Episodic fluctuations in larval supply

Science

Volumn 283, Issue 5407, 1999, Pages 1528-1530

  Dixon, Paul A ^a   Milicich, Maria J ^b,c   Sugihara, George ^a  

^a UNIVERSITY OF CALIFORNIA   (United States)

^b GRIFFITH UNIVERSITY   (Australia)

^c M2 Environmental Ltd   (Hong Kong)

Author keywords

[No Author keywords available]

Indexed keywords

FISH; REPRODUCTIVE BEHAVIOR; SPAWNING; TEMPORAL VARIATION;

ARTICLE; AUSTRALIA; BREEDING; FEEDBACK SYSTEM; FISH; ILLUMINATION; LARVA; MARINE ENVIRONMENT; NOISE; NONHUMAN; NONLINEAR SYSTEM; PRIORITY JOURNAL; SEASONAL VARIATION; STOCHASTIC MODEL;

AMPHIBIA; REPTILIA;

EID: 0033525605     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.283.5407.1528     Document Type: Article

References (27)

1. B. J. Rothschild, Dynamics of Marine Fish Populations (Harvard Univ. Press, Cambridge, MA, 1986).
2. R. Hilborn and C. Walters, Quantitative Fisheries Stock Assessment: Choice, Dynamics, and Uncertainty (Chapman & Hall, London, 1992); N. Caputi, Can. J. Fish. Aquat. Sci. 45, 178 (1988).
3. R. Hilborn and C. Walters, Quantitative Fisheries Stock Assessment: Choice, Dynamics, and Uncertainty (Chapman & Hall, London, 1992); N. Caputi, Can. J. Fish. Aquat. Sci. 45, 178 (1988).
4. D. Cushing, Climate and Fisheries (Academic Press, London, 1982); R. Lasker, Marine Fish Larvae (Univ. of Washington Press, Seattle, WA, 1981).
5. D. Cushing, Climate and Fisheries (Academic Press, London, 1982); R. Lasker, Marine Fish Larvae (Univ. of Washington Press, Seattle, WA, 1981).
6. J. M. Shenker et al. Mar. Ecol. Prog. Ser. 98, 31 (1993); M. J. Milicich and P. J. Doherty, ibid. 110, 121 (1994); S. Sponaugle and R. K. Cowen, ibid. 133, 13 (1996).
7. J. M. Shenker et al. Mar. Ecol. Prog. Ser. 98, 31 (1993); M. J. Milicich and P. J. Doherty, ibid. 110, 121 (1994); S. Sponaugle and R. K. Cowen, ibid. 133, 13 (1996).
8. J. M. Shenker et al. Mar. Ecol. Prog. Ser. 98, 31 (1993); M. J. Milicich and P. J. Doherty, ibid. 110, 121 (1994); S. Sponaugle and R. K. Cowen, ibid. 133, 13 (1996).
9. P. J. Doherty and D. M. Williams, Oceanogr. Mar. Biol. Annu. Rev. 26, 487 (1988).
10. M. C. Meekan, M. J. Milicich, P. J. Doherty, Mar. Ecol. Prog. Ser 93, 217 (1993).
11. F. Takens, in Dynamical Systems and Turbulence, D. A. Rand and L. S. Young, Eds. (Springer, New York, 1981), pp. 366-381.
12. G. Sugihara and R. M. May, Nature 344, 734 (1990).
13. G. Sugihara, Philos. Trans. R. Soc. Lond. Ser. A 348, 477 (1994).
14. M. P. Wand and M. C. Jones, Kernel Smoothing (Chapman and Hall, London, 1995); Q. Yao and H. Tong, Stat. Sin. 4, 51 (1994).
15. M. P. Wand and M. C. Jones, Kernel Smoothing (Chapman and Hall, London, 1995); Q. Yao and H. Tong, Stat. Sin. 4, 51 (1994).
16. -ed/d̄
17. Best forecasts were given by a three-term autoregressive model (AR3); however, other AR models performed similarly.
18. M. J. Milicich, Mar. Ecol. Prog. Ser. 110, 135 (1994); _, M. G. Meekan, P. J. Doherty, ibid. 86, 153 (1992).
19. M. J. Milicich, Mar. Ecol. Prog. Ser. 110, 135 (1994); _, M. G. Meekan, P. J. Doherty, ibid. 86, 153 (1992).
20. Neighborhood regression, S-maps, and linear kernel regression are all related in the use of a weighting function to control the contribution of points to local linear surface: a step function, an exponential, and (typically) a probability density function with area one such as the normal, respectively. In many situations, kernel regression has been shown to outperform simple neighborhood regression, because the contribution of points to the forecast diminishes as a smooth function of distance. S-maps also have this property, however, and in this case, give very similar results (Table 1).
21. Data on P. amboinensis larval supply contained 35 days with no individuals sampled. Decreasing the taxonomic resolution to the family level reduced this to 29 days. S-map analysis gave similar results (P. A. Dixon, M. J. Milicich, G. Sugihara, data not shown) when applied to all pomacentrid species. Also, non-linear physical models performed comparably when constructed for P. amboinensis alone.
22. The specific choices of both of these variables were not critical. Because of the deterministic nature of lunar phase, many choices of lags for nighttime irradiance performed similarly (although the best lag coincided with the mean age of the larvae). Other lag choices for cross-shelf wind speed also gave comparable results.
23. J. M. Leis, Mar. Biol. 90, 505 (1986).
24. For each point in the larval supply series, we identified its 30 nearest neighbors in model space, and calculated the average value of each physical variable and larval supply of those neighbors. We then sorted the averages for the physical variables, binned them into groups of ten, and plotted the values of the bins against the corresponding binned values of larval supply.
25. I. C. Stobutzki and D. R. Bellwood, J. Exp. Mar. Biol. Ecol. 175, 275 (1994).
26. W. C. Legett and E. DeBlois, Neth. J. Sea Res. 32, 119 (1994).
27. We are most grateful to the following: R. Alford, L. Bersier, R. Brown, M. Casdagli, H. Choat, G. Jones, H. Hastings, A. Hobday, T. Hughes, J. Lough, R. Penner, and G. Russ. Supported by endowment funds from the John Dove Isaacs Chair in Natural Philosophy, the Office of Naval Research, and Deutsche Bank NA (P.A.D. and G.S.); and by Griffith University, Lizard Island Research Station (Australian Museum); Australian Institute of Marine Science; and James Cook University (M.J.M.).