The role of L-dopa in Mucuna fruits: a deterrent or a marker Jeff Dale Department of Biology, University of Missouri ABSTRACT Agoutis (Dasyprocta punctata) are ubiquitous in Mon teverde and are known dispersers of Mucuna urens seeds despite the high levels of L-dopa the seeds c ontain. I tested the effects of L-dopa added to pej ibaye (Bactris gasipaes) on seed selection by D. punctata I hypothesized that L-dopa acts more as a marker for agouti dispersal than a deterrent for insect predat ion. In five sites of known agouti activity, I plac ed three types of seeds (i.e. those with natural L-dopa, syn thetic L-dopa, and with neither) in piles of ten approximately one meter apart. My primary data were sequences of seed removal patterns. However, the total number of seeds removed was also recorded. T he results indicate randomness in term of the sequences of seed removal. However, total seed coun t data and comparison of first and second choice da ta showed a non-random pattern. Specifically, there is for seeds with L-dopa added, and there is a signif icant trend to removing the same seed type following the first. The latter two results demonstrated a tenden cy for habituation in agouti behavior. Based on the result s, I hypothesized that the reason for habit formati on was the L-dopa consumption. A more expansive study of s equence data that evaluates seeds eaten vs. seeds cached could serve to answer a lot of the questions left open by my analyses. RESUMEN Las guatusas ( Dasyprocta punctata ) son ubicuos en Monteverde y se sabe son los disp ersores de las semillas de Mucuna urens a pesar de los niveles de la L-dopa que las semillas contienen. Prob los efe ctos de la L-dopa aadidos al pejibaye ( Bactris gasipaes ) sobre la seleccin de la semilla por D. Punctata. Presum que la L-dopa acta ms como marcador para la dispersin del agouti que un impedimento para la depredacin por insecto. En cinco sitios de activid ad conocida de la guatusa, puse tres tipos de las s emillas (es decir esas con la L-dopa natural, L-dopa sintt ica, y con ningun tratamiento) en grupos de diez separadas aproximadamente por metro. Mis datos prim arios fueron secuencias de patrones del remocin de la semilla. Sin embargo, el nmero total de semilla s removidas tambin fue registrado. Los resultados indican aleatoriedad en trmino de las secuencia de retiro de la semilla. Sin embargo, los datos del n mero de semillas y la comparacin total de los primeros y segundos datos demostraron un patrn no aleatorio Especficamente, hay para las semillas con la L-dop a agregada, una tendencia significativa a quitar el mismo tipo de la semilla que sigue la primera. Los ltimos dos resultados demostraron una tendencia pa ra la habituacin en el comportamiento de las guatusas De acuerdo con los resultados, presum que la raz n de la formacin del hbito era el consuma de la L-d opa. Un estudio ms expansivo de los datos de la secuencia que evalan las semillas comidas contra l as semillas depositadas podra servir para contesta r muchas las preguntas dejadas abiertas por mi anlis is. INTRODUCTION Co-evolution of plant-species interactions such as seed dispersal is vital to the survival of many different organisms. For example, interactions between Agaonidae wasps and Ficus trees are extremely specific and with growing habit at loss this interaction becomes more dependant on larger migration ranges of Agaonidae species for survival (Mawdsley et al.
1998). Even though plant-disperser relationships ar e drastically different from plantpollinator relationships, this example shows that e very relationship needs to be tweaked to accommodate a changing world. And from this I hy pothesized that L-dopa in Mucuna urens has further co-evolved with Dasyprocta punctata in order to increase its fitness. Mucuna urens also known as ox eye, belongs to the family Fabac eae, subfamily, Papilionoideae. It contains large golden-brown podshaped fruits. Each pod houses three hard, black seeds that contain high levels of L-dop a ranging from 5.9 to 9 % (Janzen et al. 1973). The L-dopa in the M. urens is thought to limit seed predation and decrease in sect infestations in the fruits. Mucuna urens is primarily found at the lower edges of the Monteverde Cloud Forest Reserve located near the ed ges of rivers (Zuchowski 2005). It is observed that D. punctata, or agouti, disperses M. urens seeds (Janzen 1983). Dasyprocta punctata is a diurnal rodent encountered anywhere from the low to middle elevation rain forests. Agoutis are territorial and normally make homes in small areas of two hectares. They live in monogamous pairs but usu ally forage alone. Agoutis are seedeaters with teeth that are easily capable of pe netrating hard-shelled seeds. Therefore, their favorite foods are palms seeds, especially pe jibaye ( Bactris gasipaes) seeds (Wainwright 2002). Despite the heavy levels of L-dopa that are toxic t o most potential predators, agoutis play a key role in the seed dispersal of M. urens seeds at Monteverde. Presently, the consensus is that the high levels of L-dopa are used as a chemical barrier to prevent insect attacks as shown in the study conducted by J anzen et al. (1973). Along these same lines, it is hypothesized that agoutis have develop ed a resistance to this toxicity allowing them to prey on and ultimately disperse M. urens seeds. However, few studies have been conducted and this idea is not well tested. I teste d the hypothesis that for agoutis, unlike insects, L-dopa in M. urens seeds acts not as a defense mechanism but as a mar ker for seed recognition and an attractant to promote dispe rsal. Based on data from an experiment conducted in CIEE Fall of 1997, I predicted that for the agouti the L-dopa content in the M. urens seeds would act as a marker for seed recognition and dispersal. The previous analy sis showed a conditioned behavioral preference for the Mucuna seeds after they were consumed (Franklin 1997). Re cent studies have shown that dopamine, the neurotransmit ter from which L-dopa is derived, is associated with the cognition that occurs in the an ticipation of rewards. L-dopa acts as an agent to penetrate the blood brain barrier, then co nverts to dopamine, which stimulates specific mental processes (Tran et al. 2005). If th is same process is present in agoutis, it can explain their preference for M. urens seeds over other seeds. MATERIALS AND METHODS I used 250 ripe pejibaye seeds obtained by cutting the seeds from the ripe fruit. Pejibaye seeds were chosen because agoutis are known seed pr edators of these seeds. Eighty of the seeds were treated with synthetic L-d opa by diluting four crushed tablets of pharmaceutical L-dopa (200/50 mg levodop a/carbidopa respectively) in one liter of water. These seeds were soaked for 24 hour s before use. I prepared the natural Ldopa seeds by soaking 60 Mucuna seeds in one liter of water for 24 hours. Then, I removed the Mucuna seeds and allowed 80 pejibaye seeds to soak in the solution of natural L-dopa that was created from soaking the Mucuna seeds in one liter of water
(Balaban et al. 2002). Theses methods created a 4/3 ratio of pejibaye to Mucuna seeds. Finally, 90 pejibaye seeds were immersed in one lit er of water. All seeds were left in the same solutions of water as during preparations for two weeks (the duration of the research). All seeds soaked for the same amount of time ensuring that softness was not a factor in agouti seed selection. Sites with high agouti activity were chosen study s ites. Observations were conducted from July to August. At sites with lower agouti activity, I left the seeds overnight hoping count the remaining seeds the next day and obtain viable data. For example, if 0 synthetic and natural L-dopa seeds re mained, but 7 plain seeds remained, that was viable Â“amount remainingÂ” data. At each si te, I placed all three seed types in piles of ten in a random order approximately one me ter apart. This ensured that all three seed types were equally obvious and available for a goutis to choose. I observed a total of five sites two times each, but in some cases no dat a was collected. The sites usually occurred near composts (three sites). During observ ations, I hid in a place that was out of view of the agoutis to observe their behavior. I observed a total of five sites two times each, an d watched a total of five agoutis. For each, I collected sequence data which consisted of the order in which the seeds were taken. At two sites, multiple agoutis were simultan eously seen, and I recorded the sequence for each. Any time I missed a seed that wa s taken, I went to the piles to check the numbers. I then compared those numbers to the s equence to determine which seed was taken. I used this same procedure whenever I th ought multiple seeds were taken during the same trip. In order to distinguish betwe en the three piles, I put markers (branches) behind them and determined that for exam ple, the agouti was in front of the synthetic L-dopa marker whenever a seed was taken. RESULTS Analysis of total seed removal includes four out of the 5 test sites. One site was discarded because no data were obtained. When looking at the total seed removal, there is a nonrandom pattern of seed removal based on a Chi squar ed test for goodness of fit (Fig 1, df = 2, p < 0.05). There is a strong indication that n atural L-dopa seeds were removed more often than by chance alone. In the other two cases the removal occurs randomly. In five out of the seven trials where seeds were taken, all 30 seeds placed were removed.
n rrr n "#r FIGURE 1. Comparison of expected values for seed re moval compared with observed values. Natural L-dopa, synthetic L-dopa, and no Ldopa are based on a total of 90, 70, and 90 seeds respectively. I also determined that there is a strong correlatio n between first choice of seed and second choice of seed. In six of the seven cases, t he first choice was the same as the second choice. Based on a binomial probability test there is a 6.9% chance that these events are caused by random occurrence. In all case s (N=5) where the first seed contained L-dopa the second seed removed was an ide ntical type and 50% (N=2) of the plain seeds experienced the same correlation. $%%rr
FIGURE 2. Amount of times that the second seed remo ved was the same as the first seed selected. Value includes all types of seeds, Natura l (N), Synthetic (S), and plain (P) seeds, grouped into the same category. Same seed co mbinations are S-S, N-N, and P-P. It was confirmed, however, that the sequences in s eed selection occur at random. It is important to note, that sequence data were on ly collected from three of the five tested sites and that probabilities were obtained by compa ring first seed selection to subsequent seeds selections of the same type of seed (Plain or containing L-dopa). When comparing the probability of each of the sequences of four sa me-seed choices, the probabilities show that each have about the same likelihood of occurri ng. In the case of the plain seed removal, the probability of removal is 1% as compar ed to probabilities of 3%, 3.2%, and 0.62% when referring to seeds containing L-dopa. Ev en though the sequences involving removal of L-dopa-containing seeds were observed th ree separate times, only one of the occurrences was less likely than the removal of a s imilar sequence of plain seeds (Table 1). TABLE 1. Sequences of seed removal for 5 observed a goutis. N=natural L-dopa, S=synthetic L-dopa, and P=plain seeds. Bold letters indicate the frequency four consecutive seeds removals after that same type of seed has been removed. Frequencies less than four seeds did not receive consideration. Probabilities were attained using standard fractional probabilities that were adjuste d after each subsequent seed removal. Agouti 1: Trial 1: N, N,N,N,N ,N,N,N,N,N,P Trial 2: Order not known Trial 3: N,N,N,P,S,S,S,S,S,S,S,N,N,N,N,N,N,N,S,S Agouti 2: Trial 1: P, P,P,P,P ,P,P,P,P,S (At same time as trial 3 of agouti 1) Agouti 3: Trial 1: N,N,S,S,P,S,P,S,S,N,S,S,S,P,P,S,S,P,P,P,P ,P,N,N,N,N,N,N,N Trial 2: P,P,P,P,S,S,P,N,P,P,P,P,N,N,N,S,N,N,N,S,S ,S,S Agouti 4: Trial 1: S, S,S,N,N (At same time as trial 2 of agouti 3) Agouti 5: Trial 1: P,N, N,N,N,N ,N,S DISCUSSION It has been hypothesized that agoutis coevolved by developing a resistance to the L-dopa content in Mucuna seeds. I, however, hypothesized that the L-dopa in Mucuna seeds has co-evolved further and acts as a marker for predati on and dispersal thus serving the purpose of both marker and attractant, and that wit hout the presence of L-dopa agoutis would be less likely to remove Mucuna seeds. This hypothesis is not supported by my primary data Sequence data obtained by using probabilities show that it was just as likely to see a sequence of four plain seeds
removed, as it was a sequence of four L-dopa contai ning seeds indicating chance as a key factor in seed selection. However, sequence data ca n be skewed by human presence. When agoutis acknowledged my presence, there was an obvious change in behavior. As a consequence, agoutis became more tentative during s eed selection, and frequently the seeds closest to the forest (their exit point) were taken preferentially. It is critical to understand that some of these sequence results were influenced by my presence. An interesting aspect to the sequence data puzzle i s the idea of L-dopa and associative learning (Tran et al. 2005). This conce pt was not supported by my sequence data and was confounded by seed placement methods. I switched the location of seed piles each day which could have affected the first seed selection of the next day. Further studies can be performed to evaluate congruence bet ween the first seed choice during the second day of observation, at the same site, based on the previous dayÂ’s L-dopa seed location I believe sequence data have merit and need to be p ursued further. In five of the seven trials, all seeds placed were removed. Theref ore, I suggest using more seeds to eliminate Â“over-cachingÂ” thereby producing more seq uence data, and more possibilities for Â“amount remainingÂ” data, therefore, producing l arger data sets which was a major obstacle in this study. Although sequence data presented random results, to tal seed selection data showed non-randomness. Based on the Chi-squared tes t, agoutis showed a statistical preference for natural L-dopa seeds. Also the resul ts of a previous CIEE experiment conducted in Spring 2007 show a similar amount of s eeds left for low, medium, and high level L-dopa. When considering both these results, I conclude that the level of L-dopa is not the ultimate factor deciding seed preference, b ut rather its most important aspect is its chemical composition. Since synthetic L-dopa was co nsumed less frequently, I postulate that the chemical make up of synthetic L-dopa is su bstantially different from natural Ldopa and is less preferred by agoutis. There was a strong correlation between first and se cond seed choice. In six out of the seven trials first and second seed choice were the same types of seed suggesting a strong tendency towards habit formation in agouti b ehavior. I speculate that these data provide some insight on the physiological effects o f L-dopa. In five out of the six cases seen, L-dopa containing seeds were consumed indicat ing that habit formation could be a result of L-dopa consumption. However, my methods d id not include consumption vs. cache analysis, so consumption of L-dopa was not re corded. Analysis showing difference in behavioral patterns after L-dopa seed consumptio n could prove insightful for future experimentation. Furthermore, strong specificity be tween M. urens and D. punctata would increase the fitness of M. urens substantially and indicate a long history of coevolution between the two species.
ACKNOWLEDGEMENTS First I would like to acknowledge Pablo Allen Monge and Tania Chavarra for acquiring all the supplies I needed to conduct this experiment. Also I would lik e to thank Karen Masters for letting me sit in her backyard to observe agoutis and for working with me through the problems I encountered along the way. I would like to thank everyone else not mentioned at the Estacon Biolgica for letting me stay and cond uct my research there. Jos Carlos Caldern made this e xperiment possible by suggesting that I use pejibay e seeds in my experiment, so thank you. Finally, I wo uld like to thank my parents for paying for this tr ip and allowing me to conduct these observations. LITERATURE CITED Amitra, A. K. 2007. L-dopa levels in Mucuna urens and caching behavior of agoutis (Dasyprocta puncta ta. CIEE Spring 214-222 Balaban M.O and A.A. Teixeia.2002. Potential Home a nd Industrial Process Treatments to Reduce L-dopa in Mucuna Bean. In: Food and Feed from Mucuna: Current uses and the way forward. Gainesville, Fl, pp 339-351. Franklin E. 1997. Seed Dispersal of Mucunaby Agouti s ( Dasyprocta punctata ) and its possible disruption by bananas. CIEE Fall 96-108 Janzen D. H., S.S. Rehr, and P. P. Feeney. 1973. Ldopa in Legume Seeds: A Chemical Barrier to Insect Attack. Science 18: 81-82 Mawdsley N. A., S. G. Compton, and R. J. Whittaker. 1998. Population Persistence, Pollination Mutualisms, and Figs in Fragmented Tropical Landsca pes. Conservation Biology 12: 1416-1420. Smythe N. 1983. Dasyprocta punctata and Agouti Paca (Guatusa, Cherenga, Agouti, Tepezcuintle, Paca) In: Costa Rican Natural History D.H. Janzen, ed. The University of Chicago Press, Chicago, IL, pp.463-465 Tran A., R. Tamura, T. Uwano, T. Kobayashi, M. Kats uki, T. Ono, and J.Bure 2005. Dopamine D1 receptors Involved in Locomotor Activity and Accumb ens Neural Response to Prediction of Reward Associated with Place. National Academy of S ciences. pp. 2117-2122 Wainwright M. 2002. The Natural History of Costa Ri can Mammals. Distribuidores Zona Tropical, pp. 178-180. Zuchowski W. 2005. A guide to Tropical Plants of Co sta Rica. Distribuidores Zona Tropical, pp. 293
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xlink http:www.w3.org1999xlink xmlns:xsi http:www.w3.org2001XMLSchema-instance
leader 00000nas 2200000Ka 4500
controlfield tag 008 000000c19749999pautr p s 0 0eng d
datafield ind1 8 ind2 024
subfield code a M39-00014
El papel de L-dopa en las frutas Mucuna: un elemento de disuasin o de un marcador
The role of L-dopa in Mucuna fruits: a deterrent or a marker
Agoutis (Dasyprocta punctata) are ubiquitous in Monteverde and are known dispersers of Mucuna urens seeds despite the high levels of L-dopa the seeds contain. I tested the effects of L-dopa added to pejibaye (Bactris gasipaes) on seed selection by D. punctata. I hypothesized that L-dopa acts more as a marker for agouti dispersal than a deterrent for insect predation. In five sites of known agouti activity, I placed three types of seeds (i.e. those with natural L-dopa, synthetic L-dopa, and with neither) in piles of ten approximately one meter apart. My primary data were sequences of seed removal patterns. However, the total number of seeds removed was also recorded. The results indicate randomness in term of the sequences of seed removal. However, total seed count data and comparison of first and second choice data showed a non-random pattern. Specifically, there is for seeds with L-dopa added, and there is a significant trend to removing the same seed type following the first. The latter two results demonstrated a tendency for habituation in agouti behavior. Based on the results, I hypothesized that the reason for habit formation was the L-dopa consumption. A more expansive study of sequence data that evaluates seeds eaten vs. seeds cached could serve to answer a lot of the questions left open by my analyses.
Las guatusas (Dasyprocta punctata) son ubicuos en Monteverde y son conocidos dispersores de las semillas de Mucuna urens a pesar de los altos niveles de L-dopa que las semillas contienen. Prob los efectos de L-dopa aadidos al pejibaye (Bactris gasipaes) en la seleccin de la semilla por D. Punctata.
Text in English.
Central American agouti--Costa Rica--Puntarenas--Monteverde Zone
Mucuna urens--Costa Rica--Puntarenas--Monteverde Zone
Guatusa Centro Americana--Costa Rica--Puntarenas--Zona de Monteverde
Mucuna urens--Costa Rica--Puntarenas--Zona de Monteverde
Tropical Ecology 2008
Ecologa Tropical 2008
t Monteverde Institute : Tropical Ecology