SHORT NOTES HERPETOLOGJCAL JOURNAL, Vol. 12, pp. 179-182 (2002) the territories surveyed cf. Luiselli & Angelici, 2000), situated in the easternaxis of the Niger Delta (Bayelsa DIET OF THELOTORNIS KIR TL ANDII and Rivers States), in Anambra, Akwa-Ibom, Abia and (SERPENTES: COLUBRIDAE: Cross River States. The study region is tropical, with a DISPHOLIDINI) FROM SOUTHERN wet season from May to September and a dry season fromOctober to April. The wettest period of the year is NIGERIA June-July, and the driest period between late December and February. GODFREY AKANI LUCA LUISELLI AND C. 1, 2•3 Methods used to survey the study area, capture FRANCESCO ANGELIC! M. 2·3 snakes and analyse their food items, are detailed else­ where (e.g. Luiselli et al., 1998, 2002). Fieldwork was 1 Department of Biological Sciences, Rivers State University of Science and Technology, P.M.B. 5080, Port conducted under all climatic conditions, but with a bias Harcourt (R ivers State), Nigeria toward diurnalhours (from 0800-1800 hrs) due to secu­ rity constraints related to the prevailing unstable Centre of Environmental Studies "Demetra ", Via dei 2 political situation. Field effort was almost identical in Cochi 48/B, I-00133 Rome, Italy the wet (421 field days) and dry (418 field days) sea­ 3 F.I.Z. V., via Olona 7, I-00198 Rome, Italy sons. We searched for snakes along standardized routes in Key words: Africa, feeding ecology, snake the various microhabitats frequented by snakes at the study areas. We captured snakes by hand, but additional Th elotornis kirt/andii is an arboreal, venomous free-ranging specimens were captured by pitfall traps colubrid snake with a wide distribution ranging fromthe with drift fences and by traps used by locals to capture islands of the Bijagos Archipelago, Guinea Bissau; east terrestrial animals. We always recorded the site of cap­ through forested areas of west Africaand the Congo ba­ ture and the habitat at each capture site. Each snake was sin to Uganda and southern Sudan; south to northern measured for snout-vent length (SVL, to the nearest 0.1 Angola, north-westernZambia and south-central Tanza­ cm), weighed with an electronic balance, and individu­ nia (for review, see Broadley, 2001). It is typically a ally marked by ventral scale clipping for future forest species (Broadley, 2001 ), and in southernNigeria identification. Then, the snakes were palpated in the ab­ may attain high population densities in mangrove habi­ domen until regurgitation of ingested food or defecation tats (Luiselli Akani, 2002). & occurred. In addition, specimens found already dead Although the diet of its savanna congener, during our surveys (e.g. snakes killed by farmers, or by Th elotornis capensis, is well known (Shine et al., 1996), cars, etc) were dissected to determine if prey were fe w data are available on the foodhabits of kirtlandii, T. present. We identified prey items to the lowest taxo­ possibly due to its elusive habits and relatively inacces­ nomic level possible. We estimated mass of prey items sible rainforest and mangrove habitats. Apart for at the time of its ingestion, when possible, by comparing general comments in books (e.g., "this snake feeds on the item to intact conspecifics of various sizes from our lizards, birds, etc.", e.g. see Phelps, 1981), there are own personal collection, or measuring the fresh biomass very few precise dietary records in the literature: this in perfectly preserved items. This was not possible in snake was found to prey upon nestling birds faecal samples, which generally consisted of scales of (Spermophaga ruficapilla), agamids (Acanthocercus reptiles. Although in this article we used data collected atricollis), and scincids (Mabuya maculilabris) in the from both stomachs and faeces, specimens for which Democratic Republic of Congo (Loveridge, 1942); stomach contents were used were not generally used for upon lacertids (lchnotropis capensis) in Zambia faecal pellets, to avoid multiple counts of the same food (Broadley, 1991); and upon colubrid snakes items. However, the foodcontents fromboth faeces and (Philothamnus carinatus) in Zaire DCR (Bogert, 1940). stomachs of a same snake specimen were considered in From a swamp-rainforest of south-eastern Nigeria, the cases in which, for instance, faecal samples con­ Luiselli et al. ( 1998) recorded eight food items: one tained mammal hair and the stomach a gecko). small bird ( Cisticola galactotes) and seven geckos Vouchers (of both prey and predators) were depos­ (Hemidactylus fa sciatus). Here we present a detailed ited in the herpetological collections of the Rivers State account of the diet of free-ranging kirtlandii froma T. University of Science and Technology (Port Harcourt, region situated within the continuous Guinea-Congo Nigeria), of the Department of Agricultural Sciences of rainforestbelt (i.e. southernNigeria, West Africa),with Uyo University (Uyo, Nigeria), of the Faculty of Sci­ an analysis of prey-size predator-size relationships. ences of the University of Calabar (Calabar, Nigeria), of Our fieldstudy was carried out fromSeptem ber 1996 the Ecology Lab at F.l.Z.V. (Rome, Italy), and of the to May 2002 (with additional data recorded in 1994 and Cross River National Park (collections in both Akamkpa 1995), in some localities of south-eastern Nigeria (for and Butatong, Nigeria). All statistical tests were two-tailed, with alpha set at Correspondence: L. Luiselli, F.I.Z.V., via Olona 7, 1-00 198 Rome, Italy. E-mail: [email protected] 0.05. Means are followedby ±1 SD. 180 SHORT NOTES A total of 133 specimens (44 smaller than 50 cm total sions differed significantly (heterogeneity of slopes length - assumed to be subadults in this study - and 89 test: F =4.139, P=0.026). 1.30 larger than 50 cm total length) were examined for the To the best of our knowledge, there are no other sci­ present study (see Appendix l for a list of the localities of entific reports on the diet of a large number of capture). Maximum size was 122.5 cm for males, and free-ranging specimens of T. kirtlandii from a single 131.4 cm for females. When only specimens >50 cm in geographic region (although Shine et al., 1996, pre­ length are considered, males and femaleswere similar in sented a very detailed dietary study on the savannah terms of mean length (males: 89.8±17.4 cm, n=39; fe­ congener T. capensis). Thus, it may be assumed that the males: 91.2± 18.3 cm, n=41; differences between two present study, being the largest of its type available for samples: one-way ANOVA F =2.384, P>0.3). T. kirtlandii, may have permitted finer (and perhaps 1• 77 A list of prey items is presented in Table 1. Based on different) conclusions from those already available in this table, it is evident that 100% of the dietary spectrum the literature. Nonetheless, at least in general terms, of specimens >50 cm in length consisted of vertebrates, this study confirms previous literature suggesting that whereas a very few arthropod items were found in juve­ lizards account for a remarkable portion of the diet of niles. It is not known whether these arthropod remains free-ranging T. kirtlandii. It is interesting, however, were primary feeding records, or secondary ingestion that no chameleons were eaten by these snakes, al­ following the complete digestion ofthe original predator, though they were regularly found in congeners from e.g. a gecko or amphibian. It is likely that the second op­ elsewhere (e.g. T. capensis, see Shine et al., 1996; T. tion would be more likely, as (1) specimens of mossambicanus, D. Broadley, pers. comm.). lt perhaps Th elotornis kirtlandii kept in captivity never fed upon reflects a global scarcity of these potential prey in the invertebrates; (2) these prey items were almost com­ rainforest habitats of southern Nigeria (Akani et al., pletely digested; and (3) there are no records supporting 2002). On the contrary, gekkonids (and to a lesser ex­ primary arthropod feedingby Th elotornis (e.g. see Shine tent also Scincidae) obviously represented the primary et al., 1996). Amongst vertebrates, nearly all the prey prey sources for both adult and subadult snakes. Our items were reptiles, with a single case of avian prey (the data also demonstrate that rainforest T kirtlandii do not same record was presented in Luiselli et al., 1998) and a have as wide a dietary spectrum as their savannah con­ single case of amphibian prey. Amongst reptiles, snakes geners (Shine et al., 1996), which are known to forage were rarely eaten, whereas lizards - and particularly frequently also on amphibians (especially Breviceps, gekkonids - formed the main prey. but including Chiromantis xerampelina) and on large Female body size significantly positively influenced lizards like Acanthocercus atricollis and Gerrhosaurus the size of the prey eaten (Spearman's r2=0.312, spp., in addition to snakes and mammals) (D. Broadley, ANOVA: F =6.882, P=0.0189), whereas the same re­ pers. comm.). Amphibian and snake prey were ex­ 1•18 lationship was not found in males (r2 =0.103, ANOV A: tremely rare in forest T. kirtlandii from Nigeria. F = 0.108, P=0.594). The slopes of these two regres- However, it is noteworthy that Lawson ( 1993) recorded 1•14 TABLE I. Compilation of the original data available on food items of Th elotornis kirtlandii from south-eastern Nigeria. Prey Numbers of prey items Numbers of prey items Cumulative number of in specimens <50 cm length in specimens >50 cm length prey items Invertebrates Mantodea 2 0 2 Birds Cisticola galactotes 0 Reptiles Gastropyxis smaragdina 0 J I Philothamnus sp. 0 2 2 Mabuya sp. 7 J J J8 Panaspis sp. J 0 I Agama agama 4 7 J l Hemidactylus fa sciatus 2 8 JO Lygodactylus sp.
Details
-
File Typepdf
-
Upload Time-
-
Content LanguagesEnglish
-
Upload UserAnonymous/Not logged-in
-
File Pages4 Page
-
File Size-