M.Sc. / Please refer to full text to view abstract

The River continuum concept (RCC) (Vannote et al. 1980) attempted to provide a unifying theory of river function. The Functional Feeding Group (FFG) concept (Cummins 1973, 1974) became a major component of the RCC. The FFG concept provides testable hypotheses about the changes in proportions of FFGs along a downstream gradient in a river, in response to the changing nature of food resources. The following short-comings of the FFG concept have been identified: 1) the variability of macroinvertebrate feeding, 2) problems with gut analysis as a method for assigning taxa to FFGs, and 3) inconsistent criteria defining FFGs. The objective of this study was to investigate the macroinvertebrate fauna of the Buffalo River in order to assess the applicability of aspects of the RCC and the FFG concept. The specific aims of the study were: 1) to describe the distribution of macroinvertebrate assemblages in the Buffalo River; 2) to clarify aspects of the FFG concept listed above; 3) to establish whether selected taxa could be assigned to FFGs; 4) to assess whether the proportions of different FFGs in successive reaches of a southern African river conformed to the predictions of the RCC; and 5) to test whether a functional classification is a useful alternative to a taxonomic classification. Macroinvertebrates were collected seasonally from a variety of biotopes at three sites, one each in the upper, middle and lower reaches. Riffles were sampled in summer at 16 sites. Over 100 taxa were identified and an hierarchical classification was prepared using two-way indicator species analysis. Invertebrate assemblages in the narrow headwater stream were taxonomically distinct from those of the middle/lower reaches and were not positively associated with subjectively identified biotopes. Biotopes were characterised by distinct assemblages in the wider middle/lower reaches. Sixteen abundant taxa whose feeding had not been previously investigated were selected for feeding studies, four from the headwaters and 12 from the middle/lower reaches. Methods used included gut content analysis, behavioural observations, food choice experiments and morphological studies using scanning electron microscopy. Three aspects of the functional feeding group concept were clarified. 1) Dietary variability was assessed using gut contents as an index of diet. The gut contents of both early (small) and late (large) instar larvae of all 16 taxa collected from different sites and biotopes, and in different seasons were compared using a multifactor analysis of variance. For all taxa the most consistently significant differences in gut contents were between large and small larvae. These were due to differences in the amount of material in the gut and in varying amounts of rarer dietary items. Dietary variability did not prevent taxa from being assigned to FFGs. 2) Gut content analysis satisfactorily provided basic information about the feeding biology of taxa but proved to be an inadequate single method for positively assigning taxa to FFGs. 3) Before taxa could be assigned to FFGs the definitions for some FFG categories had to be described clearly. It is suggested that the term shredder be based on the observation of shredding and a predominance of leaf fragments in the foregut. The presence of algae was not diagnostic of scrapers and a morphological basis is suggested. A morphological basis for the brusher FFG is described for the first time. All 16 taxa were assigned to FFGs. Three headwater taxa were shredders ((Goerodes caffrariae (Lepidostomatidae), Dyschimus ensifer (Pisulidae), Afronemoura spp. (Notonemouridae)) and one was a collector:brusher (Adenophlebia auriculata (Leptophlebiidae). These results were consistent with RCC predictions. All 12 of the taxa from the middle/lower reaches were filterers or collectors and this result was also consistent with RCC predictions. The Hydropsychidae, Cheumtopsyche afra and Macrostemum capense, were passive net filterers; Neurocaenis reticulatus (Tricorythidae) was a passive setal filterer; Caenidae sp. Band Pseudocloeon maculosum (Baetidae) were active filterers; caenidae sp. A, and the Baetidae, Baetis harrisoni, Centroptilum excisum and Cloeon africanum, were collector: gatherers; the Leptophlebiidae, Choroterpes elegans and Choroterpes nigrescens, were collector:brushers; and Afronurus harrisoni (Heptageniidae) was a scraper. Gut content analyses alone were insufficient to assign taxa to FFGs, but when augmented by morphological and/or behavioural data, taxa could be assigned to FFGs with confidence. In all cases the FFG designation referred to the most frequent style of feeding. N. reticulatus, A. harrisoni and A. auriculata were particularly flexible in their feeding behaviour. A functional classification of macro invertebrates in the Buffalo River was compared with a taxonomic classification. In both cases similar groups were identified, but their taxonomic and functional descriptions yielded different information. It is suggested that functional and taxonomic classifications should be viewed as complementary rather than alternative options.

This thesis presents the first intensive community-level investigation of the mesozooplankton from a biogeographical transition zone estuary. The Mgazana Estuary is located along a rural, semi-undeveloped part of South Africa’s east coast, believed to represent a transition zone between the subtropical and warm-temperate biogeographical zones. The research represented in this thesis involves data collected over a five year period between 2002 and 2006, with additional data collection made in 2008. The study focussed on five areas of investigation, the first of which investigated the physico-chemical dynamics of the Mgazana Estuary. The Mgazana Estuary exhibited marked vertical, horizontal and regional structures in the hydrological environment. Prominent vertical and horizontal stratification characterised summer months. A substantial turbidity front was observed in the lower estuary during summer and winter. The upper estuary was marked by considerable variation in multiple variables but especially salinity. These structures appeared to have shifted in position over the horizontal plane, which was attributed to variation in freshwater flow. The second study focused on the spatial dynamics of mesozooplankton in the Mgazana Estuary. The zooplankton was rich and in terms of composition typical of mangrove systems. The Copepoda were dominant, numerically and in terms of taxonomic representation. Calanoids Acartiella natalensis and Pseudodiaptomus hessei characterised middle and upper reaches in summer and mostly upper reaches in winter. On community level, a highly structured assemblage arrangement was observed during summer and winter months. These trends were further scrutinised under the theoretical framework of ecological boundaries. In so doing, an agreeable spatial association emerged between specific assemblages and their environments. These trends were concluded to reflect ecoclinal as well as ecotonal properties, the latter describing interactions over narrow spatial bands of marked changes in turbidity. The third study investigated flooding events as short-lived extreme meteorological events and the influence on zooplankton. Within this regard, second-stage multivariate statistics was used to assess year-to-year variability in assemblage structures on whole-system and regional scales. The impact of two major flooding events that flushed the estuary some days prior to sampling could clearly be elucidated. Flooding emerged as a significant source of inter-annual variability in the zooplankton of the Mgazana Estuary. Second-stage multivariate analysis proved to be an effective analytical strategy for investigating inter-annual variability in species assemblage structures. Results from the preceding study prompted a detailed investigation into the spatio-temporal dynamics of Acartiella natalensis, the most important zooplankton species of the system aimed at elucidating flood responses. Acartiella natalensis showed indication of temporal arrest in association with flooding events in a similar fashion as is observed in seasonal variation trends in estuaries towards its geographical distribution limit south of the study area. It was concluded that A. natalensis was severely affected by floods and the possibility exist for entrainment of entire populations from the estuary during flooding events. The final study tested the hypothesis of post-flood propagation from a resting egg bank in the sediments of the Mgazana Estuary. Preliminary data revealed that numerous eggs were present in the sediments. Nauplii hatched from eggs under laboratory conditions were identified as belonging to the family Acartiidae. The preliminary mode of diapause is presented as a schematic model, emphasising the role of freshwater flow and specific environmental variables. This was the first discovery of resting eggs from a subtropical estuary from South Africa. It is hoped that the findings of this study would give rise to new research initiatives investigating the importance of resting stages in estuarine and coastal Copepoda species and the role such reproductive strategies may play in estuarine functioning.

Macrozoobenthic community structure and composition was investigated along a subtidal-intertidal gradient in the Mngazana Estuary. Six transects were sampled between the spring high water mark (HWST) and the bottom of the river channel in the lower estuary. Fifteen replicate samples were collected along each transect using a Van Veen type grab (211 cm2 bite) during each of three sampling sessions. Samples were sieved through a 500 μm mesh bag and the invertebrates stored in bottles for further analysis in the laboratory. Additional grab samples were collected for sediment particle size analysis and organic matter. Physical variables measured at each transect included: salinity, temperature, dissolved oxygen, depth, pH, percentage mud, organic content and turbidity. Sediment compactness was measured at all intertidal transects and additional sediment samples were collected at mid shore and high shore transects for percentage water content analysis. A total of 104 species were recorded along the intertidal-subtidal gradient in the sampling area. Species richness was higher in the subtidal zone compared to the intertidal zone and polychaetes numerically dominated the macrozoobenthic community at most transects, during all three sessions. At high shore transects the community was characterised by having fewer species, consisting mostly of brachyurans, polychaetes and gastropods. Shannon diversity index (H’) was generally higher for subtidal transects (x¯ = 2.3; range: 2.8 to 1) than for intertidal transects (x¯ = 1.4; range: 2.2 to 0.6) indicating that the distribution of individuals among species in the intertidal zone experienced greater variability. Results for Hill’s numbers followed the same trend as Shannon diversity with subtidal communities mostly consisting of abundant species followed by very abundant species. Intertidal communities generally exhibited lower numbers of abundant and very abundant species. Sedimentary characteristics played a major role in structuring benthic communities in comparison to other physico-chemical variables. Organic content and mud content of the substrate were identified as important factors influencing community patterns observed along the subtidal-intertidal gradient. In addition, sediment compactness and water content of the substrate was found to influence intertidal community structure. Subtidal community structure possibly had a greater dependence on seasonal variations in abiotic and/or biotic factors. Cluster dendrograms used in conjunction with MDS ordination mapping revealed that macrozoobenthic communities were generally distinct between high shore intertidal transects iii and subtidal transects. Most species exhibited a broad spatial distribution along the subtidal-intertidal gradient with mid and high shore transects being the exception. Most species also exhibited marked shifts in abundance and this was especially noticeable at the transition between the subtidal and intertidal zone. Two polychaete species, Prionospio sexoculata and Capitella capitata, were very abundant species and featured amongst the most numerically dominant species collected during each sampling session.

Numerous ecological studies have used lipids to determine trophic pathways in aquatic systems, as fatty acid profiles provide time-integrated information on an organism’s assimilated diet. Many of these studies have, however, been based on sample collections with a limited temporal scale. The trophic ecology of pelagic systems has been studied intensively using fatty acid analyses, but very little work has been directed toward benthic communities, with the intertidal being especially neglected. The investigation of trophic pathways within rocky shore communities will help us to better understand system responses to environmental changes. The determination of long term temporal variation of the food web within a community could reveal the type, magnitude, duration and frequency of highly seasonal productivity. Changes in fatty acid profiles through time in primary consumers of intertidal rocky shores are poorly understood, but represent an important step towards a more comprehensive understanding of rocky shore food webs, compared with those derived from snapshot or short-term studies. The aim of this thesis was to clarify the temporal variability in the diets of rocky shore intertidal suspension-feeders (the brown mussels Perna perna and the Cape reef worm, Gunnarea gaimardi) and grazers (the Cape sea urchin Parencinus angulosus and the Goat-eye limpet, Cymbulus oculus) on the south east coast of South Africa using fatty acid profiles, and to investigate the effects of life style (e.g. feeding mode) and life cycle on temporal variations in tissue fatty acid profiles. I had three hypotheses: firstly, that suspension-feeders experience high levels of variability in their diets through time because water quality has the potential to change quickly and drastically, whereas grazers experience less variability in their diets over time since their food sources are more constant. Secondly, the reproductive cycles of the suspension-feeder P. perna and the grazer P. angulosus affect the fatty acid composition of their gonads, with temporal variations in lipid composition reflecting changes in reproduction investment. Thirdly, the total amount of energetic reserves available for reproduction are different for each gender (females allocate more energy to egg production than males allocate to gamete production). To address these aims, fatty acid profiles of suspension-feeders and grazers were investigated over a period of twelve months (from July 2010 to June 2011) at a single site on the south east coast of South Africa. The results showed high variability in the fatty acid composition of both the suspension-feeders strongly related with changes in their food source (suspended particulate material). Furthermore, similar temporal changes in fatty acid profiles of the two suspension-feeders were observed over time, reflecting their common diet and life style. There were some inter-specific differences in the suspension-feeders, likely originating from differences in their particle capturing mechanisms. Grazers showed less variability through time compared with the suspension-feeders, with the limpets being more consistent than the sea urchins. The temporal variability in the sea urchin diets may have resulted from the highly diverse and heterogeneous food sources available to them, whereas limpets may be more selective and have a limited range of diet items. Differences between the two grazer species may have arose from differences in their feeding strategies and intertidal zonation. The fatty acid compositions of gonad tissues in both P. perna and P. angulosus showed temporal variability strongly related to reproductive cycle. Differences in the fatty acid values between females and males were apparent, with females richer in total and polyunsaturated fatty acids than males. Spawning and gametogenesis influenced the variability of fatty acids through time in both species, suggesting the importance of considering the reproductive cycle when studying lipids in rocky shore species. Little evidence of lipid transfer between muscles and gonads was seen, suggesting the importance of direct lipid storage into the reproductive tissues. The influence of diet and life history of intertidal consumers on the temporal variability of their fatty acid compositions is important to understand, as it provides us with a better understanding of the functioning of rocky shore systems. There is an enormous potential for future research in this field of study.