Life cycle, early life history, fisheries and recruitment dynamics of diadromous gobies of Dominica, W.I., emphasising Sicydium punctatum Perugia

Bell, Kim Nigel Ian (1994) Life cycle, early life history, fisheries and recruitment dynamics of diadromous gobies of Dominica, W.I., emphasising Sicydium punctatum Perugia. Doctoral (PhD) thesis, Memorial University of Newfoundland.

[English] PDF (Migrated (PDF/A Conversion) from original format: (application/pdf)) - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (31MB) [English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. (Original Version)

Abstract

Fisheries for diadromous gobies occur widely in the coastal inter-tropics. Yields are rarely documented in detail, but there are many reports of and allusions to declines, which have not been satisfactorily explained. Basic life-history information has been at best sketchily known, or not at all, such that various sicydiine gobies have been incorrectly described as catadromous on the basis of assumptions alone. Sicydium punctatum Perugia is shown to be diadromous, spawning in rivers and spending 50 to 150 days at sea before migrating to fresh waters. -- Larval behaviour is described and experiments show that larvae have the ability to select particular salinity layers in stratified systems. Implications for early life history transport, survival and vulnerability to terrigenous toxins are discussed. -- Larval fish occurring in the rheoplankton are shown to be separable into five types, using pigment and other characteristics. The five types numerically correspond to the number of goby species known in Dominica, and one type is verified as S. punctatum through several captive spawning and collected nests. Separation into types permitted an analysis of mortality in rivers, using stream drift data in a manner not previously applied. The theory of this manner of estimation is discussed, and field results for S. punctatum are compared with two types of analysis of mortality in captivo. Field data are also considered for several other taxa to demonstrate the method. The mortality rates found for S. punctatum are unprecedentedly high, but the agreement among multiple samples and with the captive observations suggests that stream mortalities may be extreme, and that larvae nearest the coast have a significant advantage. Coastal habitat is therefore the most important in sustaining the fishery. Implications and questions arise for the reproductive ecology, competition and upstream migration of Sicydium spp. -- While age-at-recruitment (AAR) has thus far been treated as a constant (each species) plus error, the duration of the postlarval period of S. punctatum is shown on the basis of otolith analyses to vary systematically with time of year. There is contrary variation in size-at-recruitment, indicating strong seasonal variation in growth rate. The variation in age-at-recruitment suggests population dynamics not previously acknowledged in either fact or theory, and these are discussed in principle and the dynamics modeled by numerical simulation. The unavoidable conclusion is that, even where reproduction and daily risk of mortality remain constant in all seasons, observed levels of variation in AAR are sufficient to induce large variations in yield. The characteristics of the simulated yield patterns closely match the actual yield data. The amplitude of variation generated depends on level of (constant in each simulation) mortality. The same principle applies not only where changes are seasonal, but to any temporal trends, and may have implications in other systems. Further variation in reproduction or mortality could increase or decrease these amplitudes, but since the variation in AAR generates variations over 10- to 30-fold at plausible field mortality rates, large variations in any other factor would be required to mask these effects.

Actions (login required)

View Item