked the following morning at 8 a.m. The fates of the seeds were classified into one of t he three categories, those being eaten, cached, or lost. Eaten seeds were those with the entire embryo consumed, or if the flagging tape and fishing line were removed from the seed. Cached seeds were those that were removed from the original site and bur ied shallowly in the nearby soil. Seeds whose flagging was not found were classified as lost. A new supply of seeds was then offered, ensuring that there were twenty of each seed species presented at all times. A t test was used to determine weight diffe rence in seed species. A Chi squared test was also conducted in order to test if there was any significant difference in the fate of the seeds. RESULTS A total of 400 seeds were presented to D. punctata , 200 of each seed species. One hundred and eighty one seeds were found and collected for data. Results of the S. malaccense seeds found, 148 seeds were eaten and 43 were cached. One hundred and seventy one of the B. gasipaes seeds were eaten, while 19 were cached Table 1. The average weight for S. ma laccense seeds was 10.78 g while the average weight for B. gasipaes seeds was 3.77 g Fig. 3. An unpaired t test showed that there was a significant difference in weight between the two chosen seed species t = 15.70, p < 0.0001, df = 398. S. malaccen se was the larger seed species used, while B. gasipaes was the smaller species. A Chi squared test was used to determine if there were significant differences between the numbers of eaten or cached seeds between both species Fig. 4. There was no signif icant difference in the number of seeds eaten between the S. malaccense and B. gasipaes Chi squared = 1.65, p >0.05, df = 1. Conversely, there was a significant difference in the numbers of cached seeds between the two species Chi squared = 9.2, p <0.0 5, df = 1. The Chi squared test found that D. punctata did selectively cache the S. malaccense seeds more often than B. gasipaes . The unpaired t test found that there was a significant size difference between the two seed species; S. malaccense being t he larger of the two species. Therefore, this study showed that D. punctata selectively chose the larger seed species to cache over the smaller species.
DISCUSSION The original prediction of this experiment was that D. punctata would selectively choose the larger seed species over the small. The larger seed species, S. malaccense , was also expected, and is shown, to be cached at a higher frequency than the smaller seed species, B. gasipaes . There were a low number of seeds cached overall; most of the se eds in this study were eaten by the agouties. The high variation of seed weight in S. malaccense shows that size and weight cannot be a good measure for seed fate. If D. punctata chose to cache more S. malaccense seeds, or conversely, eat more B. gasipae s seeds than S. malaccense , there should be a reason other than size and weight upon food selection. The Optimal Foraging Theory predicts that predators choose a diet in order to maximize the net rate of energy intake. Food size is not the only factor th at is important upon choosing prey. Nutritional values, as well as handling time and predator avoidance are also important. Predators need to evaluate their prey before consumption in order to receive the greatest energetic reward. D. punctata chose to eat both the large and small seed species showing that size and weight did not play a role upon deciding prey choice. If they chose the smaller species to eat as well, another important characteristic, such as nutritional value, could have influenced the ir decision. The larger species, S. malaccense , was generally light in weight, and had a green, soft inside. The smaller species, B. gasipaes , was denser and a good source of Vitamin A, starch, and protein Zuchowski 2005. Also, as Wainwright 2002 st ates, palms are among the favorite food items of D. punctata , B. gasipaes being one of those palms. There is a possibility that B. gasipaes is more nutritionally valuable to the agoutis, thereby making the smaller seeds an important asset to their diet, as well as perhaps being more appealing in taste. Mixed diets can also be very beneficial to an individual, especially those that rely on plant material for energy and nutrients. First, an organism may accept a low quality food simply because, having enco untered them, there may be more to gain in eating them than there would be in ignoring them. Second, a mixed diet is beneficial because different food types contain different nutrients, which may be important to the individual Begon et al. 1990. A possi ble explanation as to why the agoutis did not cache more seeds overall could be due to the season in which this study took place. This study was conducted at the end of April and early May, during the dry season. Since fruits and seeds are usually scarce during the dry season, many scatter hoarding animals that remain frugivorous throughout the year, such as agouties and their allied species, the acouchies Myoprocta spp., live off of cached reserves that they collected when fruits and seeds were plentifu l, such as during the wet season Forget et al. 2002. Studies done in Panama have shown that during times of food abundance, agouties hoard more seeds than what they consume, and when fruit begins to become scarce, they rely on those hoarded seeds Forge t et al.2002. In this case, the reason why the D. punctata did not scatter more seeds overall, could have been because the dry season is not normally when caching takes place, however, it is not known whether scatter hoarding animals, such as D. punctata , show a triggered response in relation so seed fluctuations.
If S. malaccense seeds are nutritionally less beneficial to D. punctata , then one may wonder why D. punctata would have hoarded the seeds away for a later time at all. The reason for caching an y seeds could be to act as insurance. If fruiting numbers were significantly lower one year, or did not occur due to abiotic factors such as El NiÃ±o, there would be no resources for the agouties to consume; therefore, having insurance is important to ensu re survival. The agouties may then have cached the less nutritious seeds S. malaccense as opposed to the more beneficial seeds merely to act as insurance. It is also not certain that D. punctata will recover all the seeds they cached. Thus, the agouti es may eat the more nutritious seeds B. gasipaes , as opposed to caching them in order to ensure theyÂ€ll receive the nutritional benefit instead of taking the chance of losing cached reserves. Studying caching behaviors is important in understanding the foraging strategies among many scatter hoarding animals. Future studies should include differences between seasons and caching quantities. Differences in seed nutrition should also be considered. Seed preference may be better understood by picking simil ar seed types, such as those of two palm species. ACKNOWLEDGEMENTS I would like to thank Javier MÃ©ndez for his patience and guidance throughout this study, from the beginning to the end. Alan and Karen Masters, without them I would not have the supplie s to conduct this study. I would like to thank Mireya Salazar MÃ©ndez at the Santuario EcolÃ³gico for letting me use the property. Last, but not least, I would like to thank Maria JosÃ© and Ollie Hyman for helping all of my fellow students and me, in gettin g our needed materials together, and spending many hours helping us complete our projects . LITERATURE CITED Alcock, J. 1948. Animal Behavior: An Evolutionary Approach. Sinauer Associates, Inc. Saunderland, MA. B egon, M., J.L. Harper, C.R. Townsend. 1990. Ecology: Individuals, Populations, and Communities. Blackwell Scientific Publications, M.A. Brown, L. and J.F. Downhower 1951. Analysis in Behavioral Ecology. Sinauer Associates, Inc. Saunderland, MA. Forget, P.M., D.S. Hammond, T. Milleron, and R. Thomas. 2002. Seasonality of fruiting and food hoarding rodents in Neotropical forests: consequences for seed dispersal and seedling recruitment. In D.J. Levey, W.R. Silva, and M. Galetti, eds. Seed Dispersal and Fru givory: Ecology, Evolution and Conservation. CABI Publishing. Hallwachs, W. 1986. Agoutis Dasyprocta punctata : The Inheritors of Guapinol Hymenaea courbaril: Leguminosae. In A. Estrada and T.H. Fleming, eds. Frugivores and Seed Dispersal. Dr. W. Junk P ublishers, Dordrecht. Jansen, P.A., M. Bartholomeus, F. Bongers, J.A. Elizinga, J. den Ouden, and S.E. Van
Wieren 2002. The Role of Seed Dispersal by a Scatter hoarding Rodent. In D.J. Levey, W.R. Silva, and M. Galetti, eds. Seed Dispersal and Frugivory: E cology, Evolution and Conservation. CABI Publishing. Osbourne, Patrick L. 2000. Tropical Ecosystems and Ecological Concepts. Cambridge University Press, New York, NY. Smythe, N. 1978. The natural history of the Central American Agouti Dasyprocta punctata . In D.H. Janzen Ed.Costa Rican Natural History. University of Chicago Press, Chicago. Wainwright, M. 2002. The Natural History of Costa Rican Mammals. Distribuidores Zona Tropical, S.A. Zuchowski, Willow. 2005. A Guide to Tropical Plants of Costa Ric a. Distribudores Zona Tropical, S.A.
Table 1. Number of seeds eaten, cached, or lost from the study site by D. punctata for the two seed species: Syzygium malaccensis and Bractris gasipaes . Seed Species Eaten Cached Lost S. malaccensis 148 43 9 B. gasi paes 171 19 10 Figure 1. Map of the Monteverde region in Costa Rica. Study site, Santuario EcolÃ³gico is highlighted.
Figure 2. Study site in the Santuario EcolÃ³gico in Cerro Plano, Costa Rica. Study site was situated in an open understory area behind the Information Center. XÂ€s mark the two piles of seed species. Figure 3. Average weight of the two seed species used in this study; Syzygium malaccensis and Bactris gasipaes . Differences were significant p < 0.0001. 0 2 4 6 8 10 12 14 16 18 S. malaccensis B. gasipaes Seed Species Average Weight grams
Figure 4. Fate of ea ch seed, either cached or eaten, for both seed species: Syzygium malaccense and Bactris gasipaes . There was no significant difference for numbers of eaten seeds between the two species, but numbers of cached seeds between the seeds species were significan tly different Ã°c 2 = 9.2, p < 0.05, df = 1. 0 20 40 60 80 100 120 140 160 180 Eaten Cached Seed Fate Number of Seeds S. malaccense B. gasipaes