| |
| |
Custom Fields
| Title: | Structural dynamics of rainforest collembolan Assemblages in leaf litter suspended epiphytes and on the forest floor | | Year: | 1999 | | Author/s: | Denis J. Rodgers | | Degree: | Doctorate | | University/Institution: | Griffith University | | Abstract: | "This thesis is focussed on the study of the structural relationship between collembolan assemblages occurring in leaf litter on the rainforest floor and assemblages occurring in leaf litter suspended in epiphytic plants in the rainforest canopy. The rationale for the selection of this subject is detailed in the introduction (Chapter) and relies on the necessity for advancement in the scientific understanding of the ecology of rainforest arthropods if we are to conserve biodiversity. An initial hypothesis is examined suggesting that rainforest leaf litter collembolan assemblages might be vertically stratified. This hypothesis is developed from a review of the literature concerning the biology and ecology of the Collembola and consideration of relevant biotic and abiotic characteristics of rainforest environments.
Data collected in a pilot study designed to test this hypothesis are analysed using a number of recently developed methods in multivariate statistics.
A framework for extending the study of this system is then established through the identification of three questions concerning the generality of these patterns and a further three questions concerning factors generating the patterns observed. The questions are resolved as follows:
1. No significant spatial variation is found to occur in patterns of structural relationship between forest floor and canopy leaf litter collembolan assemblages in the tropics and subtropics. Temporal variation in these patterns is far stronger and winter / summer differences occur between patterns of assemblage relationship on all sites sampled.
2. The consistency of patterns of assemblage relationships among the Collembola with those among other associated arthropod assemblages is assessed using a series of Mantel's tests applied to a number of assemblage data sets drawn from the same set of samples. These analyses show that patterns of relationship between collembolan assemblages are most similar to those among the Oribatida and Mesostigmata, and least similar to those among the Coleoptera and Isopoda. A further set of analyses shows that when these data sets are sequentially pooled, the correlation increases between patterns within any single data set and any pooled data set. This suggests that ecological patterns of relationship within a data set for any arthropod Order are more informative of ecological patterns among the Arthropoda in general. than they are of ecological pattern among any other single arthropod Order. This is regarded as a highly significant result (in scientific rather than statistical terms) and an important area for further research. In answer to the question originally posed, these results indicate that patterns of relationship among leaf litter collembolan assemblages are more highly correlated to patterns of relationship within the arthropod assemblage as a whole, than they are to patterns among assemblages of any other single arthropod Order.
3. The collembolan assemblage associated with leaf litter is shown to be composed of a set of species distinct from those found in soils, on tree trunk surfaces and in the canopy in general. This result suggests that the results of all other analyses can only be interpreted as having relevance to the leaf litter microhabitat and not the rainforest in general.
4. Field observations of movements of collembolan species covering a wide range of body forms and sizes suggest that habitat selection, rather than dispersal capability, governs the species composition of forest floor and suspended litter collembolan assemblages.
5. A senes of experimental studies show that following heavy falls of rain, leaf litter in all vertical strata of the forest is uniformly saturated. However, as the leaf litter dries during fine weather, a pattern of vertical stratification in the distribution of leaf litter moisture develops. After five days of fine weather, leaf litter in the upper canopy is shown to be significantly drier than that in the lower canopy, and that in the lower canopy is in turn shown to be significantly drier than that on the forest floor. The size of suspended litter deposits is also shown to play a significant role in determining drying rates with smaller deposits drying out more quickly than large deposits.
6. The species of host tree in which epiphyte associated litter deposits are suspended has no significant effect on the species composition of collembolan assemblages in the suspended litter.
A final component of the field studies associated with this project is the establishment of an experimental system to facilitate further studies of this system.
The results all of these studies are drawn together in the development of a model of the structural dynamics of these assemblages. The model consists of four components. The first of these and the core of the model, are four phases in which the state of relationship between the assemblages may occur. The second category of components are the process links by which transition from one state to another occurs (principally vertical migration of collembolan species). Third are the input stimuli (different rainfall levels) which initiate the process components and drive the system as a whole. The fourth component is the sensitivity of the model to a site-specific degree of canopy closure.
In conclusion, the model represents the extent to which an understanding of the ecology of this system has been achieved. The model does require some refinement but the experimental system established provides the means for resolving these finer details of the functioning of this system.
" | | Type: | Thesis |
|
| |
Categories 
https://localcoinswap.net