Results and discussion
- Wildlife reintroduction: considerations of habitat quality at the release site
Estimating density of fruit trees by simply dividing the number of trees by the area of the island ignores the between-species differences in the size distribution of adult trees and the intensity of production activities, thus biasing the result. To avoid this,  have developed a productivity index:
Im = Fkm bk dk
Where Fm = the fruiting score of all sampled individuals in species 'k' during month 'm'.
bk = mean dbh (diameter at breast height) of any adult tree for species 'k'
dk = density (number per ha) of adult trees for species 'k'
Diversity of fruiting trees was estimated using the Shannon-Weaver index and the frequency of each species in a block was compared to a Poisson distribution as a check for the random distribution of the species. Shannon-Weaver (based on a formula in ):
H = - pi (ln(pi))
where pi is the relative frequency of a species in a given sample.
The Shannon-Weaver index measures species richness. If only a few species account for most of the biomass, then the Shannon-Weaver index number is low. A reading of 2.0 would be a rich, diverse plant community. S-W Index values (H) can range from 0 to ~4.6 using the natural log (versus log10). A value near 0 would indicate that every species in the sample is the same. Conversely, a value near 4.6 would indicate that the numbers of individuals are evenly distributed between all the species present.
Mather (1992b)  compared several areas of Borneo and Peninsular Malaysia for relationships with their density and distribution. Based on this, I used Mathers's data for comparative purposes. By extrapolating data from Mather's sites which had gibbons, an estimate for potential numbers of gibbons/km2 on Mintin Island was calculated based on food tree and dipterocarp availability (Table 3), to determine the potential carrying capacity. By extrapolating from the other sites, an estimate for potential numbers of gibbons/km2 on Mintin Island was calculated. This was to determine the potential carrying capacity of the island based on the keystone species of figs . Much of the data from other sites were based on one variable only (fig density, dipterocarp density) and this was used to create a best-fit for Mintin Island. Figs are keystone species due to their asynchronous fruiting pattern [10-13] i.e. all trees do not all fruit at the same time, thus there are always some fig trees in fruit in the forest, making this genus an important fall-back food for gibbons in times of food shortage (Cheyne, unpublished data). Dipterocarps are important for gibbons for singing locations, morning calling is from dipterocarp tress in 74% of singing bouts (n = 1076; [14,15]. Dipterocarps are rarely primate feeding trees (except for leaves, , Cheyne, unpublished data) and their height is important for gibbons to broadcast their morning chorus and duet.
Direct observation of the canopy is the simplest and most convenient method in open and low forest, but problems arise when the method is employed in dense, tropical forest where emergent trees are often over 50 m high (). This was not a problem on Mintin Island, where the forest is maturing secondary vegetation with a maximum tree crown height of 23 m.
Based on Mather's  index, 17.1% of the trees on Mintin are in the 10 most important gibbon food genera (Table 2), and the percentage of may be a valuable index for comparing potential food productivity between sites. Results for Mintin, show varied levels of productivity for the species/genera, perhaps because Mintin is an island of recovering secondary forest and may have more varied and less predictable productivity than pristine forest .
Adaptation to life in the forest presents many obstacles for the gibbons and one of these should not be excessive food competition through poor management of the islands resources. This method does have limitations: the extrapolation only deals with data on gibbon densities based on the percentage of trees in the important food genera for gibbons and does not account for other fauna. E.g. Barito Ulu does not have orang-utans or large groups of macaques, whereas Mintin does have macaque groups. Thus, at Barito Ulu, gibbons may occur at higher densities, because there is less competition from other species. Other fauna e.g. macaques and hornbills will need figs when other fruit sources become scarce, so, competition from these animals may reduce the fruit available for gibbons and render my estimates too high. The relationship between the four sites is not strong, as evident by the result of the regression. To counter this I present a 95% confidence interval range for the number of gibbons that could be supported on Mintin i.e. the range at which the estimated mean value from the regression analysis could lie.
Flower production was lowest during the dry season from June to August 2003 and young leaf availability increased in the lead up to the dry season. Fruit production was erratic, and would probably result in the gibbons having to feed on trees not on Mather's list. This list does contain 40 genera on which gibbons feed, thus flexible diet allows them to adapt to food shortages.
A regression analysis was carried out on the available data for gibbon group size compared to the density of figs (R2 = 54.3 p = 0.015, n = 9). Figs comprised the main fruit source for two months during this study (January-February 2003) as there was very low availability of other sources of food e.g. non-fig fruit. As a result, figs are likely to be an important fruit source for the other frugivores on the island and there may be competition in times of scarcity.
Dipterocarps are not common food sources for gibbons but they are used for sleeping and singing. Density was calculated as the number of trees/hectare. Mintin Island also supports a density of 1.8 dipterocarp trees/ha, which are important for providing sleeping and singing platforms for the gibbons (Table 3).
Mintin Island is placed into Table 4 to show the expected group size and number of groups that it could support. The fig data appear to be very variable, hence the large confidence intervals.
Mintin Island has a Shannon-Weaver H-value of 2.98 indicating that the island hosts a rich and diverse plant and tree community. Pristine forest is inevitably going to support a higher diversity than logged, recovering forest but this indicates that the diversity of Mintin Island is not quite species rich and diverse, and is probably capable of supporting a similar diversity of fauna, albeit at lower densities.
Suitability of Mintin as a release site
Based on preliminary data (Table 5), Mintin Island should be able to support at least two pairs of gibbon with a total carrying capacity of 2.5–20 gibbons/km2 and group sizes of 1.5–3.5 individuals. The island should reach carrying capacity no later than the F2 generation, thus the offspring will need to be translocated to a larger site. This area has been identified as a 1500 ha forest near the Kalaweit Gibbon Sanctuary in Central Kalimantan.
 showed that there was a positive correlation between dipterocarp abundance and the biomass of gibbons a forest can support. Dipterocarps are an important structural component of the forest for gibbons as they provide platforms for sleeping and singing. There are few large dipterocarp trees left on Mintin, which, combined with the small area of the island, means that Mintin probably cannot support more than two groups, if dipterocarp density is indeed a limiting factor for the Mintin gibbons. It is possible that the lack of dipterocarps was a factor in explaining why the gibbons did not sing very often, though I believe that the lack of singing is more likely due to the fact that the pair was in very unfamiliar territory and that they separated immediately after release. When the gibbons did start to sing, they sang from dipterocarp trees (n = 6).
Most tropical, woody plants tend to produce new leaves and flowers in bursts rather than continuously, and there are considerable seasonal variation in the abundance of these food sources . This variation affects the abundance of primary consumers (e.g. frugivores and folivores) that will also have to modify their behaviours accordingly. If the year 2003 was a bad fruit year it is possible that there will be a bigger influx of animals to the island in better years, thereby increasing food competition with the gibbons. On the other hand, there may also be more food available if 2002/03 was lower than normal for fruit production, thus mitigating the effects of food competition. Clearly data are needed to determine exact levels of competition between the gibbons and other wildlife for the island's resources. This is a post-release issue and herein I discuss the need for detailed pre-release habitat analysis.
The island also supports groups of macaques and proboscis monkeys. Both these species are transient and were observed to swim from the island to the mainland (10 sightings of macaques in two groups and 8 sightings of proboscis monkeys in one group). They were observed on the island throughout the eight months that the gibbons were present, though not every day; thus, the macaques and proboscis monkeys can be considered as semi-permanent residents of Mintin Island. It is very important to collect data on the resource use by the macaques (Macaca nemestrina and M. fascicularis) and proboscis monkeys to effectively predict the resource availability for gibbons in the same area. Hornbills, being frugivorous, could be a potential competitor for gibbons given the limited production of fruit across the year. Therefore, feeding requirements of hornbills for limited fruit trees need to be identified and quantified for determining long term survival of the released gibbon populations in the proposed release site. Ideally, feeding rates can be collected on the released gibbons and the wild macaques to obtain data on food intake at different times of year and to observe how much overlap in diet there is between the two species. These data have to be collected post-release, and will be of great importance in the final phase of release, i.e. gibbons reintroduced to contiguous forest. This study does not include data on feeding rates and food intake, as the purpose is to address the issue of habitat quality assessment pre- and post-release. Feeding rates, food competition and food intake will form part of the post-release monitoring of gibbons due for reintroduction in 2007.
rating: 3.44 from 9 votes | updated on: 9 Jul 2007 | views: 11483 |