Seed Dispersal and Predation of the Palm Acrocomia Aculeata
Total Page：16
File Type：pdf, Size：1020Kb
l99Bl SCARIOT:DISPERSAL AND PREDATIONIN ACROCOMIAACULEATA Principes,42(1\,1998, pp. 5-8 SeedDispersal and Predationof the PalmAcrocomia aculeata ArlrcrR Scenror CE N ARGE N/E M B M PA,C. P. 02 3 72, 70. 7 7 0-900 Brasilia, D F, Brazil Dispersal can increase the probability that a fruits are globose-spheric,weighing 20-64 g, seedwill find an adequatesite in which to estab- and measuring2S-49 X 26-49 mm. Generally, lish, thus reducing the level of seed predation there is one seed per fruit, but sometimestwo and introducing the plant to another still unoc- and even three seedsoccur. Several aspects of cupied habitat. The seedstage represents one of the reproductive biology have been discussed the periods of highest mortality in the life cycle elsewhereby Scariot (1987) and Scariot et al. of many plant speciesand substantialreduction (r99r, I995). "Tomahawk" of seedcrop due to predation may affect the dis- traps (490 X I70 X 160 mm), tribution and abundanceof plants. Mammalscan with fruits of Acrocomia aculeata as baito were be quite effective at transporting seedsand in- used to capture possible dispersers. One hun- creasingthe probability that a seedwill find an dred forty-three traps per night were used in the adequatesite. In disperser-freehabitats, seeds pasture and 454 traps per night in the gallery may lie under parent tree and rot (e.g.,Acroco- forest. Successwas the number of animals cao- mia, Janzen I9B3a). However, seed dispersers tured divided by the total number of traps used may also act as seed predators or increase the per night. Capturedanimals were releasedafter risk of predationby attracting predators. identification. Adjacent gallery forest was in- This study addressesthree issues:the possi- cluded in the study becausethere were few wild ble role of wild and domestic animals as seed animals in this region due to deforestedareas, consumersand dispersers,their possible effect and someplants of ,4. aculeata may occur in for- on seed predation, and the rates of seed preda- est edges.The role of domestic animals in dis- tion by bruchid beetles on the palm Acrocomia persalwas observedusing binoculars,during the aculeata(Jacq.) Lodd. ex Mart. in Central Brazil. day and by naked eye at night. It complements other studies on pollination, The rate and intensity of predation were de- phenology, and fruit set (Scariot et al. 199I, termined using 14 boxes. A box consisted of a 199s). woodframe (400 X 200 X 150 mm) coveredwith wire mesh. The mesh size (15-20 mm) allowed Methods insectsaccess to the inside of the box, while im- The study wasconducted on a privately owned peding the entranceof larger animals.The boxes farm, FazendaTabocao located 45 km (15"35'S, were placed in the pasture on 13 October 1985, 47"45'W) from Brasflia,BrazTl.The elevationat under the crownsof individual s of Acrocomiaa.c- this farm is 1,000 m. Mean annual rainfall is uleata. The boxes, each with 20 fruits, were 1,566 mm and the temperatureranges from l9o paired, one with intact fruits and the other with (June)to 23" C (September).,4.aculeata is an ar- fruits from which pulp was partially removedto borescent spiny palm reaching up to 16 m in mimic fruits chewed by cattle. The boxes were height and occurring as the dominant tree placed approximately0.5 m from the trunk at an species in pastures of the grass Hyparrhenia angle of l80o to each other. All the previously rufa. Flowering occurs from August to Decem- fallen fruits under the canopywere collected and ber, with the peak of open inflorescence(>507o) removedon the day that the boxeswere placed in occuffing from mid-October to mid-November the field. No oviposition was ever observed on (Scariot 1987, Scariot et al. 1991, 1995). The the fruits while they were still attachedto the in- PRINCIPES fYot.42 fructescences.On eachofthe first five daysafter (Fig. IA). The high number of eggsper fruit (up placing these boxes,and later at irregular inter- to 15) suggeststhat female bruchid beetles do vals, I recorded the number of fruits in which not discriminate among fruits with or without beetles had oviposited and the number of eggs eggs during oviposition (Fig. 1B). Cattle de- per fruit. When oviposition on the fruits was fin- stroyed some boxes and removed the fruits, re- ished, I collected the fruits and openedthem to ducing the final number of fruits in boxesto 97 check the seedsand identify the predators. chewedand 75 nonchewed.Of 101 seedsin 97 chewed frnits, 76.280 were preyed by insects, Results I9.BVowere intact, and.47owere nonviable (ab- Only a few speciesof animals were captured sentorotteno etc.). Of 75 seedsin 75 nonchewed and not all of them may consumethe fruits of ,4. fruits only I0.6Vo were preyed by insects and aculeata. The species of vertebrates captured 89.47owere intact. Thus, chewed fruits had a wereDidelphis albiaentris(opposum, 3l), Necto- significantly higher predation level than : : nxyssquarnipes (water rat, B); Cebusapella (ca- nonchewedfruits (12 77.17, df 1; P < puchin monkey, l), and Turd,ussp. (thrush, 2). 0.001). Eggsof bruchids may be washedaway by Overall capture success rate was 7.07b, being the rain. This may account for the observedde- higher in the gallery forest (8.370)than in the creasedpercentage offruits oviposited (Fig. lA) pasture(2.57o). Of the domesticanimals that fed and number of eggslayed (Fig. 1B), which after on fruits ofAcrocomiaaculeata, only adult cattle sometime increasesagain as the layed eggsac- ingestedthem, while horses,pigs, and youngcat- cumulate. the fruits, removing the epicarp tle only chewed Discussion and mesocarp. Fruits ingested by adult cattle were regurgitated at night in piles of up to 85 An advantage of seed dispersal by several fruits. Cattle regurgitated seeds mixed with speciesis an increasein the number of possible residuesof grasses,and the seedsusually had ei- habitats and the number of seeds dispersed ther portions of the epicarp and/or mesocarpat- (Gautier-Hion et al. 1985, Howe and Primack tached,or only the endocarp. 1975).Howevero due to their small size and feed- The bruchid beetles Pachymerus sp. and ing behavior, the wild animal species captured Caryobruchus acrocomiae (Pachimerinae, Pach- in this study can be consideredonly as possible imerinii) preyed on seeds of A. aculeata. The dispersers.As a result of environmentaldistur- adult C. acrocomiaefeeds on pollen, remaining bance and hunting, the number of species and during the day under dead leaf sheaths and in the abundanceof wild fauna have been drasti- the leaf sheath of A. aculeata. At night they cally reduced in the study area. In fact, in this oviposit on fallen fruits. After oviposition the area the wild fauna were represented only by larvae perforate the exocarp and mesocarp (if small animals that are not important for the dis- present),and finally the endocarp,reaching the persal of,4crocorniaaculeata. In the Pantanal (a seed. More than one larva can reach the seed, wetland areain Mato Grossoand Mato Grossodo but only one reachesthe adult stage,which sug- Sul states, Brazil), where fruits of A. aculeata are geststhat the first larva reachingthe seedcanni- smallerthan in my study sites (personalobserva- balizes the others. By the final instar, the larva tion), endocarpswere found to constitute637o of consumesalmost all (Pachymerussp.) or all (C. the volume of 69 fecal samples of Cerd'ocyon auocomiae) of the seed. An operculum, visible thous (crab-eatingwolf, J. C. Dalponte,personal from outside the fruit, protects the insect from communication).Schaller (1983) also found en- contactwith the external environmentand opens docarpsin224o of 36fecalsamplesof C.thousin when the adult bruchid emerges.In the experi- the same area. These observationsimplicate C. mental seedsPachym'erus sp. (13 individuals) thous as a disperser of A. aculeata in the Pan- was more abundant than C. acrocomiae(6 indi- tanal, and perhapsin other areaswhere fruits are viduals). smaller than in my study area. In the Pantanal After three days on the ground >507o of the regionothe macawsAnadorhincus hyacinthinus chewedfruits in boxes had at least one bruchid andAra chloropterafeed on seedsand transport egg, and after27 days,997ohad eggsdeposited rachillae of A. aculeala with fruits attached, on them, suggestingthat chewed fruits had a which eventually fall during the flight (C. Ya- higher level of oviposition than unchewedfruits mashita, personal communication).However, in r9981 SCARIOT:DISPERSAL AND PREDATION IN ACROCOMIAACULEATA contemporaryseed dispersers. It is also possible that the wide geographicdistribution results in E80 part from use by Amerindians (Kozak et al. t 1.979,Hill et al. 1984, Bale6 1989), as South 'F 60 American Indians used seeds of A. aculeatu a (Levi-Strauss1963), and the ChacoIndians used 'ts 40 H seedsof A. totai (M6traux 1963) as food. Also, Unchewed o20 Turner and Miksicek (1984)found a mineralized s seed of Acrocomia in Maya sites in Guatemala, which suggests possible ancient use of tlft 20 40 species by these people in Central America. Time (Days) Lentz (1990) suggestedthat it was introduced in somesites in Mexico and Central America by the Maya.It is also possible thatAcrocomio is one of the large-seededplant specieswhose fruits were E o consumed and dispersed by large Pleistocene I e 600 mammals (Janzenand Martin 1982). According to this hypothesis,cattle are now an ecological a analogueof thesePleistocene dispersers. il 400 The strong odor from recently chewed fruits may serveas a cue for oviposition.The high rate o 200 of oviposition on individual fruits may be due to low fruit abundancein comparisonto the female z - Ulchewed 0 tu bruchid population. Indeed, a bruchid female 020406080 can lay 50-100 eggs,sewing an equal number of Time (Days) fruits (Janzen 1980). Oviposition on recently fallen fruits increases the probability that the l. Ovipositionby bruchidbeetles on fruits of Acrocomia first larva to reach the seedmay have a competi- aculeata either chewed or not chewed by cattle.