- Email: [email protected]
Diet and seed dispersal by five marsupials (Didelphimorphia: Didelphidae) in a Brazilian cerrado reserve
ARTICLE IN PRESS
www.elsevier.de/mambio
ORIGINAL INVESTIGATION
Diet and seed dispersal by five marsupials (Didelphimorphia: Didelphidae) in a Brazilian cerrado reserve Leonardo Guimaraes Lessa, Fabiane Nepomuceno da Costa Department of Biological Sciences, Federal University of Jequitinhonha and Mucuri Valleys, Rua da Gloria, 187, Campus II, 39100-000 Diamantina, Minas Gerais, Brazil Received 3 July 2008; accepted 3 November 2008
Abstract The food habits and seed dispersal promoted by five didelphid marsupials, were investigated through fecal analysis in a cerrado reserve located in the central part of the Espinhac¸o mountain range in Minas Gerais state, southeastern Brazil. The studied species presented a remarkable constancy in their diets and just for C. philander a seasonal change in diet seems to be present with a high consumption of fruits during the wetter season. Arthropods such as hymenopterans, coleopterans and homopterans were the main food resource detected but seeds belonging to pioneer plants were found with high frequencies in samples of all didelphids. Fruits of pioneer plants belonging to the families Melastomataceae (Clidemia urceolata and Miconia holocericea) and Rubiaceae (Psychotria barbiflora and P. capitata) present high germination rates and were among the mostly consumed. In spite of the differences observed in food preferences, all of the didelphids studied must be considered effective seed dispersers of pioneer plants. r 2008 Deutsche Gesellschaft fu¨r Sa¨ugetierkunde. Published by Elsevier GmbH. All rights reserved. Keywords: Didelphids; Diet; Seed dispersal; Cerrado; Espinhac¸o range
Introduction The marsupial family Didelphidae includes about 55 species in Brazil (Rossi et al. 2006), but despite the commonness and wide distribution for most of them, detailed information on the diet and other aspects of their natural history is still scarce or non-existent ´ stua de Moraes 2003; (Ca´ceres et al. 2002; Vieira and A Lessa et al. in press). Didelphids have been frequently reported as omnivorous mammals consuming invertebrates, fruits, small vertebrates and occasionally carrion and other plant parts (Fonseca et al. 1996; Rossi et al. 2006). However, recent studies have yielded information that allowed researchers to describe more precisely the Corresponding author. Fax: +55 38 3532 1223.
E-mail address: [email protected] (L.G. Lessa).
feeding habits of some species, such as Caluromys spp. (Ca´ceres 2005; Casella and Ca´ceres 2006), Metachirus nudicaudatus (Santori et al. 1995; Ca´ceres 2004) and Micoureus paraguayanus (Casella and Ca´ceres 2006) which are now regarded as primarily frugivorous, insectivorous and insectivorous–omnivorous, respectively. Therefore, knowledge about the diet of any species is an important parameter to understand its ´ stua de Moraes relationship with the environment (A et al. 2003; Martins and Bonato 2004). Another important phenomenon related to feeding habit is seed dispersal (Ca´ceres 2002) which have been associated with different aspects, such as the presence of intact seeds in stomachs (Talamoni et al. 2007), intact seeds in scats (Ca´ceres et al. 2002; Pinheiro P.S. et al. 2002) and plant species (fruits) eaten (Casella and Ca´ceres 2006; Martins et al. 2006; Leiner and Silva
1616-5047/$ - see front matter r 2008 Deutsche Gesellschaft fu¨r Sa¨ugetierkunde. Published by Elsevier GmbH. All rights reserved. doi:10.1016/j.mambio.2008.11.002 Mamm. biol. 75 (2010) 10–16
ARTICLE IN PRESS L.G. Lessa, F.N. da Costa / Mamm. biol. 75 (2010) 10–16
2007). However, the authors reported did not assess the potential role of didelphids as seed dispersers by germination experiments (see Grelle and Garcia 1999; Ca´ceres and Monteiro-Filho 2000; Ca´ceres 2002, 2004). In Brazil, studies on feeding ecology of didelphid marsupials have been restricted to disturbed and nondisturbed Atlantic forests areas (e.g., Santori et al. 1995; Leite et al. 1996; Grelle and Garcia 1999; Pinheiro P.S. et al. 2002; Ca´ceres 2004; Martins and Bonato 2004; Casella and Ca´ceres 2006; Leiner and Silva 2007). Information on the diet composition of didelphids in the Brazilian cerrado, one of the most endangered ecosystems of the world (Myers et al. 2000), is limited to some few records (Martins et al. 2006; Talamoni et al. 2007). The most detailed study was conducted by Martins et al. (2006) with the little mouse opossum Gracilinanus microtarsus. Although the authors report the occurrence of fruits in the diet of G. microtarsus, information on the potential role of this species and other didelphids as seed dispersers in the Brazilian cerrado is still lacking. The objectives of this study were twofold. First, we determined the diet of five sympatric didelphid marsupials, namely Gracilinanus agilis (Burmeinster, 1854), Marmosops incanus (Lund, 1840), Metachirus nudicaudatus (Desmarest, 1817), Micoureus paraguyanus (Thomas, 1905) and Caluromys philander (Linnaeus, 1758), in a cerrado reserve in southeastern Brazil. Secondly, we evaluated the viability of the seeds found in the scats of the study species to assess their potential roles as seed dispersers in this important Brazilian ecosystem.
Material and methods Study site This study was carried out in the Parque Estadual do Rio Preto (PERP) (181090 s, 431230 W), located in the central part of the Espinhac¸o mountain range in Minas Gerais State, southeastern Brazil. The study area is an important remnant of Brazilian cerrado (savanna-like vegetation), which is defined as a xeromorphic vegetation type. The cerrado comprises five different vegetation physiognomies, namely, campo limpo (clean field), campo sujo (dirty field), campo cerrado (closed field), cerrado sensu stricto and cerrada˜o (closed woodland) (Eiten 1992). Most of the area of the PERP is occupied by campo limpo, cerrado sensu stricto and gallery forests. The local topography is characterized by a hilly relief with areas that varied between 720 and 1830 m of altitude. The annual rainfall ranges from 8.25 to 223.19 mm concentrated mainly in the wet season (October– March), although some rain may occur during the dry season (April–September).
11
Trapping of animals and data recording The five didelphid species were captured weekly from February 2005 to March 2006 using 100 mesh-wire livetraps (30 cm 15 cm 15 cm) distributed along two tracks located 50 m apart from each other. Trapping stations were placed 20 m apart from each other on each track and traps were set on the ground and on tree branches or vines at height of about 1.5 m. All traps were baited with a mixture of banana, pineapple, and cotton balls soaked with cod-liver oil during the afternoon and checked for captures during the following morning. All animals were marked with numbered ear tags (Zootech, Curitiba, PR, Brazil) and released. Scats on the trap floor and those defecated by the individuals during the manipulation were collected. One fecal sample was considered as being all feces produced by one animal in a night.
Diet analysis Scats collected were stored in numbered small plastic containers and frozen to avoid deterioration. In the laboratory, the scats samples were washed in a 0.1 mm mesh sieve. Food items were assigned to the following categories: seeds, other plant parts (flower and fruit fragments), arthropods, vertebrates and unidentified material. Items were identified at the lowest possible taxonomic category by comparison with a reference collection of invertebrates and seeds from the study area. When present, seeds were counted in each fecal sample. The bait residue was detected and excluded during the scats analyses. The relative frequency of occurrence (Korschgen 1987) was used to determine the contribution of each item in all food categories to the diet of didelphids. This statistic was calculated as the frequency of scats containing a particular food item multiplied by 100 (Korschgen 1987). The non-parametric Mann–Whitney U-test (Ayres and Ayres Jr. 2000) was used to test diet differences between the dry and the wet season. The G-test was used to compare the proportion of seeds and arthropod groups in the diet of the didelphids (Sokal and Rohlf 1995).
Germination tests Seeds found in scats were placed in closed Petri dishes (diameter ¼ 9 cm) containing wet absorbing paper to verify germination and monitored weekly. As control experiments, seeds of the same plant species consumed by the didelphids were collected directly from fruiting plants and tested for germination as described above. To test differences between germination rates of control group versus seeds found in the scats we use the w2 statistic (Sokal and Rohlf 1995).
ARTICLE IN PRESS 12
L.G. Lessa, F.N. da Costa / Mamm. biol. 75 (2010) 10–16
Results Diet A total of 157 fecal samples were obtained: G. agilis (n ¼ 56); M. incanus (n ¼ 58); M. nudicaudatus (n ¼ 11); M. paraguayanus (n ¼ 22) and C. philander (n ¼ 10) (Table 1). Arthropods were the main food items consumed and occurred in 100% of M. incanus and M. nudicaudatus samples. Hymenopterans (Formicidae) occurred in 26.8% of G. agilis samples and in 40% of C. philander. Coleopterans (Scarabeidae) were the most frequent item found for M. paraguayanus. The presence of Dipterans was detected by pupae remains found for M. incanus and G. agilis. Isopterans (Termitidae) were also largely recorded in samples of all species. Vertebrate preys (bones and feathers fragments of birds) were found with low frequencies for M. nudicaudatus and M. paraguayanus (Table 1). A total of 13 different morphotypes of seeds and other plant parts were found. The vegetal material belonging to the families: Anacardiaceae, Lauraceae, Melastomataceae, Myrtaceae, Hypericaceae, Rubiaceae and Meliaceae were identified. Seeds of Clidemia urceolata and Miconia holosericea (Melastomataceae) were found with high frequencies in samples of all didelphids (Table 1). Seeds ingested ranged from 1 to 5 mm in length. Larger seed quantities were found for C. philander (Ocotea lancifolia: n ¼ 1200 seeds; C. urceolata: n ¼ 880 seeds), M. nudicaudatus (C. urceolata: n ¼ 1204 seeds) and M. incanus (C. urceolata: n ¼ 824 seeds). Fragments of flowers were also seen in samples of M. incanus (Lauraceae) and C. philander (unidentified). No significant differences were observed in the diet between the dry and the wet season for M. incanus (M–W. U ¼ 77.00, p ¼ 0.37); G. agilis (M–W. U ¼ 25.00, p ¼ 0.11); M. nudicaudatus (M–W. U ¼ 32.00, p ¼ 0.32); and M. paraguayanus (M–W. U ¼ 81.50, p ¼ 0.89). In contrast, the diet of C. Philander differed between the dry and wet season (M–W. U ¼ 32.00, p ¼ 0.01), largely as a result of the higher consumption of fruits in the wet season. Arthropods in the diet were more frequent just for two species, M. incanus (G ¼ 6.11, df ¼ 1, p ¼ 00.1) and G. agilis (G ¼ 5.99, df ¼ 1, p ¼ 00.1).
Seed germination Germination rates of seeds from scats were higher than those of control groups (Table 2). Germination tests revealed that for the smaller didelphids such as G. agilis and M. incanus, all seeds that remained viable were less than 1 mm long (e.g., C. urceolata, M. holosericea and Psychotria barbiflora). For the larger species such as M. paraguayanus, M. nudicaudatus and C. philander, the largest intact seeds belonged to
Ocotea lancifolia (3 mm) and Amaioua guianensis (5 mm). Seeds greater than 5 mm were found destroyed (e.g., Tapirira obtuse 7 mm). Fruits of pioneer plants belonging to the families Melastomataceae (C. urceolata and M. holocericea) and Rubiaceae (P. barbiflora and P. capitata) presented high germination rates (Table 2). Fruits of C. urceolata were consumed in high proportions and germination rates of seeds from scats were significantly higher than those of control groups for the five studied species: M. incanus (w2 ¼ 128.5992, df ¼ 1, po0.01); G. agilis (w2 ¼ 13.8161, df ¼ 1 po0.01); M nudicaudatus (w2 ¼ 60.6644, df ¼ 1, po0.01); M. paraguayanus (w2 ¼ 59.2268, df ¼ 1, po0.01) and C. philander (w2 ¼ 45.0309, df ¼ 1, po0.01) (Table 2).
Discussion Food habits Our data showed that the didelphids sampled in a cerrado reserve in Brazil presented a remarkable constancy in their diets among the seasons. Just for C. philander, a seasonal change in diet seems to be present with a high frequency of fruit consumption during the wetter season. Arthropods such as hymenopterans and coleopterans were the main food resource detected. A similar pattern was also reported by Santori et al. (1995) for a restinga (shrubland) forest and Ca´ceres (2002) for a mixed ombrophilous forest in southern Brazil. Nevertheless, the relative abundance of Hymenoptera and Coleoptera in tropical forests would make these groups the most frequent prey in the diet of neotropical didelphids and does not necessarily imply that food selection is taking place (Pinheiro P.S. et al. 2002). This relative abundance may be related with two factors. First, an overestimation of this item because hard-bodied preys are easier to be detected in scats samples (Dickman and Huang 1988). Second, the high abundance of ants and beetles in the study area, although we did not estimated the availability of this items in the study site. Gracilinanus agilis is the smallest of the five species considered in this study. Its diet, based on the relative frequency of occurrence of prey in scats presented fruits and a high consumption of arthropods, mainly hymenopterans and isopterans. This diet composition is similar to that reported by the congener species, G. microtarsus (Talamoni et al. 1999; Martins and Bonato 2004; Martins et al. 2006). The high consumption of ants and termites, the most abundant insects in the cerrado (Pinheiro F. et al. 2002), could reveal a particularly high availability of these social insects in the study area and thus reflect an opportunistic food choice
ARTICLE IN PRESS L.G. Lessa, F.N. da Costa / Mamm. biol. 75 (2010) 10–16
13
Table 1. Relative frequency of occurrence (%) of food items found in the scats of five didelphids caught in a gallery forest from February 2005 to March 2006 in the PERP, Brazil. Food items
Seeds Anacardiaceae Tapirira obtusa (Benth) J. D. Mitchell Lauraceae Ocotea lancifolia (Schott) Mez Ocotea sp. Melastomataceae Clidemia urceolata DC. Miconia holosericea (L.) DC. Unidentified Melastomataceae Myrtaceae Myrcia sp. Unidentified Myrtaceae Hypericaceae Vismia glaziovii Ruhland Rubiaceae Amaioua guianensis Aubl. Psychotria barbiflora DC. Psychotria capitata Ruiz & Pav. Unidentified Rubiaceae Unidentified seeds Seeds (total) Fruits fragmentsa Meliaceae Unidentified Flowers fragmentsb Lauraceae Unidentified Vegetal fragments (total) Arthropoda Diplopoda Arachnidac Orthopterad Coleopterae Blattodea Hymenopteraf Isopterag Diptera (pupa) Hemypterah Unidentified Arthropoda Arthropoda (total) Vertebrates Birds (feathers) Vertebrates (total)
Gracilinanus agilis (n ¼ 56) – – – – – – – 5.3 1.8 – – – – – – – 3.6 3.6 1.8 7.1 19.6 – – 3.6 – – – 3.6 – – – 1.8 1.8 – 26.8 10.8 1.8 1.8 50.0 84.0 – – –
n ¼ total number of scats analyzed. a Pericarp. b Anthers and petals. c Araneidae and Acarina. d Gryllidae. e Scarabeidae and unidentified Coleoptera. f Formicidae. g Termitidae. h Heteroptera and Homoptera.
Marmosops incanus (n ¼ 58)
Metachirus nudicaudatus (n ¼ 11)
Micoureus paraguayanus (n ¼ 22)
Caluromys philander (n ¼ 10)
– – –
– – –
– – 4.5
– – –
– – – – 12.0 5.1 3.4 – – – – – – – – 3.4 1.7 3.4 29.3 – 1.7 8.7 – 3.4 5.1 19.0 – – 1.7 5.1 13.8 8.7 66.0 13.8 10.3 12.0 24.1 100.0 – – –
– – – – 27.2 9.0 – – – 9.0 – – – – – – – – 45.5 – – – – – – – – 18.1 9.0 36.4 45.5 45.5 72.8 45.5 – – 18.1 100.0 – 18.1 18.1
– – – – 9.0 9.0 4.5 – 9.0 – – – – 13.6 4.5 – – 4.5 45.5 – – 4.5 – – 22.7 27.3 – – 9.0 9.0 36.3 – 22.7 4.5 – 22.7 45.4 91.0 – 9.0 9.0
– 10.0 10.0 – 30.0 20.0 10.0 – 20.0 – – 20.0 – – – 10.0 – 10.0 90.0 – – – – 10.0 – 10.0 – – 10.0 – 30.0 – 40,0 10.0 – – 10.0 70.0 – – –
ARTICLE IN PRESS 14
L.G. Lessa, F.N. da Costa / Mamm. biol. 75 (2010) 10–16
Table 2. Number of seeds in test and frequencies of germination (%) of seeds found in the scats of five didelphids from a gallery forest in the PERP, Brazil. Family/species
Control experiments
Marmosops incanus
Gracilinanus Metachirus agilis nudicaudatus
Micoureus paraguayanus
Caluromys philander
Anacardiaceae Tapirira obtusa (Benth) J. D. Mitchell Lauraceae Ocotea lancifolia (Schott) Meza Ocotea sp.a Melastomataceae Clidemia urceolata DC.a Miconia holosericea (L.) DC.a Unidentified Melastomataceae Myrtaceae Myrcia sp.a Unidentified Myrtaceae Hypericaceae Vismia glaziovii Ruhlanda Rubiaceae Amaioua guianensis Aubl. Psychotria barbiflora DC.a Psychotria capitata Ruiz & Pav.a Unidentified Rubiaceae
– 6 (33)
– –
– –
– –
– 4 (50)
– –
16 (37)
– –
– –
– –
– –
– 100 (51)
45 (22) 45 (26)
– – 298 (92) 48 (81)
– – 22 (73) –
– – 120 (87) 10 (90)
– – 60 (97) 64 (55)
30 (73) – 130 (79) 10 (40)
25 (12)
10 (100)
10 (60)
–
–
–
– 30 – – 10 – 10 10 10
– – – – – – – – 8 (75)
– – – – – – – 3 (67) 4 (50)
– – – – – – – – –
– 30 (73) – – – – 9 (44) 15 (80) –
– 10 (60) 4(50) – 10 (60) – – – 6 (60)
–
2 (50)
–
–
–
a
28 (25)
(20)
(10) (20) (20) (20)
10 (50)
Pioneer plants.
by G. agilis (see Martins et al. 2006, for seasonality data of G. microtarsus). Marmosops incaus is also a small didelphid and an insectivorous–omnivorous diet is indicated for all species within the genus (Fonseca et al. 1996). However, a preliminary analysis on the relative frequency of occurrence of arthropods among the dry and the wet season revealed a remarkable constancy on the consumption of arthropods, mainly hymenopterans. Despite the high consumption of arthropods, present in 100% of the fecal samples, fruits seem to be an important resource. A similar pattern was found by Leiner and Silva (2007) working in an area of Brazilian Atlantic forest for the congeneric species M. paulensis, suggesting the importance of this item as an energetic supplement in both seasons (Julien-Laferie`rre and ´ stua Atramentowics 1990; Pinheiro P.S. et al. 2002; A de Moraes et al. 2003). The consumption of flowers, especially during the dry season was observed for M. paulensis (Leiner and Silva 2007) and the same trend would occur here for M. incanus, but to a lesser extent. Metachirus nudicaudatus seems to be more insectivorous, consuming a reasonable number of invertebrates, as was expected for a terrestrial didelphid (Santori et al. 1995). However, the local food availability could
determine variations in the diet of certain species (Ca´ceres et al. 2002). Although M. nudicaudatus has been regarded as an insectivorous species (Santori et al. ´ stua de Moraes 2003), 1995; Ca´ceres 2004; Vieira and A fruits of pioneer plants such as Clidemia urceolata occurred at a relatively high proportion (45.5%) in scats collected in the cerrado (present study) but were detected in low proportions in restinga (10.5%) and Atlantic forest (7%) studies (Santori et al. 1995; Ca´ceres 2004, respectively). Micoureus paraguayanus has been reported as insectivorous–omnivorous, in general consuming large quantities of insects, a selected group of fruits and rarely vertebrates (Leite et al. 1996; Ca´ceres et al. 2002; Pinheiro P.S. et al. 2002; Casella and Ca´ceres 2006) which is in agreement with our results, although it also fed on fruits and flowers fragments in a reasonable proportion (45.5% and 22.7%, respectively). As was emphasized by Santori et al. (1995), ‘‘there is a tendency towards the insectivory in smaller neotropical marsupial species’’ although, sometimes the consumption of vertebrate prey such as birds has been reported by this specie (Ca´ceres et al. 2002; Casella and Ca´ceres 2006; this study). Caluromys philander had a more frugivorous diet (Leite et al. 1996; Casella and Ca´ceres 2006) as was
ARTICLE IN PRESS L.G. Lessa, F.N. da Costa / Mamm. biol. 75 (2010) 10–16
15
expected for an arboreal species (Casella and Ca´ceres 2006), but arthropods were found in its diet at high frequencies (70%). Julien-Laferie`rre (1999) showed the importance of insects as nitrogen resources for C. philander in French Guiana. Coleopterans, hymenopterans and arachnids were important prey items for this species, mainly during the drier season. Fruits of pioneer plants such as Miconia holosericea, Ocotea lancifolia and Myrcia sp. were mainly consumed during the wetter season. Caluromys philander seems to be frugivorous but also use other food items in different proportions, the same trend would occur for the congeneric C. lanatus (Ca´ceres 2005; Casella and Ca´ceres 2006). These results are in agreement with Fonseca et al. (1996) that proposed a frugivorous–omnivorous diet for Caluromys spp.
regenerating several types of environments through seed dispersal (Pinheiro P.S. et al. 2002; Ca´ceres et al. 2002; Leiner and Silva 2007), given that the necessary conditions of movement and appropriateness of site of feces/seed deposition are satisfied.
Role in seed dispersal
References
Previous studies detected a tendency towards insectivory in smaller neotropical didelphids (Santori et al. 1995; Martins and Bonato 2004); however, fruits seem to be an important supplement on its diets (Julien´ stua de Laferie`rre 1999; Pinheiro P.S. et al. 2002; A Moraes et al. 2003; Leiner and Silva 2007; this study) also for the more insectivorous species, such as G. agilis, M. incanus and M. nudicaudatus. The five didelphids studied fed on fruits in varying frequencies, mainly fruits of pioneer plants. Moreover, seeds consumed remained viable after passing undamaged through the digestive tract of the animals and presented a high germination rate. As evidenced in other studies (Ca´ceres and Monteiro-Filho 2000; Ca´ceres 2002; Casella and Ca´ceres 2006), there seems to be a relation among the seed size and the animal body size. For the smaller didelphids sampled such as G. agilis and M. incanus, just the small seeds (1 mm) remain viable. For the largest species such as M. nudicaudatus, M. paraguayanus and C. Philander, large seeds (between 3 and 5 mm) remain viable after cross undamaged the gut. The potential role of didelphid marsupials on seed dispersal, especially for small seeds belonging to pioneer plants, has been show in disturbed and non-disturbed Atlantic forests remnants (Leite et al. 1996; Ca´ceres et al. 2002; Pinheiro P.S. et al. 2002; Casella and Ca´ceres 2006; Leiner and Silva 2007). However, just a few studies confirmed this potential role with germination tests (Grelle and Garcia 1999; Ca´ceres and MonteiroFilho 2000; Ca´ceres 2002). Our results confirmed that the germination rates of seeds from scats are higher than germination rates of seeds collected from fruits (control experiments), especially for pioneer species belonging to the families Melastomataceae and Rubiaceae. These results stress the importance of neotropical didelphids in
´ stua de Moraes, D., Santori, R.T., Finotti, R., Cerqueira, R., A 2003. Nutritional and fibre contents of laboratory-established diets of neotropical opossums (Didelphidae). In: Jones, C., Dickman, J.M., Archer, M. (Eds.), Predators with Pounches: The Biology of Carnivorous Marsupials. CSIRO Publishing, Melborne, pp. 229–237. Ayres, M., Ayres Jr., M., 2000. Bioestat 5.0: aplicac¸o˜es estatı´sticas nas a´reas das cieˆncias biolo´gicas e me´dicas. Sociedade Civil Mamiraua´, CNPq, Bele´m. Ca´ceres, N.C., 2002. Food habits and seed dispersal by the white-eared-opossum, Didelphis albiventris (Mammalia, Marsupialia) in southeastern Brazil. Stud. Neotrop. Fauna Environ. 37, 97–104. Ca´ceres, N.C., 2004. Diet of three didelphid marsupials (Mammalia, Didelphimorphia) in southeastern Brazil. Mam. Biol. 69, 430–433. Ca´ceres, N.C., 2005. Comparative lengths of digestive tracts of seven didelphid marsupials (Mammalia) in relation to diet. Rev. Bras. Zool. 22, 181–185. Ca´ceres, N.C., Monteiro-Filho, E.L.A., 2000. The common opossum, Didelphis aurita, as a seed disperser of several plants in southern Brazil. Cieˆnc. Cult. 52, 41–44. Ca´ceres, N.C., Ghizoni, I.R., Graipel, M.E., 2002. Diet of two marsupials, Lutreolina crassicaudata and Micoureus demerarae, in coastal Atlantic Forest island of Brazil. Mammalia 66, 331–340. Casella, J., Ca´ceres, N.C., 2006. Diet of four small mammals species from Atlantic forest patches in south Brazil. Neotrop. Biol. Cons. 1, 5–11. Dickman, C.R., Huang, C., 1988. The reliability of fecal analysis as a method for determining the diet of insectivorous mammals. J. Mammal. 69, 108–113. Eiten, G., 1992. Natural Brazilian vegetation types and their causes. Ann. Acad. Bras. Cieˆnc. 64 (Suppl. I), 35–65. Fonseca, G.A.B., Herrmann, G., Leite, Y.R.L., Mittermier, R.A., Rylands, A.B., Patton, J.L., 1996. Lista anotada de mamı´feros do Brasil. Occas. Pap. Cons. Biol. 4, 1–38. Grelle, C.E., Garcia, Q.S., 1999. Potential dispersal of Cecropia hololeuca by the common opossum (Didelphis
Acknowledgements We thank Soˆnia Talamoni and Lena Geise for suggestions to the early version of the manuscript. IEF-MG (Minas Gerais State Forest Institute) for allowing the access at PERP. We are also indebted to two anonymous reviewers for their comments and suggestions that greatly improved and clarify the manuscript. The study was supported by FAPEMIG (process number CRA 133-03).
ARTICLE IN PRESS 16
L.G. Lessa, F.N. da Costa / Mamm. biol. 75 (2010) 10–16
aurita) in Atlantic Forest, southeastern Brazil. Rev. Ecol. (Terre Vie) 54, 327–332. Julien-Laferie`rre, D., 1999. Foraging strategies and food partitioning in the neotropical frugivorous mammals Caluromys philander and Potos flavus. J. Zool. (London) 247, 71–80. Julien-Laferie`rre, D., Atramentowics, M., 1990. Feeding and reproduction of three didelphid marsupials in two neotropical forests (French Guiana). Biotropica 22, 404–415. Korschgen, J.L., 1987. Procedimientos para el ana´lisis de los ha´bitos alimentarios. Manual de te´cnicas de gestio´n de vida silvestre. The Wildlife Society, Washington, DC. Leiner, N.O., Silva, W.R., 2007. Seasonal variation in the diet of the Brazilian slender opossum (Marmosops paulensis) in a montane Atlantic Forest area, southeastern Brazil. J. Mammal. 88, 158–164. Leite, Y.R.L., Costa, L.P., Stallings, J.R., 1996. Diet and vertical space use of three sympatric opossums in a Brazilian Atlantic Forest reserve. J. Trop. Ecol. 12, 435–440. Lessa, L.G., Costa, B.M.A., Rossoni, D.M., Tavares, V.C., Dias, L.G., Moares-Junior, E.A., Silva, J.A., in press. Mamı´feros da Cadeia do Espinhac¸o, riqueza, ameac¸as e estrate´gias para conservac¸a˜o. Megadiversidade. Martins, E.G., Bonato, V., 2004. On the diet of Gracilinanus microtarsus (Marsupialia, Didelphidae) in an Atlantic rainforest fragment in southeastern Brazil. Mam. Biol. 69, 58–60. Martins, E.G., Bonato, V., Pinheiro, H.P., Dos Reis, S.F., 2006. Diet of the gracile mouse opossum (Gracilinanus microtarsus) (Didelphimorphia: Didelphidae) in a Brazilian Cerrado: patterns of food consumption and intrapopulation variation. J. Zool. (London) 269, 21–28.
Myers, N., Mittermeier, R.A., Mittermeier, C.G., Fonseca, G.A.B., Kent, J., 2000. Biodiversity hotspots for conservation priorities. Nature 403, 853–858. Pinheiro, F., Diniz, I.R., Coelho, D., Bandeira, M.P.S., 2002. Seasonal pattern of insect abundance in the Brazilian cerrado. Aust. Ecol. 27, 132–136. Pinheiro, P.S., Carvalho, F.M.V., Fernandez, F.A.Z., Nessimian, J.L., 2002. Diet of the marsupial Micoureus demerarae in small fragments of Atlantic Forest in southeastern Brazil. Stud. Neotrop. Fauna Environ. 37, 213–218. Rossi, R.V., Bianconi, G.V., Pedro, W.A., 2006. Ordem Didelphimorphia. In: Reis, R.R., Peracchi, A.L., Pedro, W.A., Lima, I.P. (Eds.), Mamı´feros do Brasil. Parana´, Londrina, pp. 27–60. Santori, R.T., Moraes, D.A., Cerqueira, R., 1995. Diet composition of Metachirus nudicaudatus and Didelphis aurita (Marsupialia, Didelphoidea) in southeastern Brazil. Mammalia 59, 511–516. Sokal, R.R., Rohlf, F.J., 1995. Biometry. W.H. Freeman and Company, New York, 887pp. Talamoni, S.A., Couto, D., Lopes, M.O.G., Cordeiro-Ju´nior, D.A., 1999. Dieta de algumas espe´cies de pequenos mamı´feros do sudeste brasileiro. Bios 7, 51–56. Talamoni, A.S., Couto, D., Ju´nior, D.A.C., Diniz, F.M., 2007. Diet of some species of neotropical small mammals. Mam. Biol. ´ stua de Moraes, D., 2003. Carnivory and Vieira, E.M., A insectivory in neotropical marsupials. In: Jones, C., Dickman, J.M., Archer, M. (Eds.), Predators with Pounches: The Biology of Carnivorous Marsupials. CSIRO Publishing, Melborne, pp. 267–280.
www.elsevier.de/mambio
ORIGINAL INVESTIGATION
Diet and seed dispersal by five marsupials (Didelphimorphia: Didelphidae) in a Brazilian cerrado reserve Leonardo Guimaraes Lessa, Fabiane Nepomuceno da Costa Department of Biological Sciences, Federal University of Jequitinhonha and Mucuri Valleys, Rua da Gloria, 187, Campus II, 39100-000 Diamantina, Minas Gerais, Brazil Received 3 July 2008; accepted 3 November 2008
Abstract The food habits and seed dispersal promoted by five didelphid marsupials, were investigated through fecal analysis in a cerrado reserve located in the central part of the Espinhac¸o mountain range in Minas Gerais state, southeastern Brazil. The studied species presented a remarkable constancy in their diets and just for C. philander a seasonal change in diet seems to be present with a high consumption of fruits during the wetter season. Arthropods such as hymenopterans, coleopterans and homopterans were the main food resource detected but seeds belonging to pioneer plants were found with high frequencies in samples of all didelphids. Fruits of pioneer plants belonging to the families Melastomataceae (Clidemia urceolata and Miconia holocericea) and Rubiaceae (Psychotria barbiflora and P. capitata) present high germination rates and were among the mostly consumed. In spite of the differences observed in food preferences, all of the didelphids studied must be considered effective seed dispersers of pioneer plants. r 2008 Deutsche Gesellschaft fu¨r Sa¨ugetierkunde. Published by Elsevier GmbH. All rights reserved. Keywords: Didelphids; Diet; Seed dispersal; Cerrado; Espinhac¸o range
Introduction The marsupial family Didelphidae includes about 55 species in Brazil (Rossi et al. 2006), but despite the commonness and wide distribution for most of them, detailed information on the diet and other aspects of their natural history is still scarce or non-existent ´ stua de Moraes 2003; (Ca´ceres et al. 2002; Vieira and A Lessa et al. in press). Didelphids have been frequently reported as omnivorous mammals consuming invertebrates, fruits, small vertebrates and occasionally carrion and other plant parts (Fonseca et al. 1996; Rossi et al. 2006). However, recent studies have yielded information that allowed researchers to describe more precisely the Corresponding author. Fax: +55 38 3532 1223.
E-mail address: [email protected] (L.G. Lessa).
feeding habits of some species, such as Caluromys spp. (Ca´ceres 2005; Casella and Ca´ceres 2006), Metachirus nudicaudatus (Santori et al. 1995; Ca´ceres 2004) and Micoureus paraguayanus (Casella and Ca´ceres 2006) which are now regarded as primarily frugivorous, insectivorous and insectivorous–omnivorous, respectively. Therefore, knowledge about the diet of any species is an important parameter to understand its ´ stua de Moraes relationship with the environment (A et al. 2003; Martins and Bonato 2004). Another important phenomenon related to feeding habit is seed dispersal (Ca´ceres 2002) which have been associated with different aspects, such as the presence of intact seeds in stomachs (Talamoni et al. 2007), intact seeds in scats (Ca´ceres et al. 2002; Pinheiro P.S. et al. 2002) and plant species (fruits) eaten (Casella and Ca´ceres 2006; Martins et al. 2006; Leiner and Silva
1616-5047/$ - see front matter r 2008 Deutsche Gesellschaft fu¨r Sa¨ugetierkunde. Published by Elsevier GmbH. All rights reserved. doi:10.1016/j.mambio.2008.11.002 Mamm. biol. 75 (2010) 10–16
ARTICLE IN PRESS L.G. Lessa, F.N. da Costa / Mamm. biol. 75 (2010) 10–16
2007). However, the authors reported did not assess the potential role of didelphids as seed dispersers by germination experiments (see Grelle and Garcia 1999; Ca´ceres and Monteiro-Filho 2000; Ca´ceres 2002, 2004). In Brazil, studies on feeding ecology of didelphid marsupials have been restricted to disturbed and nondisturbed Atlantic forests areas (e.g., Santori et al. 1995; Leite et al. 1996; Grelle and Garcia 1999; Pinheiro P.S. et al. 2002; Ca´ceres 2004; Martins and Bonato 2004; Casella and Ca´ceres 2006; Leiner and Silva 2007). Information on the diet composition of didelphids in the Brazilian cerrado, one of the most endangered ecosystems of the world (Myers et al. 2000), is limited to some few records (Martins et al. 2006; Talamoni et al. 2007). The most detailed study was conducted by Martins et al. (2006) with the little mouse opossum Gracilinanus microtarsus. Although the authors report the occurrence of fruits in the diet of G. microtarsus, information on the potential role of this species and other didelphids as seed dispersers in the Brazilian cerrado is still lacking. The objectives of this study were twofold. First, we determined the diet of five sympatric didelphid marsupials, namely Gracilinanus agilis (Burmeinster, 1854), Marmosops incanus (Lund, 1840), Metachirus nudicaudatus (Desmarest, 1817), Micoureus paraguyanus (Thomas, 1905) and Caluromys philander (Linnaeus, 1758), in a cerrado reserve in southeastern Brazil. Secondly, we evaluated the viability of the seeds found in the scats of the study species to assess their potential roles as seed dispersers in this important Brazilian ecosystem.
Material and methods Study site This study was carried out in the Parque Estadual do Rio Preto (PERP) (181090 s, 431230 W), located in the central part of the Espinhac¸o mountain range in Minas Gerais State, southeastern Brazil. The study area is an important remnant of Brazilian cerrado (savanna-like vegetation), which is defined as a xeromorphic vegetation type. The cerrado comprises five different vegetation physiognomies, namely, campo limpo (clean field), campo sujo (dirty field), campo cerrado (closed field), cerrado sensu stricto and cerrada˜o (closed woodland) (Eiten 1992). Most of the area of the PERP is occupied by campo limpo, cerrado sensu stricto and gallery forests. The local topography is characterized by a hilly relief with areas that varied between 720 and 1830 m of altitude. The annual rainfall ranges from 8.25 to 223.19 mm concentrated mainly in the wet season (October– March), although some rain may occur during the dry season (April–September).
11
Trapping of animals and data recording The five didelphid species were captured weekly from February 2005 to March 2006 using 100 mesh-wire livetraps (30 cm 15 cm 15 cm) distributed along two tracks located 50 m apart from each other. Trapping stations were placed 20 m apart from each other on each track and traps were set on the ground and on tree branches or vines at height of about 1.5 m. All traps were baited with a mixture of banana, pineapple, and cotton balls soaked with cod-liver oil during the afternoon and checked for captures during the following morning. All animals were marked with numbered ear tags (Zootech, Curitiba, PR, Brazil) and released. Scats on the trap floor and those defecated by the individuals during the manipulation were collected. One fecal sample was considered as being all feces produced by one animal in a night.
Diet analysis Scats collected were stored in numbered small plastic containers and frozen to avoid deterioration. In the laboratory, the scats samples were washed in a 0.1 mm mesh sieve. Food items were assigned to the following categories: seeds, other plant parts (flower and fruit fragments), arthropods, vertebrates and unidentified material. Items were identified at the lowest possible taxonomic category by comparison with a reference collection of invertebrates and seeds from the study area. When present, seeds were counted in each fecal sample. The bait residue was detected and excluded during the scats analyses. The relative frequency of occurrence (Korschgen 1987) was used to determine the contribution of each item in all food categories to the diet of didelphids. This statistic was calculated as the frequency of scats containing a particular food item multiplied by 100 (Korschgen 1987). The non-parametric Mann–Whitney U-test (Ayres and Ayres Jr. 2000) was used to test diet differences between the dry and the wet season. The G-test was used to compare the proportion of seeds and arthropod groups in the diet of the didelphids (Sokal and Rohlf 1995).
Germination tests Seeds found in scats were placed in closed Petri dishes (diameter ¼ 9 cm) containing wet absorbing paper to verify germination and monitored weekly. As control experiments, seeds of the same plant species consumed by the didelphids were collected directly from fruiting plants and tested for germination as described above. To test differences between germination rates of control group versus seeds found in the scats we use the w2 statistic (Sokal and Rohlf 1995).
ARTICLE IN PRESS 12
L.G. Lessa, F.N. da Costa / Mamm. biol. 75 (2010) 10–16
Results Diet A total of 157 fecal samples were obtained: G. agilis (n ¼ 56); M. incanus (n ¼ 58); M. nudicaudatus (n ¼ 11); M. paraguayanus (n ¼ 22) and C. philander (n ¼ 10) (Table 1). Arthropods were the main food items consumed and occurred in 100% of M. incanus and M. nudicaudatus samples. Hymenopterans (Formicidae) occurred in 26.8% of G. agilis samples and in 40% of C. philander. Coleopterans (Scarabeidae) were the most frequent item found for M. paraguayanus. The presence of Dipterans was detected by pupae remains found for M. incanus and G. agilis. Isopterans (Termitidae) were also largely recorded in samples of all species. Vertebrate preys (bones and feathers fragments of birds) were found with low frequencies for M. nudicaudatus and M. paraguayanus (Table 1). A total of 13 different morphotypes of seeds and other plant parts were found. The vegetal material belonging to the families: Anacardiaceae, Lauraceae, Melastomataceae, Myrtaceae, Hypericaceae, Rubiaceae and Meliaceae were identified. Seeds of Clidemia urceolata and Miconia holosericea (Melastomataceae) were found with high frequencies in samples of all didelphids (Table 1). Seeds ingested ranged from 1 to 5 mm in length. Larger seed quantities were found for C. philander (Ocotea lancifolia: n ¼ 1200 seeds; C. urceolata: n ¼ 880 seeds), M. nudicaudatus (C. urceolata: n ¼ 1204 seeds) and M. incanus (C. urceolata: n ¼ 824 seeds). Fragments of flowers were also seen in samples of M. incanus (Lauraceae) and C. philander (unidentified). No significant differences were observed in the diet between the dry and the wet season for M. incanus (M–W. U ¼ 77.00, p ¼ 0.37); G. agilis (M–W. U ¼ 25.00, p ¼ 0.11); M. nudicaudatus (M–W. U ¼ 32.00, p ¼ 0.32); and M. paraguayanus (M–W. U ¼ 81.50, p ¼ 0.89). In contrast, the diet of C. Philander differed between the dry and wet season (M–W. U ¼ 32.00, p ¼ 0.01), largely as a result of the higher consumption of fruits in the wet season. Arthropods in the diet were more frequent just for two species, M. incanus (G ¼ 6.11, df ¼ 1, p ¼ 00.1) and G. agilis (G ¼ 5.99, df ¼ 1, p ¼ 00.1).
Seed germination Germination rates of seeds from scats were higher than those of control groups (Table 2). Germination tests revealed that for the smaller didelphids such as G. agilis and M. incanus, all seeds that remained viable were less than 1 mm long (e.g., C. urceolata, M. holosericea and Psychotria barbiflora). For the larger species such as M. paraguayanus, M. nudicaudatus and C. philander, the largest intact seeds belonged to
Ocotea lancifolia (3 mm) and Amaioua guianensis (5 mm). Seeds greater than 5 mm were found destroyed (e.g., Tapirira obtuse 7 mm). Fruits of pioneer plants belonging to the families Melastomataceae (C. urceolata and M. holocericea) and Rubiaceae (P. barbiflora and P. capitata) presented high germination rates (Table 2). Fruits of C. urceolata were consumed in high proportions and germination rates of seeds from scats were significantly higher than those of control groups for the five studied species: M. incanus (w2 ¼ 128.5992, df ¼ 1, po0.01); G. agilis (w2 ¼ 13.8161, df ¼ 1 po0.01); M nudicaudatus (w2 ¼ 60.6644, df ¼ 1, po0.01); M. paraguayanus (w2 ¼ 59.2268, df ¼ 1, po0.01) and C. philander (w2 ¼ 45.0309, df ¼ 1, po0.01) (Table 2).
Discussion Food habits Our data showed that the didelphids sampled in a cerrado reserve in Brazil presented a remarkable constancy in their diets among the seasons. Just for C. philander, a seasonal change in diet seems to be present with a high frequency of fruit consumption during the wetter season. Arthropods such as hymenopterans and coleopterans were the main food resource detected. A similar pattern was also reported by Santori et al. (1995) for a restinga (shrubland) forest and Ca´ceres (2002) for a mixed ombrophilous forest in southern Brazil. Nevertheless, the relative abundance of Hymenoptera and Coleoptera in tropical forests would make these groups the most frequent prey in the diet of neotropical didelphids and does not necessarily imply that food selection is taking place (Pinheiro P.S. et al. 2002). This relative abundance may be related with two factors. First, an overestimation of this item because hard-bodied preys are easier to be detected in scats samples (Dickman and Huang 1988). Second, the high abundance of ants and beetles in the study area, although we did not estimated the availability of this items in the study site. Gracilinanus agilis is the smallest of the five species considered in this study. Its diet, based on the relative frequency of occurrence of prey in scats presented fruits and a high consumption of arthropods, mainly hymenopterans and isopterans. This diet composition is similar to that reported by the congener species, G. microtarsus (Talamoni et al. 1999; Martins and Bonato 2004; Martins et al. 2006). The high consumption of ants and termites, the most abundant insects in the cerrado (Pinheiro F. et al. 2002), could reveal a particularly high availability of these social insects in the study area and thus reflect an opportunistic food choice
ARTICLE IN PRESS L.G. Lessa, F.N. da Costa / Mamm. biol. 75 (2010) 10–16
13
Table 1. Relative frequency of occurrence (%) of food items found in the scats of five didelphids caught in a gallery forest from February 2005 to March 2006 in the PERP, Brazil. Food items
Seeds Anacardiaceae Tapirira obtusa (Benth) J. D. Mitchell Lauraceae Ocotea lancifolia (Schott) Mez Ocotea sp. Melastomataceae Clidemia urceolata DC. Miconia holosericea (L.) DC. Unidentified Melastomataceae Myrtaceae Myrcia sp. Unidentified Myrtaceae Hypericaceae Vismia glaziovii Ruhland Rubiaceae Amaioua guianensis Aubl. Psychotria barbiflora DC. Psychotria capitata Ruiz & Pav. Unidentified Rubiaceae Unidentified seeds Seeds (total) Fruits fragmentsa Meliaceae Unidentified Flowers fragmentsb Lauraceae Unidentified Vegetal fragments (total) Arthropoda Diplopoda Arachnidac Orthopterad Coleopterae Blattodea Hymenopteraf Isopterag Diptera (pupa) Hemypterah Unidentified Arthropoda Arthropoda (total) Vertebrates Birds (feathers) Vertebrates (total)
Gracilinanus agilis (n ¼ 56) – – – – – – – 5.3 1.8 – – – – – – – 3.6 3.6 1.8 7.1 19.6 – – 3.6 – – – 3.6 – – – 1.8 1.8 – 26.8 10.8 1.8 1.8 50.0 84.0 – – –
n ¼ total number of scats analyzed. a Pericarp. b Anthers and petals. c Araneidae and Acarina. d Gryllidae. e Scarabeidae and unidentified Coleoptera. f Formicidae. g Termitidae. h Heteroptera and Homoptera.
Marmosops incanus (n ¼ 58)
Metachirus nudicaudatus (n ¼ 11)
Micoureus paraguayanus (n ¼ 22)
Caluromys philander (n ¼ 10)
– – –
– – –
– – 4.5
– – –
– – – – 12.0 5.1 3.4 – – – – – – – – 3.4 1.7 3.4 29.3 – 1.7 8.7 – 3.4 5.1 19.0 – – 1.7 5.1 13.8 8.7 66.0 13.8 10.3 12.0 24.1 100.0 – – –
– – – – 27.2 9.0 – – – 9.0 – – – – – – – – 45.5 – – – – – – – – 18.1 9.0 36.4 45.5 45.5 72.8 45.5 – – 18.1 100.0 – 18.1 18.1
– – – – 9.0 9.0 4.5 – 9.0 – – – – 13.6 4.5 – – 4.5 45.5 – – 4.5 – – 22.7 27.3 – – 9.0 9.0 36.3 – 22.7 4.5 – 22.7 45.4 91.0 – 9.0 9.0
– 10.0 10.0 – 30.0 20.0 10.0 – 20.0 – – 20.0 – – – 10.0 – 10.0 90.0 – – – – 10.0 – 10.0 – – 10.0 – 30.0 – 40,0 10.0 – – 10.0 70.0 – – –
ARTICLE IN PRESS 14
L.G. Lessa, F.N. da Costa / Mamm. biol. 75 (2010) 10–16
Table 2. Number of seeds in test and frequencies of germination (%) of seeds found in the scats of five didelphids from a gallery forest in the PERP, Brazil. Family/species
Control experiments
Marmosops incanus
Gracilinanus Metachirus agilis nudicaudatus
Micoureus paraguayanus
Caluromys philander
Anacardiaceae Tapirira obtusa (Benth) J. D. Mitchell Lauraceae Ocotea lancifolia (Schott) Meza Ocotea sp.a Melastomataceae Clidemia urceolata DC.a Miconia holosericea (L.) DC.a Unidentified Melastomataceae Myrtaceae Myrcia sp.a Unidentified Myrtaceae Hypericaceae Vismia glaziovii Ruhlanda Rubiaceae Amaioua guianensis Aubl. Psychotria barbiflora DC.a Psychotria capitata Ruiz & Pav.a Unidentified Rubiaceae
– 6 (33)
– –
– –
– –
– 4 (50)
– –
16 (37)
– –
– –
– –
– –
– 100 (51)
45 (22) 45 (26)
– – 298 (92) 48 (81)
– – 22 (73) –
– – 120 (87) 10 (90)
– – 60 (97) 64 (55)
30 (73) – 130 (79) 10 (40)
25 (12)
10 (100)
10 (60)
–
–
–
– 30 – – 10 – 10 10 10
– – – – – – – – 8 (75)
– – – – – – – 3 (67) 4 (50)
– – – – – – – – –
– 30 (73) – – – – 9 (44) 15 (80) –
– 10 (60) 4(50) – 10 (60) – – – 6 (60)
–
2 (50)
–
–
–
a
28 (25)
(20)
(10) (20) (20) (20)
10 (50)
Pioneer plants.
by G. agilis (see Martins et al. 2006, for seasonality data of G. microtarsus). Marmosops incaus is also a small didelphid and an insectivorous–omnivorous diet is indicated for all species within the genus (Fonseca et al. 1996). However, a preliminary analysis on the relative frequency of occurrence of arthropods among the dry and the wet season revealed a remarkable constancy on the consumption of arthropods, mainly hymenopterans. Despite the high consumption of arthropods, present in 100% of the fecal samples, fruits seem to be an important resource. A similar pattern was found by Leiner and Silva (2007) working in an area of Brazilian Atlantic forest for the congeneric species M. paulensis, suggesting the importance of this item as an energetic supplement in both seasons (Julien-Laferie`rre and ´ stua Atramentowics 1990; Pinheiro P.S. et al. 2002; A de Moraes et al. 2003). The consumption of flowers, especially during the dry season was observed for M. paulensis (Leiner and Silva 2007) and the same trend would occur here for M. incanus, but to a lesser extent. Metachirus nudicaudatus seems to be more insectivorous, consuming a reasonable number of invertebrates, as was expected for a terrestrial didelphid (Santori et al. 1995). However, the local food availability could
determine variations in the diet of certain species (Ca´ceres et al. 2002). Although M. nudicaudatus has been regarded as an insectivorous species (Santori et al. ´ stua de Moraes 2003), 1995; Ca´ceres 2004; Vieira and A fruits of pioneer plants such as Clidemia urceolata occurred at a relatively high proportion (45.5%) in scats collected in the cerrado (present study) but were detected in low proportions in restinga (10.5%) and Atlantic forest (7%) studies (Santori et al. 1995; Ca´ceres 2004, respectively). Micoureus paraguayanus has been reported as insectivorous–omnivorous, in general consuming large quantities of insects, a selected group of fruits and rarely vertebrates (Leite et al. 1996; Ca´ceres et al. 2002; Pinheiro P.S. et al. 2002; Casella and Ca´ceres 2006) which is in agreement with our results, although it also fed on fruits and flowers fragments in a reasonable proportion (45.5% and 22.7%, respectively). As was emphasized by Santori et al. (1995), ‘‘there is a tendency towards the insectivory in smaller neotropical marsupial species’’ although, sometimes the consumption of vertebrate prey such as birds has been reported by this specie (Ca´ceres et al. 2002; Casella and Ca´ceres 2006; this study). Caluromys philander had a more frugivorous diet (Leite et al. 1996; Casella and Ca´ceres 2006) as was
ARTICLE IN PRESS L.G. Lessa, F.N. da Costa / Mamm. biol. 75 (2010) 10–16
15
expected for an arboreal species (Casella and Ca´ceres 2006), but arthropods were found in its diet at high frequencies (70%). Julien-Laferie`rre (1999) showed the importance of insects as nitrogen resources for C. philander in French Guiana. Coleopterans, hymenopterans and arachnids were important prey items for this species, mainly during the drier season. Fruits of pioneer plants such as Miconia holosericea, Ocotea lancifolia and Myrcia sp. were mainly consumed during the wetter season. Caluromys philander seems to be frugivorous but also use other food items in different proportions, the same trend would occur for the congeneric C. lanatus (Ca´ceres 2005; Casella and Ca´ceres 2006). These results are in agreement with Fonseca et al. (1996) that proposed a frugivorous–omnivorous diet for Caluromys spp.
regenerating several types of environments through seed dispersal (Pinheiro P.S. et al. 2002; Ca´ceres et al. 2002; Leiner and Silva 2007), given that the necessary conditions of movement and appropriateness of site of feces/seed deposition are satisfied.
Role in seed dispersal
References
Previous studies detected a tendency towards insectivory in smaller neotropical didelphids (Santori et al. 1995; Martins and Bonato 2004); however, fruits seem to be an important supplement on its diets (Julien´ stua de Laferie`rre 1999; Pinheiro P.S. et al. 2002; A Moraes et al. 2003; Leiner and Silva 2007; this study) also for the more insectivorous species, such as G. agilis, M. incanus and M. nudicaudatus. The five didelphids studied fed on fruits in varying frequencies, mainly fruits of pioneer plants. Moreover, seeds consumed remained viable after passing undamaged through the digestive tract of the animals and presented a high germination rate. As evidenced in other studies (Ca´ceres and Monteiro-Filho 2000; Ca´ceres 2002; Casella and Ca´ceres 2006), there seems to be a relation among the seed size and the animal body size. For the smaller didelphids sampled such as G. agilis and M. incanus, just the small seeds (1 mm) remain viable. For the largest species such as M. nudicaudatus, M. paraguayanus and C. Philander, large seeds (between 3 and 5 mm) remain viable after cross undamaged the gut. The potential role of didelphid marsupials on seed dispersal, especially for small seeds belonging to pioneer plants, has been show in disturbed and non-disturbed Atlantic forests remnants (Leite et al. 1996; Ca´ceres et al. 2002; Pinheiro P.S. et al. 2002; Casella and Ca´ceres 2006; Leiner and Silva 2007). However, just a few studies confirmed this potential role with germination tests (Grelle and Garcia 1999; Ca´ceres and MonteiroFilho 2000; Ca´ceres 2002). Our results confirmed that the germination rates of seeds from scats are higher than germination rates of seeds collected from fruits (control experiments), especially for pioneer species belonging to the families Melastomataceae and Rubiaceae. These results stress the importance of neotropical didelphids in
´ stua de Moraes, D., Santori, R.T., Finotti, R., Cerqueira, R., A 2003. Nutritional and fibre contents of laboratory-established diets of neotropical opossums (Didelphidae). In: Jones, C., Dickman, J.M., Archer, M. (Eds.), Predators with Pounches: The Biology of Carnivorous Marsupials. CSIRO Publishing, Melborne, pp. 229–237. Ayres, M., Ayres Jr., M., 2000. Bioestat 5.0: aplicac¸o˜es estatı´sticas nas a´reas das cieˆncias biolo´gicas e me´dicas. Sociedade Civil Mamiraua´, CNPq, Bele´m. Ca´ceres, N.C., 2002. Food habits and seed dispersal by the white-eared-opossum, Didelphis albiventris (Mammalia, Marsupialia) in southeastern Brazil. Stud. Neotrop. Fauna Environ. 37, 97–104. Ca´ceres, N.C., 2004. Diet of three didelphid marsupials (Mammalia, Didelphimorphia) in southeastern Brazil. Mam. Biol. 69, 430–433. Ca´ceres, N.C., 2005. Comparative lengths of digestive tracts of seven didelphid marsupials (Mammalia) in relation to diet. Rev. Bras. Zool. 22, 181–185. Ca´ceres, N.C., Monteiro-Filho, E.L.A., 2000. The common opossum, Didelphis aurita, as a seed disperser of several plants in southern Brazil. Cieˆnc. Cult. 52, 41–44. Ca´ceres, N.C., Ghizoni, I.R., Graipel, M.E., 2002. Diet of two marsupials, Lutreolina crassicaudata and Micoureus demerarae, in coastal Atlantic Forest island of Brazil. Mammalia 66, 331–340. Casella, J., Ca´ceres, N.C., 2006. Diet of four small mammals species from Atlantic forest patches in south Brazil. Neotrop. Biol. Cons. 1, 5–11. Dickman, C.R., Huang, C., 1988. The reliability of fecal analysis as a method for determining the diet of insectivorous mammals. J. Mammal. 69, 108–113. Eiten, G., 1992. Natural Brazilian vegetation types and their causes. Ann. Acad. Bras. Cieˆnc. 64 (Suppl. I), 35–65. Fonseca, G.A.B., Herrmann, G., Leite, Y.R.L., Mittermier, R.A., Rylands, A.B., Patton, J.L., 1996. Lista anotada de mamı´feros do Brasil. Occas. Pap. Cons. Biol. 4, 1–38. Grelle, C.E., Garcia, Q.S., 1999. Potential dispersal of Cecropia hololeuca by the common opossum (Didelphis
Acknowledgements We thank Soˆnia Talamoni and Lena Geise for suggestions to the early version of the manuscript. IEF-MG (Minas Gerais State Forest Institute) for allowing the access at PERP. We are also indebted to two anonymous reviewers for their comments and suggestions that greatly improved and clarify the manuscript. The study was supported by FAPEMIG (process number CRA 133-03).
ARTICLE IN PRESS 16
L.G. Lessa, F.N. da Costa / Mamm. biol. 75 (2010) 10–16
aurita) in Atlantic Forest, southeastern Brazil. Rev. Ecol. (Terre Vie) 54, 327–332. Julien-Laferie`rre, D., 1999. Foraging strategies and food partitioning in the neotropical frugivorous mammals Caluromys philander and Potos flavus. J. Zool. (London) 247, 71–80. Julien-Laferie`rre, D., Atramentowics, M., 1990. Feeding and reproduction of three didelphid marsupials in two neotropical forests (French Guiana). Biotropica 22, 404–415. Korschgen, J.L., 1987. Procedimientos para el ana´lisis de los ha´bitos alimentarios. Manual de te´cnicas de gestio´n de vida silvestre. The Wildlife Society, Washington, DC. Leiner, N.O., Silva, W.R., 2007. Seasonal variation in the diet of the Brazilian slender opossum (Marmosops paulensis) in a montane Atlantic Forest area, southeastern Brazil. J. Mammal. 88, 158–164. Leite, Y.R.L., Costa, L.P., Stallings, J.R., 1996. Diet and vertical space use of three sympatric opossums in a Brazilian Atlantic Forest reserve. J. Trop. Ecol. 12, 435–440. Lessa, L.G., Costa, B.M.A., Rossoni, D.M., Tavares, V.C., Dias, L.G., Moares-Junior, E.A., Silva, J.A., in press. Mamı´feros da Cadeia do Espinhac¸o, riqueza, ameac¸as e estrate´gias para conservac¸a˜o. Megadiversidade. Martins, E.G., Bonato, V., 2004. On the diet of Gracilinanus microtarsus (Marsupialia, Didelphidae) in an Atlantic rainforest fragment in southeastern Brazil. Mam. Biol. 69, 58–60. Martins, E.G., Bonato, V., Pinheiro, H.P., Dos Reis, S.F., 2006. Diet of the gracile mouse opossum (Gracilinanus microtarsus) (Didelphimorphia: Didelphidae) in a Brazilian Cerrado: patterns of food consumption and intrapopulation variation. J. Zool. (London) 269, 21–28.
Myers, N., Mittermeier, R.A., Mittermeier, C.G., Fonseca, G.A.B., Kent, J., 2000. Biodiversity hotspots for conservation priorities. Nature 403, 853–858. Pinheiro, F., Diniz, I.R., Coelho, D., Bandeira, M.P.S., 2002. Seasonal pattern of insect abundance in the Brazilian cerrado. Aust. Ecol. 27, 132–136. Pinheiro, P.S., Carvalho, F.M.V., Fernandez, F.A.Z., Nessimian, J.L., 2002. Diet of the marsupial Micoureus demerarae in small fragments of Atlantic Forest in southeastern Brazil. Stud. Neotrop. Fauna Environ. 37, 213–218. Rossi, R.V., Bianconi, G.V., Pedro, W.A., 2006. Ordem Didelphimorphia. In: Reis, R.R., Peracchi, A.L., Pedro, W.A., Lima, I.P. (Eds.), Mamı´feros do Brasil. Parana´, Londrina, pp. 27–60. Santori, R.T., Moraes, D.A., Cerqueira, R., 1995. Diet composition of Metachirus nudicaudatus and Didelphis aurita (Marsupialia, Didelphoidea) in southeastern Brazil. Mammalia 59, 511–516. Sokal, R.R., Rohlf, F.J., 1995. Biometry. W.H. Freeman and Company, New York, 887pp. Talamoni, S.A., Couto, D., Lopes, M.O.G., Cordeiro-Ju´nior, D.A., 1999. Dieta de algumas espe´cies de pequenos mamı´feros do sudeste brasileiro. Bios 7, 51–56. Talamoni, A.S., Couto, D., Ju´nior, D.A.C., Diniz, F.M., 2007. Diet of some species of neotropical small mammals. Mam. Biol. ´ stua de Moraes, D., 2003. Carnivory and Vieira, E.M., A insectivory in neotropical marsupials. In: Jones, C., Dickman, J.M., Archer, M. (Eds.), Predators with Pounches: The Biology of Carnivorous Marsupials. CSIRO Publishing, Melborne, pp. 267–280.