What controls tropical reef fish populations: recruitment or benthic mortality? An example in the Caribbean reef fish Haemulon flavolineatum

Abstract

Recru~tment from a planktonic larval stage has been proposed to b e a n important factor in lim~ting populations of marine organisms, part~cularly troplcal reef fishes. We monitored r e c r u ~t ~n e n t and populat~on densities of juven~le size classes In French grunt Haemulon flavolineatum (Haemulidae) from October 1978 through December 1980 In a portion of Tague Bay, St. Croix, U. S. Vlrgin Islands. Withln our study area, 95 % of new recrults settled onto the sand and seagrass lagoon floor and w i t h ~n a few weeks migrated to nearby reefs; the remalnlng 5' 10 settled d~rectly onto reef structures. hIean annual recruitment rate was 1.8 recrults per m2 of lagoon floor, equivalent to 44 recruits per m2 of luvenlle (backreef) habitat per yr and among t h c h ~g h e s t recruitment rates yet reported for reef fishes. Post-settlement mortality was also very high. d u r ~n g the first month of benthic life it was 0.9 and during the first year was > 0.992. We propose a s e n s ~t ~v i t y analysis criterion for determining the relative ~nfluence of multiple factors or processes that llmit adult population slze The critenon w e used is the relatlve senslt~vity of adult population size to proportionate changes in those factors or processes. Using an open, d e n s ~t y -~n d e p e n d e n t model of benthlc populat~on dynamics for French grunts, we estimate that changes In post-recru~tment mortality rate \v111 have a 100 tlmes greater effect on the abundance of large juveniles than a proportionate change In recruitment rate Therefore w e conclude that, whether or not mortality 1s density-~ndependent, the factors controlling b e n t h ~c mortality are much more Important in the b e n t h ~c populat~on d y n a m ~c s of French grunts than 1s the recruitment rate. The same analysis is applied to other coral reef flshes for which appropriate data are ava~lable; \ye conclude that for 2 out of 3 species for whlch recruitment limitation has been demonstrated, changes in b e n t h ~c mortahty are hkely to have a stronger influence on adult populat~on sizes than are propoi-tlonately equivalent changes in recruitment rate