The sprat fat content variability in connection with

The sprat fat content variability in connection with long-term environmental changes in the Black Sea V. N. Nikolsky and G. E. Shulman Institute of Biology of the Southern Seas (IBSS), Nakhimov Av. 2, Sevastopol 99011, Ukraine e-mail: nikolsky@ibss. iuf. net shulman@ibss. iuf. net

Main goals of time series analysis: (a) identifying the nature of the phenomenon studied (b) forecasting (predicting future values of the time series variable)

The objectives (i) Weather the observed variations of the indicator reflect adequately the real interannual variability of the sprat food supply? (ii) In what degree the observed variations of the indicator are connected to long-term environment variability? (iii) Can we predict the fat content in Black Sea sprat population using available data?

Annual dynamics of the Black sea sprat fat content (M ± SD) 16. 0 14. 0 Fat content, % 12. 0 10. 0 8. 0 6. 0 4. 0 2. 0 0. 0 spawning I feeding period II IV V VI VII Month VIII spawning IX X XI XII I

Sample locations of the Black sea sprat (after Minyuk et al. , 1997) 1 2 3

Summary of spatial variability in the sprat fat content data (single factor analysis of variance)

Summary of interannual variability in the sprat fat content data (single factor analysis of variance) Spatial variability exceeds 56 % of total variability Standard error of single observation amounts 1. 3

Long-term dynamics of sprat fat content from 1960 to 2001 16 Fat content, % 14 +SD 12 10 -SD M=11. 74 SD=1. 71 8 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005

Long-term dynamics of sprat fat content compared with two first principal components of environmental variability (acordingly to Daskalov, 2003) 3 16 2 14 0 12 FAT, % PC 1 1 -1 10 -2 -3 8 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 3 16 2 PC 2 1 12 0 10 -1 -2 8 1960 1965 1970 1975 1980 Principal components (Daskalov, 2003) 1985 1990 1995 2000 2005 Sprat fat content FAT, % 14

Sprat fat content dynamics (detrended) in comparison with the 2 nd principal component of the Black sea ecosystem variability 4 2 3 2 1 1 0 0 -1 -1 -2 -3 1960 -2 1970 1980 1990 All data: R = 0. 548 Except data with < 3 points per year: R = 0. 727 Except data with < 5 points per year : R = 0. 871 2000 PC 2 Fat content anomalies, % (acordingly to Daskalov, 2003)

Main loadings of the input variables to the 2 nd principal component (according to Daskalov, 2003) Sprat fat content 3 + Sea level pressure + Total river inflow + Inorganic phosphorus 16 2 14 12 FAT, % PC 2 1 + Hypoxia zone (? ) + Hypoxia zone 0 10 -1 -2 8 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 - Hydrogen sulphide - Zooplankton (E) - Pleurobrachia pileus - Phytoplankton during bloom - Mytilus biomass + Phytoplankton + Whiting recruitment + Anchovy recruitment + Horse mackerel recruitment

Summary of autocorrelation test Autocorrelation function Partial autocorrelation 0. 4 0. 2 0 0 -0. 2 -0. 4 0 4 8 12 Time lag, years 16

Summary of correlation tests for sprat fat content Variables Sprat fat content (t-1) Sprat fat content (t-3) Sprat biomass (t-1) Sprat biomass (t-2) Sprat biomass (t-3) Mean annual SST(t-4) Mean winter SST(t-2) Phytoplankton NW (t-1) Phytoplankton NW (t-2) Phytoplankton E (t-4) Correlation coefficient 0. 43 0. 42 0. 34 0. 35 0. 36 – 0. 46 – 0. 31 0. 52 0. 39 0. 48 Only significant correlation coefficients (p<0. 05) are presented

Observed and predicted indices of sprat fat content (linear model) 16, 0 Observed Predicted 15, 0 14, 0 13, 0 ? 12, 0 11, 0 ? 10, 0 9, 0 8, 0 1964 1968 1972 1976 1980 1984 FATt = f (FATt-1, FATt-3, SSTt-4) 1988 1992 1996 2000 2004 r = 0. 69 (R 2 = 0. 47)

Conclusion We are need long-term series data. Let they will be good data.

References Afifi A. A. , Azen S. P. (1979). Statistical analysis. A computer oriented approach. Academic Press, New York, San Francisco, London. Daskalov G. M. (2003). Long-term changes in fish abundance and indices in the Black Sea. Mar. Ecol. Prog. Ser. , 255, 259– 270. Daskalov G. M. , Grishin A. , Mihneva V. Ecosystem time-series analyses in the Black sea // Mediterranean biological time series. CIESM Workshop monographs. – Monaco, 2003. – P. 31– 36. Minyuk G. S. , Shulman G. E. , Shchepkin V. Ya. , Yuneva T. V. (1997). Black Sea sprat: the relationship between lipid dynamics, biology and fishery. Ekosi-Hydrophysica, Sevastopol, Ukraine (in Russian) Shulman G. E. (1974). Life cycles of fish. Physiology and biochemistry. Hulsted Press, John Wiley and Sons, New York. Shulman G. E. , Chashchin A. K. , Minyuk G. S. , Shchepkin V. Ya. , Nikolsky V. N. , Dobrovolov I. S. , Dobrovolova S. G. , Zhigunenko A. S. (1994). Long-term monitoring of Black Sea sprat condition. Doklady Akademii Nauk, 335, 124– 126 (in Russian). Shulman G. E. , Love R. M. (1999). The Biochemical Ecology of Marine Fishes. In: Advances in marine biology, vol. 36, Academic Press, London. Shulman G. E, Nikolsky V. N, Yuneva T. V. , Minyuk G. S. , Shchepkin V. Ya. , Shchepkina A. M, Ivleva E. V. , Yunev O. A. , Dobrovolov I. S. , Bingel F. , Kideys A. E. (2005). Fat content of Black Sea sprat as an indicator of fish and ecosystem condition. Mar. Ecol. Prog. Ser. , 293, 201– 212. Zar J. H. (1984). Biostatistical analysis. 2 nd edn. , Prentice Hall, Englewood Cliffs, NJ.