Trophic Associations of a Dung Beetle Assemblage (Scarabaeidae ...

COMMUNITY AND ECOSYSTEM ECOLOGY

Trophic Associations of a Dung Beetle Assemblage (Scarabaeidae: Scarabaeinae) in a Woodland Savanna of Botswana B. POWER TSHIKAE,1 ADRIAN L. V. DAVIS,

AND

CLARKE H. SCHOLTZ

Department of Zoology and Entomology, University of Pretoria, Lynwood Road, Pretoria 0002, South Africa

Environ. Entomol. 37(2): 431Ð441 (2008)

ABSTRACT Species richness and abundance of dung beetles were assessed across a range of bait types that acted as surrogates for the food resources available in Chobe National Park, Botswana. These bait types were comprised of the dung of pig (omnivore), cattle (ruminant herbivore dropping Þne-Þberd pads), sheep (pellet-dropping ruminant herbivore), and elephant (monogastric, nonruminant herbivore producing coarse-Þbered droppings), and chicken livers (carrion). Species richness was similar between traps baited with pig, cattle, and elephant dung but was relatively lower in those baited with sheep dung and carrion. In traps baited with pig dung, abundance was relatively greater than in all other bait types. A cluster analysis of species abundance distributions for the 30 most abundant species identiÞed four different patterns of bait type association at a 60% level of similarity. All but 1 of the 15 species in cluster A were attracted primarily to the dung of omnivores and pad-dropping ruminant herbivores (pig and cattle). All seven species of cluster B were attracted primarily to coarse-Þbered, nonruminant herbivore dung (elephant). All four species of cluster C were primarily carrion and pig dung associated, whereas all four species of cluster D were carrion specialists. In conclusion, the most abundant species were attracted to all bait types, but most species were largely specialized to different dung types or carrion, with dung attracting the majority of the fauna in terms of both species richness and abundance. KEY WORDS African savannas, Botswana, bait types, trophic associations, Scarabaeinae

Dung beetles of the subfamily Scarabaeinae are a functionally important group of invertebrates that use both dung and carrion (Estrada et al. 1993, Davis 1994, Martin-Piera and Lobo 1996, Vernes et al. 2005). Because they have a long history of ecological specialization to feeding and breeding in dung (Davis et al. 2002), their diversity is tightly linked to mammal dung diversity (Hanski and Cambefort 1991). Empirical evidence indicates that a large and diverse mammal fauna is crucial for maintenance of a large and diverse dung beetle fauna (Klein 1989, Hanski and Cambefort 1991, Estrada et al. 1998). Generally, the principal dung types available in African savannas are omnivore dung (bush pig, monkey, and baboon); ÞneÞbered dung of ruminant herbivores dropped as pads (buffalo and cattle); pellets of ruminant herbivores (wildebeest, impala, sable, kudu, gemsbok, sheep, and goat); and coarse-Þbered dung of nonruminant herbivores (elephant, rhinoceros, horse, donkey, and zebra). Therefore, the savannas of Africa are well suited for studies of trophic resource partitioning because they show a greater mammal and dung beetle diversity than any other single region in the world (Hanski and Cambefort 1991, du Toit and Cumming 1999).

1

Corresponding author, e-mail: [email protected].

Dung and carrion are ephemeral resources with a patchy distribution in space and time (Hanski and Cambefort 1991). Although vision is essential for ßight navigation in search for fresh dung (McIntyre and Caveney 1998), dung beetles locate their food and breeding resource primarily by olfactory cues (Dormont et al. 2004) that are provided by the many chemical compounds that are released from dung as volatiles, either in small or large quantities (Yasuhara et al. 1984). Thus, in areas where many species of mammals occur, dung odors constitute an important niche dimension (Gittings and Giller 1998). Both dung-frequenting ßies (Mulla and Ridsdill-Smith 1986) and dung beetles respond selectively to these chemical components. Because different odors are produced by different food types (Davis 1994), dung beetles are able to select one dung type or the other. This may vary with the purpose of use: feeding versus nesting (Kryger et al. 2006). In addition to odor proÞle, dung beetles partition their food and breeding resource according to its physico-chemical attributes. These include water content, Þber size, dropping size, and nutritional quality (Halftter and Matthews 1966, Mulla and RidsdillSmith 1986, Davis 1989, Hanski and Cambefort 1991, Sowig and Wassmer 1994, Sowig 1997, Gittings and Giller 1998, Dormont et al. 2004). These qualities vary

0046-225X/08/0431Ð0441$04.00/0 䉷 2008 Entomological Society of America

432

ENVIRONMENTAL ENTOMOLOGY

Vol. 37, no. 2

Fig. 1. Location map showing the Chobe National Park in the northeast of Botswana.

according to animal digestive system, body size, and diet (Davis 1989, Edwards 1991, Estrada et al. 1993, Gittings and Giller 1998). Dung beetle trophic association has been previously studied in various ecosystems, i.e., Central America (Estrada et al. 1993), South Africa (Davis 1994), Europe (Martin-Piera and Lobo 1996), Australia (Hill 1996, Vernes et al. 2005), and South Asia (Borghesio et al. 2001). However, the only study (Davis 1994) in southern Africa, which describes trophic preference in Afro-tropical dung beetles species in savanna woodland, was conducted in an agro-ecosystem where only two principal dung types were available (pad-dropping herbivore [cattle] and omnivore [baboons]). Therefore, this study assesses diversity and trophic preference of dung beetles across a range of dung types in a natural woodland savanna region in Botswana, with a full complement of indigenous African mammals dropping a full range of dung types. This study intends to provide preliminary information for selecting bait types suitable for dung beetle faunal inventory. Materials and Methods Study Area and Sites. The study was conducted at an altitude of 936 Ð1,033 m a.s.l. in the Chobe National Park (Fig. 1), northeast Botswana (17⬚54⬘52.4⬙ S, 25⬚01⬘09.4⬙ E; hereafter Chobe NP) in December 2004. Chobe NP covers 1,057,000 ha of patchy and relatively undisturbed Baikiaea plurijuga (Harms) woodlands on deep Kalahari sands. It lies on the southern bank of the Chobe River near its conßuence with the Zambezi River and stretches southward to the inland delta of

the Okavango and Linyanti river systems, much of which is also conserved. Together these areas represent a large area of pristine sandveld (a region characterized by sandy soils) into which ßow several large river systems. Chobe National Park has a rich mammal fauna in comparison to other localities in Botswana (T. Taolo, personal communication). The wide variety of large mammals includes elephant (Loxodonta africana) zebra (Equus burchelli), buffalo (Syncerus caffer), blue wildebeest (Connochaetes taurinus), tsessebe (Damaliscus lunatus), baboon (Papio ursinus), vervet monkey (Cercopithecus aethiops), and several species of buck and antelopes (reed buck [Redunca arundinum], bush buck [Tragelephus scriptus], puku [Kobus vardonii], impala [Aepyceros melampus], sable [Hippotragus niger], and kudu [Tragelaphus strepsiceros]). Based on regional rainfall patterns, Chobe NP lies in the midsummer rainfall region of southern Africa (Davis 1994, Davis and Scholtz 2004) and is characterized by a hot wet summer season from October to April followed by the dry winter spell from May to July. Annual rainfall ranges from 600 to 690 mm, with high variability in space and time. Heavy downpours are often followed by a strong sunshine causing loss of moisture by evaporation and transpiration. Periodic droughts also occur. The summer temperature ranges from 19 to 33⬚C and winter from 5 to 23⬚C. Two trapping sites were chosen in Chobe NP. Site 1 (dense woodland) was situated along a single track from the main road to the Chobe River, and site 2 (open woodland) was along a Þre break from the main road to the river. Even though the study area constituted a single habitat, these sites were considered to

April 2008

TSHIKAE ET AL.: TROPHIC ASSOCIATION IN DUNG BEETLES

represent good spatial replicates because they were located 5 km apart. Bait Resources. Dung beetle associations with bait type were studied using naturally decaying animal matter (chicken offal) and four different dung types. The four dung types consisted of (1) nonÞbrous strong-smelling pig dung as a surrogate for the dung of omnivores (baboons and monkey); (2) Þne-Þbered dung of ruminant herbivores (cattle) as a surrogate for buffalo dung; (3) pellets of ruminant herbivores (sheep) as a surrogate for pellet producers (impala, sable, kudu); and (4) coarse-Þbered dung of a nonruminant herbivore (elephant). Surrogates were used for all dung types except elephant because it was difÞcult to Þnd other suitable dung types and carrion in sufÞcient quantities. Dung for baits was collected from an intensively managed pig farm, from pasture grazing cattle and sheep on the University of Pretoria Experimental Farm, and from elephant droppings found on the roads in Chobe NP. Elephant dung was collected fresh in the park, whereas chicken offal and the other dung types were collected fresh, made into baits, and deep frozen. After defrosting, chicken offal was permitted to decay for 6 h in the sun before use. Trapping and Collection of Samples. Sampling was conducted in December 2004, during the season of high dung beetle activity (Davis 2002). Dung beetles were sampled using 25 baited pitfall traps at each site. Traps were placed 50 m apart along the access tracks and 25 m into the bush from either edge of the track. They thus formed a 2 by 12 ⫹ 1 grid over a distance of 600 m. Trapping was standardized according to microhabitat by placing traps in similar less shaded situations because dung beetles are strongly inßuenced by insolation (Doube 1983, Steenkamp and Chown 1996), and the greatest proportion of savanna species occurs in unshaded situations (Davis 1994). Baits were randomly allocated to the traps, either ⬇250 ml of dung (elephant, cattle, sheep, or pig) or 50 ml of carrion (chicken livers), each wrapped in chiffon to exclude dung beetles. Trapping was done at each site on a single 24-h occasion after substantial rainfall (40 mm) in Þne, sunny weather. Traps were baited between 0600 and 0800 h and rebaited between 1500 and 1800 h so that both diurnal and crepuscular/ nocturnal species were presented with fresh baits. Samples were preserved in 70% alcohol for identiÞcation and counting. Voucher specimens will be deposited at the University of Pretoria Insect Collection and the Botswana National Museum. Data Analysis. Rank abundance curves were used to compare abundance patterns and species evenness across bait types, and the IndVal method (Dufreˆ ne and Legendre 1997, see also van Rensburg et al. 1999, McGeoch et al. 2002) was used to determine indicator or characteristic species for each dung type and carrion. This is expressed as IndValij ⫽ Aij ⫻ Bij ⫻ 100, in which the speciÞcity measure, Aij ⫽ Nindividualsij/ Nindividualsi, is the abundance of species i in bait type j divided by the total abundance of species i across all Þve bait types, and the Þdelity measure, Bij ⫽ Ntrapsij/ Ntrapsj, is the number of traps in which species i is

433

present in bait type j divided by the total number of traps in bait type j (n ⫽ 10). A species was classiÞed as a specialist for a certain bait type if the IndVal for that particular bait type was ⬎70%. Those with an IndVal range between 50 and 70% were considered as feeding with a degree of preference, whereas species with an IndVal between 5 and 50% were classiÞed as generalists. Trophic association was only determined for the adequately sampled species (i.e., ⱖ25 individuals) because anything less gave an unclear picture. Trophic associations have been classiÞed by arranging the trap data as a matrix of 30 species by total numbers attracted to each of Þve bait types. The data matrix included only species with an overall abundance of ⬎25 individuals. A Bray-Curtis similarity matrix was calculated from the species data matrix and was subjected to cluster analysis using the agglomerative method of group average linking (PRIMER v5.0; Clarke and Warwick 2001). The results were summarized using a dendrogram from which groups of species with similar trophic associations were deÞned. A simple correspondence analysis (SCA) (StafSoft 2004) was used to determine the species association pattern with each bait type. The null hypothesis assumed that species associations were independent of bait type. The results were depicted as an ordination plot using a two-dimensional map construction, where groups of species drawn to a particular bait type are plotted closer to the bait and to each other on the map than those that are highly dissimilar. Results Assemblage Structure. A total of 69,075 beetles were trapped in Chobe NP, representing 67 species and 8 tribes (Appendix 1). Of the tribes represented in Africa, only the Gymnopleurini were absent from the present data set. The 67 species of beetles were comprised of 52 (78%) in cattle, 51 (76%) in pig, 44 (66%) in elephant, 37 (55%) in sheep dung, and 22 (33%) recorded in carrion (Fig. 2b; Appendix 1). Dung attracted more individuals than carrion. Pig dung baits attracted 38,670 (56%) individuals, followed by cattle with 16,393 (23.7%). The other two dung types, elephant and sheep, attracted 7,773 (11.3%) and 5,564 (8%) individuals, respectively. Only 675 (1%) individuals preferred carrion baits (Fig. 2a; Appendix 1). Rank abundance curves of dung beetle species trapped in Chobe NP (Fig. 3) show the patterns of species diversity in the dung and carrion baits. Dominance was lower in sheep dung and carrion than in pig, cattle, and elephant dung. Although Caccobius species were most abundant in pig, cattle, and elephant dung, Metacatharsius opacus was dominant in carrion-baited traps. The curves show greater evenness among species with intermediate (⬎10 ⬍ 100) and low (⬍10) abundance. Trophic Association. Individuals of most beetle species were found in dung-baited traps. When all sampled beetles were considered, exclusive feeding was encountered in all bait types. However, when beetle species represented by few specimens (⬍25 individ-

434


Vol. 37, no. 2

Fig. 2. Number of species (a) and number of individuals (b) trapped on different bait types.

uals or ⬍0.01% of the total abundance) were excluded, only one species was found exclusively in one type of bait, Catharsius melancholicus (41 individuals), collected only on carrion (Table 1). There were in general more species with IndVal of ⬎70% in carrion than in all dung types (Table 2). Among the dung types, only pig dung had a species with IndVal ⬎70%. Characteristic species in carrion had higher IndVal than the one in pig dung. The majority of species showed some degree of preference between different dung types (IndVal 50 Ð70%), although there were nine generalists, three carrion specialists, and one pig dung specialist. Faunal similarity at different sites is summarized by the dendrogram (Fig. 4) and indicates that dung beetle assemblages were strongly structured by food type. For each bait, the fauna was most similar to that attracted to the same bait type at the other site. Species abundance patterns in cattle and pig dung were very

similar, with those in sheep and elephant dung showing greater differences. There was extreme dissimilarity between dung and carrion community composition (Fig. 4). The dendrogram in Fig. 5 shows similarities in bait association of the 30 most abundant species. It supports the Þndings in Fig. 4 in that one major group is associated with carrion baits and another with dung baits at the 40% level of similarity. However, at the 60% level of similarity, four clusters were shown on the dendrogram (Fig. 5), indicating more detailed association of beetles with various bait types. All but 3 of 15 taxa in the species-rich cluster A were associated primarily with pig dung. However, the whole group showed a distribution skewed to pig and cattle dung baits (Fig. 6A). Most species formed a single cluster, whereas Caccobius nigritulus occured as an outlier because of the relatively large number of individuals (Appendix 1). The second species-rich cluster B was

Fig. 3. Rank-abundance curves for dung beetle species on Þve bait types (Eleph, elephant; H, Shannon-Weiner; E, evenness).

April 2008 Table 1.


435

Proportion showing bait specificity of dung beetle species collected in numbers of 25 or >25 across the bait types Percentage on bait type

Species

Species code

Carrion

Pig

Elephant

Cattle

Sheep

Total

H (4, N ⫽ 50)

Anachalcos convexus Caccobius cavatus Caccobius ferrugineus Caccobius nigritulus Catharsius melancholicus Cleptocaccobius convexifrons Drepanocerus laticollis Euonthophagus sp. 1 Kheper lamarcki Metacatharsius opacus Metacatharsius sp. 1 Metacatharsius troglodytes Onthophagus anomalus Onthophagus plebejus Onthophagus quadraticeps Onthophagus signatus Onthophagus sp. 2 Onthophagus sp. 3 Onthophagus sp. 4 Onthophagus sp. k Onthophagus sp. nr probus Onthophagus sp. nr pullus (a) Onthophagus sp. nr variegatus Onthophagus verticalis Onthophagus vinctus Onthophagus virescens Pachylomerus femoralis Pedaria sp. 1 Scarabaeus flavicornis Scarabaeus zambesianus

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z AA ab AC AD

39.1 0.0a 0.1a 0.0a 100.0a 0.1ac 0.0a 0.0 0.0a 61.3a 91.3a 1.9a 0.0a 0.0 0.0a 0.0b 98.6a 98.6 25.6ab 0.0a 0.0a 0.1ac 3.8 0.0a 0.0a 0.0a 3.5a 0.0 64.0a 0.0a

26.1 40.4b 28.4bc 70.7b 0.0b 51.3b 7.6ab 47.1 40.4b 19.8ab 2.2b 35.1b 27.9ab 13.2 39.4b 43.8a 1.4b 0.0 39.6a 57.0b 47.5b 55.1b 46.2 14.5a 18.6b 41.2b 30.3b 8.8 30.0ac 17.4bc

4.3 16.1b 44.5bc 12.4bc 0.0b 4.9c 69.3b 14.3 11.0ab 5.0b 0.0b 5.8ab 53.6b 68.4 10.1ab 3.7bc 0.0b 0.0 21.8ab 0.0a 2.7a 6.7c 7.7 47.4bc 41.3b 10.9b 6.5ac 67.6 0.0b 4.6ac

19.6 26.7b 20.2ac 15.8bc 0.0b 28.3bc 20.3b 38.6 42.1b 7.5b 6.5ab 30.2b 15.9ab 13.2 25.6b 38.1a 0.0b 1.4 9.9ab 32.9ab 23.9ab 27.6bc 30.8 25.0ac 37.5bc 31.8b 37.6b 8.8 5.0bc 47.6b

10.9 16.8ab 6.7a 1.1ac 0.0b 15.4bc 3.0a 0.0 6.5a 6.4b 0.0b 27.0ab 2.6a 5.3 24.8b 14.4c 0.0b 0.0 3.0b 10.1ab 25.9ab 10.6bc 11.5 13.2a 2.7ac 10.0ab 22.1cb 14.7 1.0b 30.5bc

46 7603 820 31204 41 3823 301 70 907 439 46 430 308 38 355 7994 73 74 101 79 1203 8403 26 76 264 2856 766 34 100 328

6.26 28.93*** 28.93*** 44.06*** 48.23*** 35.24*** 26.31*** 17.08 40.14*** 21.67** 33.37*** 17.28* 22.8** 10.08* 21.83** 38.07*** 44.84*** 27.92 14.13* 24.39*** 29.4*** 33.99*** 9.36 22.24** 27.49*** 29.75*** 36.39*** 11.8 30.88*** 25.26***

Values in a row followed by a different letter are signiÞcantly different: *P ⬍ 0.001, **P ⬍ 0.01, ***P ⬍ 0.05.

comprised of seven species in which all but one was attracted primarily to elephant dung, with a distribution strongly skewed to elephant dung baits (Fig. 6B). Cluster C consisted of four species that associated primarily with carrion, with a distribution skewed to carrion and pig dung baits (Fig. 6C). Cluster D constituted four species, all carrion specialists (Fig. 5; Table 1), with a distribution strongly skewed to carrion baits (Fig. 6D). The SCA indicated that two dimensions signiÞcantly explained 92% of the difference in species association with bait type (␹2 dimension 1 ⫽ 2,313.50, P ⫽ 0.00001, df ⫽ 116; ␹2 dimension 2 ⫽ 887.58, P ⫽ 0.00001, df ⫽ 116; Fig. 7). It showed that species distribution is dependent on bait type. It also showed the formation of three groups important in trophic association. Statistically, the third contributed little to proportional variance because relatively few individuals were involved compared with the Þrst two groups. However, biologically, species contributing to the third group are radically different to those contributing to the Þrst two. Discussion This study is considered preliminary because it was limited both temporally and spatially. Nonetheless, the results clearly indicated that dung beetle assemblages in Chobe NP are comprised of both specialist and generalist species. As in other studies (Davis 1994,

Martin-Piera and Lobo 1996), strong trophic partitioning was shown by this study on deep sand in a southern African reserve, but the manner of partitioning differed to some extent from that shown on deep sand in a savanna agro-ecosystem (Davis 1994). With high mammal diversity being a characteristic feature of Chobe NP, dung category (herbivore versus omnivores) was expected to be a signiÞcant factor in resource partioning, where species would segregate according to locally available trophic resources (Davis 1994). However, many species of Scarabaeinae showed niche overlap on distantly related dung types (cattle, sheep, and pig). Patterns of trophic partitioning ranged from indiscriminate attraction to omnivore and ruminant herbivore dung, to a clear association with nonruminant herbivore dung (elephant) through to a specialized association in carrion fauna. This is not the only study where perhaps lack of dichotomy between species attracted to omnivore and ruminant dung was observed. A rather similar trophic association pattern was observed by Botes et al. (2006) and MartinPiera and Lobo (1996) in other biogeographical regions. In the former, the suites of species recorded in the human-cattleÐ dominated habitat were equally attracted to the two dung types, whereas in the elephant-dominated habitat, a suite of different species showed a clear elephant dung association. Similarly, in the latter study, a different suite of species were again equally attracted to human and

436 Table 2.


Vol. 37, no. 2

Percentage indicator value (IndVal) of Scarabaeidae species for five bait types comparisons Species

Carrion

Pig

Elephant

Cattle

Sheep

Anachalcos convexus Boheman Caccobius cavatus dÕOrbigny Caccobius ferrugineus Fahraeus Caccobius nigritulus Klug Catharsius melancholicus Boheman Cleptocaccobius convexifrons Raffray Drepanocerus laticollis Fahraeus Euonthophagus sp. 1 Kheper lamarcki Macleay Metacatharsius opacus Waterhouse Metacatharsius sp. 1 Metacatharsius troglodytes Boheman Onthophagus anomalus Klug Onthophagus plebejus Klug Onthophagus quadraticeps Harold Onthophagus signatus Fahraeus Onthophagus sp. 2 Onthophagus sp. 3 Onthophagus sp. 4 Onthophagus sp. k Onthophagus sp. nr probus Onthophagus sp. nr pullus (a) Onthophagus sp. nr variegatus Onthophagus verticalis Fahraeus Onthophagus vinctus Erichson Onthophagus virescens Harold Pachylomerus femoralis Kirby Pedaria sp. 1 Scarabaeus flavicornis Boheman Scarabaeus zambesianus Peringuey

23.48 0.00 0.01 0.00 100.00 0.01 0.00 0.00 0.00 61.28 82.17 1.12 0.00 0.00 0.00 0.01 98.63 69.05 15.45 0.00 0.00 0.02 0.38 0.00 0.00 0.00 3.17 0.00 57.60 0.00

13.04 40.43 28.41 70.68 0.00 51.35 4.58 28.29 40.35 13.87 0.22 28.09 19.55 5.26 39.44 43.85 0.14 0.00 35.64 39.87 47.55 55.05 27.69 4.34 14.85 41.04 30.29 1.76 24.00 13.90

0.87 16.11 44.51 12.38 0.00 4.89 62.19 7.14 9.92 4.01 0.00 4.07 48.21 34.21 8.11 3.65 0.00 0.00 8.71 0.00 1.86 6.70 1.54 47.37 41.29 10.85 6.53 40.59 0.00 2.29

9.78 26.66 16.20 15.80 0.00 28.33 16.21 23.14 42.12 6.01 1.96 27.21 9.55 3.95 25.63 38.07 0.00 0.14 2.97 26.33 21.47 27.57 12.31 12.50 26.25 38.13 37.60 2.65 1.50 47.56

4.35 16.78 1.34 1.12 0.00 15.38 1.20 0.00 5.85 5.10 0.00 24.28 0.26 0.53 17.35 14.41 0.00 0.00 0.89 3.04 20.75 10.59 2.31 1.32 1.33 8.98 22.06 1.47 0.10 24.39

Bold IndVals denote specialist species with IndVal ⬎ 70%.

cattle dung. The reason for the remarkably similar trend in association patterns between ruminant and omnivore dung assemblages from different biogeographical regions remains unknown.

The results showed that there is a distinct dung and carrion fauna of Scarabaeinae captured in Chobe NP. However, species showing copro-necrophagous or necrophagous feeding habits are in the minority, be-

Fig. 4. Dendrogram for cluster analysis of dung beetle species abundance data for Þve bait types at two sampled sites in Chobe National park (S1, site 1; S2, site 2).

April 2008


437

Fig. 5. Dendrogram for cluster analysis of the most abundant species trapped with Þve different bait types in Chobe National Park (AÐAD are species codes; full species names and authority are given in Table 2).

cause the majority fed on dung. The high proportion of coprophagous species seems to be a predominant feature of Afro-tropical dung beetle assemblages, which contrasts with the greater proportions of largely necrophagous dung beetles found in the Asian and Neotropical regions (Hanski and Cambefort 1991,

Hanski and Krikken 1991). This observation is also consistent with the general view of Halftter and Matthews (1966) on the trophic structure of Afro-tropical dung beetles, that few species make use of carrion because of the presence of large carnivores and vultures. Furthermore, the results of this study are sup-

Fig. 6. Percentage distribution of numbers across Þve different bait types in four clusters (AÐD) of dung beetle species deÞned from a dendrogram (Fig. 5) at the 60% level of similarity and results for Kruskal-Wallis tests (columns with different letter differed signiÞcantly, P ⬍ 0.05).

438


Vol. 37, no. 2

Fig. 7. Ordination plot showing the statistical distance between food-association patterns in 25 dung beetle species (E) and their relative association with Þve different bait types (f): carrion plus cattle, elephant, pig, and sheep dung. Acon, Anachalcos convexus; Ccav, Caccobius cavatus; Cfer, Caccobius ferrugineus; Cnig, Caccobius nigritulus; Cmel, Catharsius melancholicus; Ccon, Cleptocaccobius convexifrons; Dlat, Drepanocerus laticollis; Esp1, Euonthophagus sp. 1; Klam, Kheper lamarcki; Mopa, Metacatharsius opacus; MspI, Metacatharsius sp. 1; Mtro, Metacatharsius troglodytes; Oano, Onthophagus anomalus; Ople, Onthophagus plebejus; Oqua, Onthophagus quadraticeps; Osig, Onthophagus signatus; Osp2, Onthophagus sp. 2; Osp3, Onthophagus sp. 3; Osp4, Onthophagus sp. 4; Ospk, Onthophagus sp. K; Opro, Onthophagus sp. nr probus; Opul, Onthophagus sp. nr pullus (a); Ovar, Onthophagus sp. nr variegates; Over, Onthophagus verticalis; Ovin, Onthophagus vinctus; Ovir, Onthophagus virescens; Pfem, Pachylomerus femoralis; Psp1, Pedaria sp. 1; Sfla, Scarabaeus flavicornis; Szam, Scarabaeus zambesianus.

ported by the Þndings of several studies on dung beetle trophic preferences carried out in the Palaearctic region (Martin-Piera and Lobo 1996, Al-Houty and Al-Musalam 1997, Barbero et al. 1999, Galante and Cartagena 1999), Australia (Hill 1996, Vernes et al. 2005), and Africa (Cambefort 1991, Davis 1994, Paetel 2002). They indicate that dung beetles feeding on ruminant, nonruminant plus omnivore dung, and carrion do so with a degree of preference. Caution is needed in interpreting the differences in abundance patterns between dung types obtained from standardized baits and those from the actual situation that would be encountered in naturally dropped dung pats. Finn and Giller (2000) concluded that density of beetles increased with patch size, suggesting that size differences between standardized dung baits and naturally dropped dung pats would undoubtedly inßuence trophic structure of beetles. In these data, species composition was similar across dung types, but abundance varied considerably with respect to dung types analyzed. Overall abundance was high in omnivore dung because of the exceptional abundance of the pig specialist (C. nigritulus); nonetheless, even when C. nigritulus was removed, abundance remained higher in pig dung than other dung types. This difference in distribution of numbers of beetles between locally occurring dung types has been

attributed to suites of physico-chemical characteristicsÑmost importantly size, volatile compounds, water content, and resource type (Peck and Howden 1984, Edwards 1991, Lumaret et al. 1992, Davis 1994, Sowig and Wassmer 1994, Hill 1996, Gittings and Giller 1998 Dormont et al. 2004, Errouissi et al. 2004). In fact, the inßuence of size and smell was clearly shown by Vernes et al. (2005), who captured more dung beetle individuals in bandicootsÕ fecal pellets, which were much larger and smell more strongly than pellets from two other mammals in the same locality. The results of this study are consistent with this observation because most individuals were found in pig dung, which smelled more strongly than other dung types. In addition to odor of pig dung, the size of the baits was relatively larger than the normally small and ßat droppings of the baboons and monkeys prevalent in the study area, whereas for the other dung types (elephant, buffalo), baits were much smaller than the normal pat size. Some studies on dung beetle trophic preferences suggest that generalist behavior (polyphagy) is a consequence of resource scarcity (Al-Houty and Al-Musalam 1997, Dormont et al. 2004). Although this might be true for some areas, this study was conducted in an area where most of the principal dung types and to a lesser extent carrion were in abundance, thus making

April 2008


it a suitable location to observe clear trophic preferences. Nonetheless, there was no evidence of exclusive preference to any one dung type. Conversely, exclusive preference has been recorded only in carrion, which is a very limited resource in savannas, because of the presence of large carnivores and vultures that leave little carrion available for beetles (Halftter and Matthews 1966, Martin-Piera and Lobo 1996, Barbero et al. 1999). However, numerous factors invariably affect fresh dung availability including the migration of mammals (Tribe 1976, Hanski and Cambefort 1991). Seasonal migration of mammals in Chobe NP between permanent and temporary wetlands that take place at the onset of rains may lead to dung resource scarcity or even the absence of fresh dung. This was noticeable because it was difÞcult to Þnd fresh elephant and buffalo dung during the trapping period. In addition, some dung beetle species (e.g., Kheper lamarcki) were frequently observed feeding on rehydrated elephant dung. Clearly, dung-baited traps were the only source of fresh food for dung beetles. Perhaps clear trophic preferences for most species may be observed if trapping is conducted throughout the dung beetle activity season. In conclusion, this study indicated that mammal diversity in Chobe NP is very important for the maintenance of the diverse dung beetle community because the most abundant species were attracted to all bait types, with dung attracting the majority of the fauna in terms of both species richness and abundance. Moreover, because of some specialized trophic associations on carrion, dung beetle studies should perhaps include carrion baits, because dung alone cannot be used to achieve an accurate inventory of dung beetle species. Although species abundances and patterns of trophic association were established for some species, only 45% of the trapped species were used in this analysis. Hence, there is a need for further study of the associations of less common species. Acknowledgments This research was Þnancially supported by the University of Pretoria, NRF, and UNDP-SGP in Botswana. We thank the Ministry of Environment Tourism and Wildlife, Department of Wildlife and National Parks, for granting permission to work in the protected areas and for generous assistance; F. Escobar for statistical input; M. Mokhatla, the Department of Zoology and Entomology, and the Scarab Research Group for support; and two anonymous reviewers for input.

References Cited Al-Houty, W., and F. Al-Musalam. 1997. Dung preferences of the dung beetle Scarabaeus cristatus Fabricius (Coleoptera: Scarabaeidae) from Kuwait. J. Arid Environ. 35: 511Ð516. Barbero, E. C., C. Palestrini, and A. Rolando. 1999. Dung beetle conservation: effects of habitat and resource selection (Coleoptera: Scarabaeoidea). J. Insect Conserv. 3: 75Ð84. Borghesio, L., C. Palestrini, and P. Passerin d’Entreves. 2001. Dung beetles of the Grand Paradiso National Park: a

439

preliminary analysis (Insecta: Coleoptera: Scarabaeoidea). J. Mount. Ecol. 6: 41Ð 8. Botes, A., M. A. McGeoch, and B. J. van Rensburg. 2006. Elephant and human-induced changes to dung beetle (Coleoptera: Scarabaeidae) assemblages in the Maputaland Centre of Endemism. Biol. Conserv. 139: 573Ð583. Cambefort, Y. 1991. Dung beetles in tropical savannas. In I. Hanski and Y. Cambefort (eds.), Dung beetle ecology, pp. 156 Ð178. Princeton University Press, Princeton, NJ. Clarke, K. R., and R. M. Warwick. 2001. Change in marine communities: an approach to statistical analysis and interpretation, 2nd ed. PRIMER-E, Plymouth, UK. Davis, A.L.V. 1989. Nesting of Afrotropical Oniticellus (Coleoptera: Scarabaeidae) and its evolutionary trend from soil to dung. Ecol. Entomol. 14: 11Ð21. Davis, A.L.V. 1994. Association of Afrotropical Coleoptera (Scarabaeidae: Aphodiidae: Staphylinidae: Hydrophilidae: Histeridae) with dung and decaying matter: implications for selection of ßy control agents for Australia. J. Nat. Hist. 28: 383Ð99. Davis, A.L.V. 2002. Dung beetle diversity in South Africa: inßuential factors. Conservation status, data inadequecncies and survey design. Afr. Entomol. 10: 53Ð 65. Davis, A.L.V., and C. H. Scholtz. 2004. Local and regional species ranges of a dung beetle assemblage from semi-arid Karoo/Kalahari margins South Africa. J. Arid Environ. 57: 61Ð 85. Davis, A.L.V., C. H. Scholtz, and T. K. Philips. 2002. Historical biogeography of scarabaeinae dung beetles. J. Biogeogr. 29: 1217Ð1256. Dormont, L., G. Epinat, and J. P. Lumaret. 2004. Trophic preferences meditated by olfactory cues in dung beetles colonizing cattle and horse dung. Environ. Entomol. 33: 370 Ð377. Doube, B. M. 1983. The habitat preference of some bovine dung beetles (Coleoptera: Scarabaeidae) in Hluhluwe Game Reserve, South Africa. Bull. Entomol. Res. 73: 357Ð 371. du Toit, J. T., and D.H.M. Cumming. 1999. Functional signiÞcance of ungulates diversity in African savannas and the ecological implications of the spread of pastoralism. Biodivers. Conserv. 8: 1643Ð1661. Dufreˆne, M., and P. Legendre. 1997. Species assemblages and indicator species: the need for a ßexible asymmetrical approach. Ecol. Monogr. 67: 345Ð366. Edwards, P. B. 1991. Seasonal variation in the dung of African grazing mammals, and its consequences for coprophagous insects. Funct. Ecol. 5: 617Ð28. Errouissi, F., S. Haloti, P. Jay-Robert, A. Janati-Idrissi, and J. P. Lumaret. 2004. Effects of the attractiveness for dung beetle of dung pat origin and size along a climatic gradient. Environ. Entomol. 33: 45Ð53. Estrada, A., G. Halffter, R. Coates-Estrada, and D. A. Meritt. 1993. Dung beetles attracted to mammalian herbivore (Alouatta palliata) and omnivore (Nasua narica) dung in the tropical rainforest of Los Tuxtlas, Mexico. J. Trop. Ecol. 9: 45Ð54. Estrada, A., R. Coates-Estrada, A. A. Dadda, and P. Cammarano. 1998. Dung and carrion beetles tropical rainforest fragments and agricultural habitats in Los Tuxtlas Mexico. J. Trop. Ecol. 14: 577Ð593. Finn, J. A., and P. S. Giller. 2000. Patch size and colonization patterns: an experimental approach using coprophagous north temperate dung beetles. Ecography 23: 315Ð327. Galante, E., and C. Cartagena. 1999. Comparisons of Mediterranean dung beetle (Coleoptera: Scarabaeoidae) in cattle and rabbit dung. Environ. Entomol. 28: 420 Ð 424.

440


Gittings, T., and P. S. Giller. 1998. Resource quality and the colonisation and succession of coprophagous dung beetles. Ecography 21: 581Ð592. Halftter, G., and E. G. Matthews. 1966. The natural history of dung beetles family Scarabaeinae (Coleoptera : Scarabaeidae), Folia Entomol. Mex. 12-. 14: 1Ð312. Hanski, I., and J. Krikken. 1991. Dung beetles in tropical forests in Southeast Asia. In I. Hanski and Y. Cambefort (eds.), Dung beetle ecology, pp. 179 Ð197. Princeton University Press, Princeton, NJ. Hanski, I., and Y. Cambefort. 1991 Resource partitioning. In I. Hanski and Y. Cambefort (eds.), Dung beetle ecology, pp. 330 Ð349. Princeton University Press, Princeton, NJ. Hill, C. J. 1996. Habitat speciÞcity and food preferences of an assemblage of tropical Australian dung beetles. J. Trop. Ecol. 12: 449 Ð 460. Klein, B. C. 1989. Effects of forest fragmentation on dung and carrion beetle communities on Central Amazonia. Ecology 70: 1715Ð1725. Kryger, U., K. S. Cole., R. Tukker, and C. H. Scholtz. 2006. Biology and ecology of Circellium bacchus (Fabricius 1781) (Coleoptera: Scarabaeidae), a South African dung beetle of conservation concern. Trop. Zool. 19: 185Ð207. Lumaret, J. P., N. Kadiri, and M. Bertrand. 1992. Changes in resources: consequences for the dynamics of dung beetle communities. J. Appl. Ecol. 29: 349 Ð356. Martin-Piera, F., and J. M. Lobo. 1996. A comparative discussion of trophic preferences in dung beetle communities. Misc. Zool. 19. 1: 13Ð31. McGeoch, M., B. J. van Rensburg, and A. Botes. 2002. The veriÞcation and application of bioindicators: a case study of dung beetles in a savanna ecosystem. J. Appl. Ecol. 39: 661Ð 672. McIntyre, P., and S. Caveney. 1998. Superposition optics and the time of ßight in Onitine dung beetles. J. Comp. Physiol. A. 183: 45Ð 60. Mulla, M. S., and T. J. Ridsdill-Smith. 1986. Chemical attractants tested against the Australian bush ßy Musca

Vol. 37, no. 2

vetustissima (Diptera: Muscidae), J. Chem. Ecol. 12: 261Ð 270. Paetel, C. 2002. Ecological aspects of the radiation of coprophagous Scarabaeoidea as “follow up” evolution of the evolutionary differentiation of ungulates. PhD thesis, Humboldt-Universita¨t zu Berlin, Berlin, Germany. Peck, S. B., and H. F. Howden. 1984. Response of a dung beetle guild to different size of dung bait in a Panamian rainforest. Biotropica 16: 235Ð238. Sowig, P. 1997. Niche separation in coprophagous beetles: a comparison of two multivariate approaches. Bull. Entomol. Res. 87: 625Ð 631. Sowig, P., and T. Wassmer. 1994. Resource partitioning in coprophagous beetles from sheep dung: phenology and microhabitat preferences. Zool. J. Syst. 121: 171Ð192. StatSoft. 2001. STATISTICA (data analysis software system), version 6 (www.statsoft.com). Steenkamp, H. E., and S. L. Chown. 1996. Inßuence of dense stands of an exotic tree, Prosopis glandulosa Benson, on a savanna dung beetle (Coleoptera: Scarabaeidae) assemblage in southern Africa. Biol. Conserv. 78: 305Ð311. Tribe, G. D. 1976. The ecology and ethology of ball-rolling dung beetles (Coleoptera: Scarabaeidae) MSc thesis, University of Natal, Pietermaritzburg, South Africa. van Rensburg, B. J., M. A. McGeoch, S. L. Chown, and A. S. van Jaarsveld. 1999. Conservation of heterogeneity among dung beetles in the Maputaland Centre of Endemism South Africa. Biol. Conserv. 88: 145Ð153. Vernes, K., L. C. Pope., C. J. Hill, and F. Barlocher. 2005. Seasonality, dung speciÞcity and competition in dung beetle assemblages in the Australian Wet Tropics, northeastern Australia. J. Trop. Ecol. 21: 1Ð 8. Yasuhara, A., K. Fuwa, and M. Jimbu. 1984. IdentiÞcation of odorous compounds in fresh and rotten swine manure. Agric. Biol. Chem. 48: 3001Ð3010. Received 12 June 2007; accepted 17 December 2007.

April 2008


441

Appendix 1. Abundance and overall percentage of 67 species of Scarabaeinae dung beetles trapped over 24 h in savanna woodland, Chobe National Park Tribe Canthonini Coprini

Dichothomini

Oniticellini

Onitini

Onthophagini

Scarabaeini

Sisyphini

a b

Species Anachalcos convexus Boheman Catharsius calaharicus Kolbe Catharsius melancholicus Boheman Copris elphenor Klug Copris evanidus Klug Copris vilhenai Ferreira Metacatharsius exiguus Boheman Metacatharsius opacus Waterhouse Metacatharsius sp. 1 Metacatharsius sp. 2 Metacatharsius troglodytes Boheman Heliocopris atropos Boheman Heliocopris japetus Klug Pedaria sp. 1 Pedaria sp. A Pedaria sp. VI Drepanocerus fastiditus Peringuey Drepanocerus freyi Janssens Drepanocerus laticollis Fahraeus Euoniticellus intermedius Reiche Euoniticellus sp. Oniticellus formosus Chevrolat Tiniocellus spinipes Roth Cheironitis sp. nr scabrosus Onitis alexis Klug Onitis deceptor Peringuey Onitis orthopus Langsberge Onitis viridulus Boheman Caccobius cavatus dÕOrbigny Caccobius ferrugineus Fahraeus Caccobius nigritulus Klug Caccobius sp. 1 Cleptocaccobius convexifrons Raffray Digitonthophagus gazella Fabricius Euonthophagus sp. 1 Hyalonthophagus alcyon dÕOrbigny Onthophagus anomalus Klug Onthophagus apiciosus dÕOrbigny Onthophagus flavolimbatus Klug Onthophagus plebejus Klug Onthophagus quadraticeps Harold Onthophagus signatus Fahraeus Onthophagus sp. 1 Onthophagus sp. 2 Onthophagus sp. 3 Onthophagus sp. 4 Onthophagus sp. 5 Onthophagus sp. 8 Onthophagus sp. B Onthophagus sp. k Onthophagus sp. nr probus Onthophagus sp. nr pullus (a) Onthophagus sp. nr variegatus Onthophagus suffusus Klug Onthophagus sugillatus Klug Onthophagus verticalis Fahraeus Onthophagus vinctus Erichson Onthophagus virescens Harold Phalops boschas Klug Proagoderus bicallosus Klug Kheper lamarcki Macleay Kheper prodigiosus Erichson Pachylomerus femoralis Kirby Scarabaeus anderseni Waterhouse Scarabaeus flavicornis Boheman Scarabaeus zambesianus Peringuey Sisyphus goryi Harold Total individualsa Total speciesb

Abundance from 10 traps Carrion

Pig

Elephant

Cattle

Sheep

Total

18 0 41 0 0 0 0 269 42 0 8 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 4 0 2 0 0 0 0 1 0 0 0 2 0 72 73 26 0 0 2 0 0 7 1 3 7 0 0 0 0 0 0 0 27 4 64 0 0 675a 22 (33)

12 3 0 0 0 0 1 87 1 6 151 1 2 3 6 0 2 1 23 1 5 0 1 0 0 0 0 0 3074 233 22056 3 1963 1 33 0 86 0 1 5 140 3505 0 1 0 40 1 6 1 45 572 4626 12 4 11 11 49 1172 2 9 366 7 232 10 30 57 0 38,670bc 51 (76)

2 2 0 0 1 1 0 22 0 1 25 1 3 23 0 8 19 17 208 5 1 0 2 0 3 1 6 1 1225 365 3864 0 187 3 10 0 165 0 0 26 36 292 0 0 0 22 0 0 0 0 32 563 2 1 0 36 109 310 2 0 100 6 50 0 0 15 0 7,773bc 44 (66)

9 1 0 1 0 1 0 33 3 1 130 0 1 3 7 0 0 2 61 1 1 0 1 0 1 1 0 0 2027 166 4930 1 1083 2 27 2 49 0 0 5 91 3043 2 0 1 10 0 5 2 26 287 2317 8 0 3 19 99 1089 4 1 382 3 288 2 5 156 0 16,393bc 52 (78)

5 0 0 2 0 2 0 28 0 1 116 0 0 5 7 0 0 0 9 2 0 1 0 7 0 0 0 0 1276 55 350 0 588 0 0 0 8 0 0 2 88 1152 0 0 0 3 0 3 0 8 312 890 3 0 2 10 7 285 0 1 59 6 169 0 1 100 1 5,564ac 37 (55)

46 6 41 3 1 4 1 439 46 9 430 2 6 34 20 8 21 20 301 9 7 1 4 7 4 2 6 1 7603 820 31204 4 3823 6 70 2 308 1 1 38 355 7994 2 73 74 101 1 14 5 79 1203 8403 26 8 23 76 264 2856 8 11 907 22 766 16 100 328 1 69,075 67

Abundance values with different letters are signiÞcantly different 关Kruskal-Wallis test: H (4, N ⫽ 335) ⫽ 29.67, P ⬍ 0.05兴. Numbers in parentheses refer to percentages of total species trapped in each bait type.

Percent 0.1 0.0 0.1 0.0 0.0 0.0 0.0 0.6 0.1 0.0 0.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 11.0 1.2 45.2 0.0 5.5 0.0 0.1 0.0 0.4 0.0 0.0 0.1 0.5 11.6 0.0 0.1 0.1 0.1 0.0 0.0 0.0 0.1 1.7 12.2 0.0 0.0 0.0 0.1 0.4 4.1 0.0 0.0 1.3 0.0 1.1 0.0 0.1 0.5 0.0