Oecologia (2002) 131:89–93 DOI 10.1007/s00442-001-0866-4
P O P U L AT I O N E C O L O G Y
Laurent Vigliola · Mark G. Meekan
Size at hatching and planktonic growth determine post-settlement survivorship of a coral reef fish
Received: 11 June 2001 / Accepted: 17 December 2001 / Published online: 31 January 2002 © Springer-Verlag 2002
Abstract Coral reef fishes, like many marine organisms, have a complex life history that consists of a planktonic larvae stage and a benthic juvenile or adult stage. We used the growth records in the otoliths of a common damselfish to investigate the extent to which processes in the plankton determined the outcome of events after benthic settlement. Sequential samples of the same cohort showed that individuals that survived intense selective mortality 1–3 months after settlement were those fish that were the larger members of the cohort at hatching and grew faster during planktonic life. Such links between life history phases are likely to occur in reef fishes whenever there is selection for a trait that is cumulative, such as size. They may not only operate between life history stages in the same individuals, but even between those of different generations via maternal effects on size at hatching. Keywords Size selection · Maternal effects · Coral reef fish · Post-settlement mortality · Otoliths
Introduction In the sea, the life cycles of a great range of both vertebrate and invertebrate species are divided into a larval phase that develops in the plankton and an adult phase that is associated with the benthos. Such life cycles are typical of >90% of marine fishes and are particularly widespread in those species that inhabit coral reefs (Leis 1991). While the duration of pelagic life varies among species, so that in some it lasts little more than a week L. Vigliola (✉) · M.G. Meekan Australian Institute of Marine Science, Townsville MC Queensland, Australia 4810 e-mail:
[email protected] Tel.: +1-352-3922449, Fax: +1-352-3923704 Present address: L. Vigliola. University of Florida, Department of Zoology, 223 Bartram Hall, P.O. Box 118525, Gainesville, FL 32611-8525, USA
and in others many months, a significant part of the development of most coral reef fishes (either in terms of age or size) occurs in the plankton (Leis 1991). At present, we know far more about the adult and juvenile phase of coral reef fishes than we do about the larval phase that precedes it. This is not surprising, given that young larvae can be difficult to identify and inhabit a dynamic, heterogeneous environment that is notoriously complex to sample. These problems have led most reef ecologists to treat the larval stage as a “black box”, where inputs (newly hatched larvae) and outputs (settlement) have been quantified, but the identity and importance of processes occurring within the plankton, and their effects on subsequent life history phases, remain largely unknown (Leis 1991; Cowen and Sponaugle 1997). Recent evidence from temperate environments suggests that it is unlikely that the alternate phases of reef fish life histories can be regarded as independent phenomena. Theoretical, experimental and field studies show that the high rate of mortality (nearly 100%) undergone by larval and juvenile fishes is size selective, so that small fish have lower survivorship than larger fish of the same age [the growth-predation theory (Anderson 1988; Miller et al. 1988; Bailey and Houde 1989; Cushing 1990)]. If such processes also operate in tropical environments, there are likely to be strong links between the planktonic and benthic phases of coral reef fishes, since size at age is a cumulative variable and represents the sum of growth rates at all times prior to that age. The growth histories recorded in otoliths of reef fishes provide a means to open and illuminate the interior of the larval “black box” to test this prediction. By repeatedly sampling the same cohort of fish and analyzing otoliths it is possible to back-calculate growth rates throughout early life and to examine the influence of planktonic growth on survivorship during benthic stages (Hovencamp 1992; Meekan and Fortier 1996). We used this approach with a common damselfish, Neopomacentrus filamentosus (Macleay), and show that the post-set-
90
tlement survivorship of this species is strongly dependent on initial size at hatching and growth rate during the planktonic phase.
Materials and methods N. filamentosus is widely distributed in the Indo-Australian Archipelago (Allen 1991). Settlement of this species occurs on new moons during summer months (L. Vigliola and M. G. Meekan, unpublished data). Typically, settlers form schools at the bottom of large coral colonies (2–4 m diameter) where they remain closely associated with the substrate for several weeks. Schools of adults and older juveniles occupy the top of these same coral colonies and feed in the water column between the coral and the surface (L. Vigliola and M. G. Meekan, personal observation). Fish at the end of the pelagic stage were collected at a site in the Dampier Archipelago, Western Australia (20°30′S, 116°40′E), using light traps [Doherty (1987); see Fig. 1 in Meekan et al. (2001) for trap design]. This technique captures young fish in the few days just prior to their settlement on reefs and catches in traps are strongly correlated with the magnitude of larval supply to benthic populations (Milicich et al. 1992; Meekan et al. 1993). Two traps were deployed at the surface on moorings and cleared of catches every morning for a period of 2 weeks around the time of the new moon in November 1998. Divers using hand-nets and an anaesthetic sampled individuals of this same lunar cohort 1, 2 and 3 months after settlement from a large (500 m) fringing reef on an isolated island 50–100 m from the site where light traps were deployed. Sufficient numbers of fish were usually obtained from five to seven randomly chosen large coral colonies during any single collection. Coral colonies were not re-sampled at later dates and effort was spread along the reef and over the depth range (4–10 m) occupied by the species to account for potential spatial variation in traits of the cohort. The area from which fish were removed and the total numbers collected represented very small proportions (a fraction of 1%) of the total habitat and abundance of fish in the study site. Fish from each of the collections were measured to the nearest 0.01 mm standard length (SL) and sub-sampled for otolith analysis. A total of 220 settlers were collected using light traps in November, of which 100 were selected for analysis in proportion to the abundance and sizes of fish collected in each trap. Most of these fish (77%) were caught on only three nights of sampling (15–17 November). Samples of benthic juveniles collected by divers were composed of 408 fish in December, 637 in January and 750 in February. These collections of post-settlement fish included many individuals that were not part of the November cohort. In order to restrict our analysis to fish settling in November, we sub-sampled collections by generating a preliminary growth curve. This was done by aging 36 fish that ranged in size from 9 to 45 mm SL in millimeter increments. These fish were haphazardly selected from all collections. The growth curve was then used to predict the range in size of fish 1 month after settlement, given the range in size at age of fish collected from the light traps. One hundred fish were then sub-sampled from the individuals collected 1 month after settlement in proportion to the numbers occurring in 1-mm size classes of SL within the predicted size range. Once these fish were aged, the process was repeated to subsample fish collected 2 and 3 months after settlement. Sub-samples were generated from the preliminary growth trajectory estimated from the growth series of 36 fish, the variability in size at age of samples of fish collected at previous ages and the numbers occurring in 1-mm size classes of SL within each predicted size range. As predicted size ranges were very broad, so that both slow- and fast-growing individuals were included in the subsample, aging revealed some fish that did not settle in November. These were removed prior to analysis, so that our sub-samples were composed of 98 fish collected in December, 85 in January and 71 in February.
Lapillae were removed from each of the 354 fish selected for analysis, mounted on a glass slide, ground on lapping film to produce a thin transverse section that contained the nucleus and viewed using a compound microscope under transmitted light with immersion oil at 1,000× magnification. Increments within the otolith were measured to the nearest 1 µm along the longest axis of the section using an image analysis system. Each otolith was read once by the senior author. The daily periodicity of deposition of increments within the otoliths of N. filamentosus has been validated experimentally in a mark-recapture study that used a chemical tag (alizarin) that was incorporated into otoliths (A. Hansen and A. Retzel, unpublished data). We assumed that the increment closest to the core of the otoliths of this species was formed at the day of hatching, as is the case in many other tropical and temperate reef fishes (Campana and Neilson 1985; Wellington and Victor 1989). Back-calculation of size from otoliths assumes that there is proportionality between otolith and somatic growth rates (Vigliola et al. 2000). The assumption was verified by calculating a regression relationship between otolith radius and SL. This analysis was highly significant and there was a strong (r2=0.95, P
