<back

 

Dr. Christian Schulze: Effects of anthropogenic disturbance on the diversity of herbivores – an analysis of moth species assemblages along habitat gradients in East Malaysia (2000)

Summary

Using two extremely species-rich moth taxa, i.e. Pyraloidea and Geometridae, as well as three less species rich groups, viz. Sphingidae, Arctiinae and Syntominae (both Arctiidae), as model taxa for tropical biodiversity research the present study dealed with the following topics: (1.) vertical stratification of moths within closed rain forest, (2.) alpha and beta diversity of moth species assemblages along gradients of anthropogenic disturbance, (3.) moth diversity in relation to vegetational structure and diversity, (4.) diversity of moth species assemblages along elevational gradients, (5.) importance of various habitats, including secondary forests, for Bornean endemics and (6.) suitability of selected moth groups for indicating diversity of different moth taxa at higher taxonomic levels.

The study was conducted in the years 1993-1999 at Mt Kinabalu National Park in the north-east of the Sundanian island of Borneo (Sabah, East Malaysia). Two gradients of disturbance were investigated, one at Poring and the other in Serinsim, both situated at the edge of the National Park in the transition between lowland forest and hill dipterocarp forest. Moths were attracted by light, using a gauze tower illuminated by a 15 W blacklight tube, and hand-sampled. The following numbers of specimens and species were available for analysis of diversity: Pyraloidea: 11.380 specimens (779 species), Geometridae: 4.558 specimens (500 species), Sphingidae: 373 specimens (30 species), Arctiinae: 354 specimens (16 species), Syntominae: 343 specimens (15 species).

Williams’ Alpha of the logarithmic series has been frequently used and recommended for the analysis of diversity of temperate-zone as well as tropical moth samples. However, with the present data base values of Williams’ Alpha depended strongly on sample size in the species-rich Pyraloidea and still not reached stable diversity estimates after 13 light-trapping nights per site at sample sizes as large as >2000 individuals. Hence, Hurlbert Rarefaction was used as principal tool for analysing alpha diversity of moth species assemblages. As measures for beta diversity, abundance-based Morisita and NESS indices yielded superior resolution than the incidence-based Sørensen index.

The following main results were obtained:

  1. Stratification within the closed dipterocarp forests could be demonstrated for moth taxa as well as guilds (as defined according to adult nutrient requirements). A higher abundance and species richness in the canopy was shown just for the Sphingidae. On the other hand, taxa like Geometridae and Syntominae (Arctiidae) exhibited higher diversity in primary forest understorey. Nectar-feeding moths showed highest abundances at canopy level. The proportion of Pyraloidea, Arctiinae and Geometridae species recorded at canopy level but noted in understorey of closed forest as well, ranged between 50.0 und 65.2%. Of the predominantly canopy-bound hawkmoths only 20% of the ‘canopy species’ could also be recorded from the ground, wheres among the Syntominae, which are poorly represented in the canopy, all species were recorded also at understorey level. An additional proportion of ‘canopy species’ of 9.9 (Pyraloidea), 13.3-38.1 (Geometridae), 33.3 (Arctiinae) and 50.0% (Sphingidae), respectively, was noted at other ground based light trapping sites. Hence, in all studied taxa only a minor fraction of the species totals seems to represent a ‘true’ canopy fauna. As a consequence, a representative assessment of moth diversity can be achieved from ground level, without the need of sampling the canopy extensively.
    While beta diversity analysis for Pyraloidea showed – like in Sphingidae, Arctiinae and Syntominae – a relatively distinct canopy species assemblage, which in turn closely resembled species assemblages of open areas, canopy Geometridae species assemblages shared most species with understorey of closed forest nearby. In no case an increase of similarity of canopy Geometridae with geometrid species assemblages sampled at ground in more disturbed habitats could be observed. This may indicate differences in habitat requirements as well as dispersal strategies among the studied moth taxa.

  2. Uniformly, moth taxa showed an unexpectedly high alpha diversity in secondary forests and extensively cultivated sites near to closed forest areas. A dramatic decrease of moth diversity only occurred at cultivated sites far away from forest remnants. This indicates that the distance to old-grown forests may play a major role for recolonization by dispersing moths, and emphasizes the contribution of structurally heterogeneous cultivated landscapes or secondary forests in the vicinity of natural forests – acting as sources for recolonization – to preserve a substantial proportion of tropical moth diversity.

  3. The only positive relationship between the diversity of a moth taxon and one of the three recorded vegetation parameters (diversity of understorey vegetation, diversity of larger woody plants, relative coefficient of variation of tree stem diameters in breast height) was found between diversity of Geometridae moths and understorey vegetation diversity. Taking into account the high dominance of Dipterocarpaceae at canopy level, which do not contribute very much as Lepidoptera food plants, a close relationship between moth and tree diversity could not be expected.

  4. Comparing the elevational distributions of the three studied taxa Sphingidae, Arctiinae and Geometridae, two different patterns of diversity emerged, showing diversity peaks either in lower montane forest (Sphingidae and Arctiinae) or in dipterocarp hill forest (Geometridae). No moth taxon reached maximum diversity in the lowlands. The least pronounced elevational zonation of the three studied taxa was found for the highly mobile hawkmoths. In the less diverse Arctiinae a weak segregation between lower sites and montane forest was noted. The most conspicuous zonation could be recorded for Geometridae. They segregate into four distinct species sets in the lowlands, the hill forest, montane forest, and an additional set of species at the upper margin of the montane forest.

  5. The proportion of endemics varied enormously among taxa. While endemic Geometridae contribute less than 10% to species totals as well as individuals of the recorded Geometridae at individual light trapping sites, endemic Syntominae reached a proportion of up to 100%. However, in none of the studied moth taxa the proportion or abundance of endemics showed any correlation with the extent of anthropogenic habitat disturbance. Specifically, endemics were not more prominent in primary forest as compared to secondary forests or extensively cultivated landscapes.

  6. Although species numbers recorded at individual light trapping sites always correlated between different taxa, only relationships between sub-taxa (e.g. subfamilies) and their superordinated taxonomic levels (e.g. families) were close enough to allow for a reliable estimation of species richness of higher taxa. The available data showed a particularly strong linear relationship between the subfamily Geometrinae and diversity of the total Geometridae fauna. Hence, Geometrinae (low diversity, relatively easy determination) may well serve as diversity indicators for the entire family Geomatridae, which comprise a major fraction of Bornean moth diversity.

Finally, the results of the present study are discussed with respect to their significance for the implementation of practicable conservation, management, and monitoring strategies for highly diverse communities of tropical insects.