The determination of fish age and growth is fundamental in fisheries biology and management. Such age-determined parameters as mortality and growth underlie the population dynamics models used in fishery analyses. Age studies can furnish other basic data such as stock age structure, age at first maturity, spawning frequency, individual and stock responses to changes in the habitat, recruitment success, etc. Age and growth data also permit the determination of population changes due to fishing rates.
Age can be determined by one or more of the following methods.
Anatomical method: counting the regular growth marks formed in hard tissues such as scales, otoliths, vertebrae, spines and tail bones.
Length-frequency analysis: monitoring the progression through time of the identifiable modes in size classes.
Direct estimate: through direct measurements of growth rate of specific specimens extrapolated to the stock as a whole. Marking and subsequent recapture of fish, or monitoring the growth of captive fish of known age are two direct estimation methods.
Ageing tropical fishes was until recently assumed to be virtually impossible due to continuous spawning and the absence of growth cycles (Mohr, 1921), making the application of the anatomical method and of length-frequency analyses difficult if not impossible. It has been demonstrated, however, that though tropical fish have a longer spawning period than temperate fish (Lowe-McConnell, 1987), recruitment is limited to one or two seasons of the year. This limitation may arise out of spawning fluctuations or out of the juvenile and larval mortality which governs and limits recruitment to specific periods (Bakun et al., 1982; Victor, 1982; Robertson et al., 1988).
The deposition of annual growth rings (annulae) in the calcified tissues of bony fishes is at least partly caused by seasonal changes in the environment. These periodic changes (temperature cycles, available food) are less regular and less severe in tropical than in temperate zones. Several authors do, however, mention the presence of annual growth rings in tropical fish otoliths (Poinsard and Troadec, 1966; Quasim, 1973; Manooch III, 1987). The causes of this cyclical annual growth are unclear: some authors link them to spawning periods and others to water temperature changes. As annual growth rings are present in immature fish, ring formation probably follows an internal rate of growth synchronized to seasonal environmental variations.
A new application of otolith growth structure analysis was developed by Pannella (1971; 1974; 1980), who showed that the concentric shells (microscopic lamellae) (Hickling, 1931) were formed daily. Daily growth increments offer a very promising field of study with many applications (Campana and Neilson, 1985).
These anatomical methods make it relatively easy to determine age and growth (Bagenal and Tesch, 1978; Casselman, 1983; Beamish and McFarlane, 1987). Nonetheless, annulae and growth studies cannot assume specific periodicity in growth marks, and so this must be determined for each age class of the stock studied (Beamish and McFarlane, 1983).
The method of separation of the modal classes (considered to correspond to distinct age classes) of the length-frequency distribution was the first to be applied to the determination of growth (Petersen, 1891). The early tropical fish growth studies used length-frequency analyses. However, the superimposition of successive modal classes and the difficulty of collecting representative, non-selective population samples are a frequent source of problems in the application of this method (Mathews, 1974; Morgan, 1983, 1985; Morales-Nin, 1989).
The method of estimating growth directly by marking and subsequent capture has rarely been used for tropical species. Diseases due to handling or marking the fish, as well as erroneous measurements, can affect the estimate. Additionally, many specimens must be marked to ensure a sufficient number of recaptures for growth estimates. In practice, marking and recapture experiments are limited to hardy species which can survive the stress of handling out of the water and also be recaptured in sufficient quantities (Munro, 1982).
The right method to apply in each case will depend on the available data and the characteristics of the population under study, plus technical and cost factors (Mathews, 1987; Gulland, 1987). Additional time and experienced, expert staff will be needed for the interpretation of the otoliths (Williams and Bedford, 1974), whereas length analyses are based on rather easily obtained data which can be quickly processed. The evaluation of the relative cost of each method should, however, bear in mind the degree of precision of the findings (Gulland, 1987).
The relative uncertainty inherent in all growth determination methods suggests the use of two independent techniques to confirm the findings. The use of length-frequency analyses and the simultaneous interpretation of growth marks probably offers the best results. The international ICLARM/KSIR meeting on the theory and application of stock assessment methods based on length-frequency analyses, which was held in 1985 (Pauly, 1987), concluded that length-frequency analyses methods are made much more precise by the inclusion of information on growth obtained through the use of an independent method, usually based on otolith reading (Morgan, 1987).
This paper describes a growth determination method based on otolith microstructure analysis. Growth determination by length-frequency analysis has been described in other FAO publications (Sparre et al., 1989) and by other authors (Pauly, 1982; Morgan and Pauly, 1987; Csirke et al., 1987).
