| Summary: | Movement of fishes is an integral part of their daily life, but has significant implications for fishery management. As with nearly all coastal countries, South Africa relies on coastal fisheries as a renewable resource, but many stocks have been overexploited for decades. Although it has long been recognised that an understanding of fish movement is necessary for effective management, it is with some difficulty that the subject has been studied in the past. In recent years, however, improvements in technology have provided the means for more in-depth investigations into fish movement. This research has revealed a range of complex movement behaviours. Movement in fishes occurs on a variety of temporal and spatial scales leading to the characteristic patterns of distribution and abundance observed in marine ecosystems. Fishes move nearly constantly in search of food, shelter or reproductive opportunities. Observations of behaviours such as long-term site fidelity, longdistance migration and natal homing are enabling ecologists to understand patterns of distribution and abundance within a species' range. Fish movement around the South African coast has been studied on numerous occasions but this has largely been confined to studies on single species. Movement behaviour of multiple species has been studied, but this has been limited to spatially localised marine protected area research. There has been little attempt in southern Africa to synthesize interspecific movement behaviour over wide spatial scales.Unprecedented concern over the biological effects of overexploitation, together with rapid technological advances in biotelemetry, have provided the impetus for much research, on a global scale, into the movement of marine animals. I reviewed 101 marine and estuarine fish movement studies from southern Africa, published from 1928 to 2014, with the aim of synthesising research trends and findings. Trends showed an increasing emphasis on fish movement research in publications in the sub-tropical and warm-temperate biogeographic regions along the south and east coasts of southern Africa. Although 63% of publications featured only marine studies, research into fine-scale habitat use in estuaries has been on the increase, concomitant with increasing accessibility of biotelemetry. Overall, 26 fish families were identified in the surveyed literature with regionally endemic sparids featuring in 32% of the publications. Ten movement themes were identified in the surveyed literature, including broad-scale movement patterns, which featured in 68% of studies, followed by fine- scale habitat usage (33%) and protected areas (26%). The most prominent phenomenon, emerging from this research, is that of partial migration, which describes the occurrence of resident and migratory behaviour within a coexisting animal population. Substantial progress has also been made in unravelling the complexities of fine-scale movements in marine reserves and habitat usage in estuaries. While this knowledge has enabled more effective management of South Africa's multi-user, multi-species fisheries, focus should now be directed at improving our understanding of the commonalities in movement behaviour, the associated driving forces behind this behaviour and the extent of movement across reserve boundaries. Mark-recapture data, collected over the past 30 years by the Oceanographic Research Institute’s Cooperative Fish Tagging Project (ORICFTP), were used to investigate broad- scale movement patterns of 30 prominent coastal fishery species (Chapter 4). Fishes were tagged with plastic dart tags along the coastline of southern Africa from Angola to Mozambique. This exercise yielded more than 10000 recaptures. The 30 chosen species represented 14 families, although 12 species belonged to a single family (Sparidae). Overall, 67% of recaptures were reported from the original tagging locality and 73% were recorded within 5 km of the tagging locality. The remaining observations extended from 6-3000 km. Movements were assigned to four distance bins (0-5 km, 6-50km, 51-500km and >500 km) and modelled with an ordinal logistic regression. Species, life-stage (juvenile/adult) and time- at-liberty were included as predictor variables. Model coefficients were then included in a cluster analysis, which produced two primary groupings of species (Category I and II), with two sub-groupings (Category IIa and IIb). Category I species were characterised by wide-ranging movements, greatest median recapture length and highest trophic levels. Category II species were characterised by residency, lower median recapture length and lower trophic levels. These findings have implications for fisheries management. Exploitation of resident species may lead to localised depletion, but their diffuse spatial distribution may offer some resilience. In contrast, even localised targeting of migratory species may pose a population level risk if individuals are known to aggregate. Life-cycle diversity or intra-population variability describes the existence of alternative strategies or tactics among coexisting individuals within an animal population (Chapter 5). Partial migration is a specific case of life-cycle diversity where coexisting groups exhibit either resident or migratory (wide-ranging) behaviour within a single population. Mark- recapture data collected under the auspices of the ORICFTP were used to investigate the occurrence and nature of life-cycle diversity in the movement behaviour of five non- diadromous fish species around the coastline of southern Africa. Among the five species were three teleosts (Category I and IIa) and two elasmobranchs (Category I). A fish was considered to have remained resident if recaptured within 0-5 km after 365 days at liberty. A fish was considered to have undertaken a wide-ranging movement if recaptured more than 50 km away from the release site in 365 days or less. A total of 1848 individuals from the five study species were recaptured during the study, of which 73% of the observations were classified as being resident. Binomial logistic regression confirmed that species, life-stage (juvenile/adult) and ecoregion were significant (p < 0.001) predictors of the probability of wide-ranging behaviour. A Gaussian model confirmed that species and ecoregion were also significant (p < 0.001) predictors of direction and distance of wide-ranging movement. However, the direction and distance of wide-ranging movements in juveniles did not differ significantly (p > 0.05) to that of adults. The median growth rate was mostly greater in wide- ranging individuals; however, this was only statistically significant (p < 0.05) in two cases. These findings provide unequivocal evidence of life-cycle diversity in five fish species, with vastly different life-histories. This ecological phenomenon may provide species resilience at the population level and needs to be considered in fisheries management initiatives. The movement of fishes is a fundamental aspect to consider when designing fishery management regimes. Unfortunately, traditional management strategies have often disregarded movement behaviour to the detriment of fish populations (Chapter 6). As a case study, the management of Lichia amia (Category I: wide-ranging) was evaluated in the context of its movement behaviour. Long-term catch-per-unit-effort (CPUE) datasets were examined for three South African recreational fishery sectors. The CPUE was standardised using generalized linear models (delta-Gamma/hurdle approach) to reduce the effect of factors other than abundance. Factors that were available for this purpose were year, month and locality/zone. Year was included in every model as the primary objective was to detect trends in abundance over time. Although standardised CPUE for all sectors showed an overall long-term decline, there was considerable variability in trends between the different recreational sectors and between datasets. Contrasting trends between competitive shore angling and general shore angling datasets were ascribed to hyperstability in competitive CPUE data. Hyperstability in this case was mostly influenced by rapidly improving technology, techniques and communication networks amongst competitive anglers. Month and locality were significant factors explaining the probability of catching L. amia. This suggests that the predictable aggregatory behaviour of this species could further compound the observed CPUE hyperstability. Although the CPUE responded positively for six years after implementation of the first minimum size and bag limits, and for one year after the decommercialisation of the species, these regulations and their amendments failed to arrest a long-term decline in the CPUE despite the ample evidence for hyperstability. It is clear from this case study that the predictable nature of wide-ranging behaviour in L. amia has made the population vulnerable to exploitation. This has led to the demise in the population, which could have been worse if not for the occurrence of intra-population variability in its movement behaviour, which may provide some measure of resilience.
|
|---|
