Adult Tree Mortality Research Articles

Landscape-scale disturbances and regeneration in semi-arid woodlands of southwestern Australia

Woodlands dominated by Eucalyptus salmonophloia occur both in the fragmented landscapes of the Western Australian wheatbelt and in the adjacent unfragmented goldfields area. We examined the responses of the unfragmented woodlands to landscape-scale disturbances caused by fire, floods, windstorms and drought. Sites known to have experienced disturbances of these types over the past 50 years all had cohorts of sapling-stage E. salmonophloia and other dominant Eucalyptus species. Sites disturbed either by fire, flood or storm during 1991-92 displayed adult tree mortality and extensive seedling establishment, although rates of establishment and survival varied between sites. No regeneration was observed at equivalent undisturbed sites. These results indicate that landscape-scale disturbances of several types are important drivers of the dynamics of these semi-arid woodlands. Lack of regeneration of fragmented woodlands in the wheatbelt is likely to be due to changed disturbance regimes coupled with altered physical and biotic conditions within remnants. We argue that it may be difficult to identify processes which are important for the long-term persistence of natural ecosystems in fragmented landscapes without reference to equivalent unfragmented areas.

Fruiting Patterns, Seed Germination, and Distribution of Sclerocarya caffra in an Elephant-Inhabited Woodland

This study, conducted in the Luangwa Valley, Zambia, examines the ecological influence of elephants on a population of Sclerocarya caifra and certain demographic features of this same population. Patterns of fruiting and seed germination are described, with an experimental design for the latter to test elephants' role as a seed dispersal agent. Size class distributions and spatial patterns of trees are also investigated. Results suggest that seed germination and seedling survivorship are enhanced from fruit ingestion by elephants. Frequency of size classes in the population indicate that younger trees are relatively scarce and that mortality of seedlings and saplings may be sufficiently high to prevent recruitment. The presence of seedlings in an area where the composition of mammalian browsers differs from the study area suggests a species other than elephant may be contributing to the low numbers of small trees. Dispersion pattern of trees was strongly aggregated and correlated with well-drained deep soils. Interpretations of these results focus on the ecological relationship of S. caifra with elephants and the management requirements of S. caifra in Luangwa Valley. SCLEROCARYA CAFFRA (ANACARDIACEAE) is a species indigenous to Africa and a favored food plant of the elephant (Van Wyk & Fairall 1969, Coetzee et al. 1979). Its fruit is readily taken and extensive bark and branch browsing is also incurred. It is a medium-sized tree, reaching 15 m under favorable conditions (Palgrave 1983). Flowers have separate sexes on the same or different trees. From March to June large, fleshy fruits up to 3.5 cm in diameter and approximately 42 g in weight, ripen and fall to the ground with as many as 8000 fruits/tree. The entire fruit is consumed by the elephant, with its large stone, containing 2 to 3 seeds, passing through the elephant's digestive tract. No other recorded wild vertebrate that eats these fruits disperses the seeds away from the tree. Despite the general claim that elephants have caused a decline of woodland habitats in Luangwa Valley (Astle et al. 1969, Caughley 1976), the lack of life history data for those tree populations utilized by elephants prevents full substantiation. This applies to S. caifra, whose adult trees succomb to elephant browsing by either debarking or tree felling in Luangwa Valley (pers. obs.) and whose populations elsewhere may even be threatened from overbrowsing (Coetzee et al. 1979). Besides adult tree mortality, other demographic processes that might influence S. caifra's response to browsing by elephants (e.g., fruiting patterns, germination success) have been largely unstudied. Using both observational and experimental data, this paper provides information on the ecological basis of coexistence between elephants and S. caifra. Coexistence is likely influenced by mechanisms involving tree recruitment. For this reason and because elephants are the principal seed-dispersing agent, the role of elephants on seed germination and seedling survivorship was investigated. In particular, the following questions were asked: Does any relationship exist between median number of fruits found in an elephant bolus and the observed growth rate of seedlings from boluses with varying number of fruits? Does the ingestion and passage of fruits by an elephant in any way facilitate germination? Is germination facilitated from some property of the dung itself, which houses the seeds? Other variables considered that might influence coexistence were edaphic factors, mortality on age classes less affected by elephant browsing, and fruiting behavior of S. caifra. Results pertain not only to possible ecological relationships between elephants and S. caifra but also to the management of this tree species in elephant-inhabited woodlands. THE SITE AND VEGETATION The sampled tree population included all S. caifra trees within 15 km2 of the Lupande Research Area on the east bank of the Luangwa River (31.380 E, 13.22? S; see Fig. 1) at an altitude of approximately 540 m. Twelve distinct soil types are classified for this area (Table 1); vegetation characteristics for these soil types are given by Astle et al. (1969). The 15-km2 area includes soil types and vegetation classes representive of the entire Luangwa Valley (Astle et al. 1969) because of the compressed catena of I Received 4 January 1985, revision accepted 24 October 1985. 2 Mailing address: % Nyamaluma Camp, Box 18, Mfuwe, Zambia. 50 BIOTROPICA 19(1): 50-56 1987 This content downloaded from 40.77.167.32 on Fri, 13 May 2016 05:13:01 UTC All use subject to http://about.jstor.org/terms