A light trap design for stratum-specific sampling of reef fish larvae

Abstract

The vertical distribution of late stage reef fish larvae may potentially influence their dispersal, recruitment success and energetic expenditure during the recruitment process. To date, methods of examining the vertical distribution of reef fish larvae either under-sample late stage individuals, or are incapable of discretely sampling the water column. The aim of this study was to develop a light trap able to sample a narrow depth range enabling fine scale patterns of vertical distribution to be examined. The experimental traps radiated light in a relatively narrow beam with a maximum vertical angle of radiation of 7.5°, indicating that the traps could be placed 4.8 m apart and still sample discrete depth strata. Their catch efficiency was similar to conventional light traps, indicating that they are adequate sampling units. Preliminary data showed that most families are more abundant near the surface, although many have significant numbers of individuals lower in the water column. Some families (inc. Apogonidae) occurred in higher abundance at greater depths. Our experimental light traps permit increased resolution of the vertical distribution of late stage larval reef fishes in the field.

Introduction

It is now widely accepted that behaviour by reef fish larvae during their pelagic phase can potentially have a significant impact on the overall dispersal patterns and recruitment success of reef fishes. There are three main aspects of larval behaviour that may influence their ability to locate and return to a reef. These are: (i) the orientation cues of the larvae, (ii) the potential for larvae to influence their horizontal distribution via active swimming and (iii) their vertical distribution in the water column. A considerable amount of attention has been paid to horizontal swimming Stobutzki and Bellwood, 1994, Stobutzki and Bellwood, 1997, Leis and Carson-Ewart, 1997, Leis and Carson-Ewart, 1998, Fisher et al., 2000, Fisher and Bellwood, 2001, however, the vertical distribution and migration behaviour of reef fish larvae has received less attention (Leis, 1991).

Modelling studies have shown that if larvae can exploit the vertical structure of water currents at different depths, then the extent to which they can influence their dispersal is greatly increased (Armsworth, 2001). Furthermore, the energetic expenditure required to find and settle on a reef may be greatly reduced (Armsworth, 2001). Recent work on the development of the visual abilities of larval reef fishes indicates that late stage larvae may be capable of feeding at considerable depths in the water column, possibly up to 250 m in depth (Job and Bellwood, 2000). The presence of well-developed swim bladders (Leis and Rennis, 1983) and advanced swimming abilities Stobutzki and Bellwood, 1994, Leis and Carson-Ewart, 1997 suggests that larval reef fishes should be capable of fairly rapid vertical movements and precise vertical maneuvering. Indeed, there is evidence to suggest that larval reef fish may utilize vertical movement as part of a mechanism to influence their position during settlement (Cowen et al., 2000).

Despite the apparent ability of larval reef fish to modify their vertical position and the potential importance of this behaviour on their dispersal patterns, the distribution of larval fish in the field has only been briefly examined (but see Leis, 1986, Leis, 1991, Doherty and Carleton, 1997, Hendriks et al., 2001). There are two methods that have been used to describe the vertical distribution of reef fish larvae in the field: towed nets and light traps. Towed nets have been used to describe the vertical distribution of larval fishes around Lizard Island (see Leis, 1986, Leis, 1991). These studies found that reef fishes show taxon-specific patterns of vertical distribution that changed little during ontogeny. However, the problem with towed net methods is that they tend to under-represent late-stage larvae Doherty, 1987, Choat et al., 1993. It is these late-stage larvae, or “competent” larvae that are most likely to be actively returning to reefs. The ability to accurately describe the vertical distribution of these late-stage larvae will lead to a better understanding of the behavioural mechanisms that different reef fish taxa adopt during the settlement process and the ability of reef fish larvae to influence their recruitment success.

Light traps have been used extensively for sampling reef fish larvae in their last few days before settlement, and they have been used to a limited extent to investigate the vertical distribution of reef fish larvae in the field (e.g. Doherty and Carleton, 1997, Hendriks et al., 2001). These studies have shown that while most taxa appear to be most abundant at the surface, there are clear differences between taxa, with some taxa preferring deeper traps. However, this sampling is confounded by the fact that conventional light traps (see Doherty, 1987, Stobutzki and Bellwood, 1997) sample a range of depth strata simultaneously because they illuminate in all directions. In relatively clear water, such traps may sample nearly the entire water column, limiting the utility of depth comparisons. Even if such traps are placed simultaneously at different depths, an accurate description of the vertical distribution of larvae will only occur under the assumption that larvae swim towards the brightest (closest) trap. Considering that there is an evidence that larval fish may change their photo-tactic response as the orientation of the light source is altered (above or below) (Olla and Davis, 1990), this assumption may not be reasonable.

Aside from potential problems due to the differing larval behaviour, the broad vertical sampling of conventional traps also precludes their use in finer scale examinations of vertical distribution patterns. Reef fish larvae are known to show taxon specific differences within the water column as well as between surface and bottom waters Leis, 1986, Leis, 1991. Given that hydrodynamic regimes are also likely to differ on such a scale, it is important to examine the vertical distribution of larvae throughout the water column, rather than just at the surface and near the bottom. While many taxa appear to be caught in highest abundances in surface traps (e.g. Doherty and Carleton, 1997) whether or not these larvae are distributed near the surface or actually occur in mid-water remains to be determined. In order to answer such questions a method of sampling is required that is able to sample a relatively narrow, discrete depth stratum, so that the sampling units can be placed at multiple levels throughout the water column.

The aim of this study, therefore, was to develop a light trap able to sample discrete depths in the water column with a sufficiently narrow sampling band to allow fine-scale patterns of vertical distributions to be examined. The light trap design is described and tested, and data are presented that demonstrate the ability of traps to sample discrete depth strata.