IDENTIFYING CLIMATIC FACTORS INFLUENCING 

 COMMERCIAL FISH AND SHELLFISH LANDINGS IN MARYLAND 1 



Robert E. Ulanowicz, 2 Mohammed Liaquat Ali, 3 Alice Vivian, 4 Donald R. Heinle, 5 

 William A. Richkus, 6 and J. Kevin Summers 6 



ABSTRACT 



In five of the seven most important commercial fisheries of Maryland an appreciable portion of the 

 annual variations in catch can be linked with past fluctuations in the physical environment. The 

 harvest figures were compared with appropriate annual characterizations of 40 years of daily 

 environmental records using a variation of the stepwise multiple linear regression technique. The 

 criterion for entry of a term into the regression was how well the given variable improved the pre- 

 diction of a randomly chosen independent subset of catch figures. The identification of spurious 

 predictor variables becomes less probable under this criterion. The results should help in the 

 organization of further research and management concerning these species and may afford esti- 

 mates of catches 1 or more years into the future. 



Annual population levels of commercially har- 

 vested fish and shellfish usually fluctuate widely 

 over the years. Such variation is often attributed 

 to the influence of important environmental var- 

 iables, such as water temperature, upon spawn- 

 ing success (Sissenwine 1978). Environmental 

 variables may directly affect the mortality rates 

 of prerecruits or indirectly exert influence by 

 altering the abundance of forage or predators. 

 Many other aspects of the ecosystem may also 

 alter population levels (Cushing 1975); however, 

 exact causative mechanisms in most fisheries 

 are seldom known. 



Year-to-year fluctuations in the abundance of 

 exploited species will determine in part the 

 magnitude of annual harvest of those species. 

 But the relationship will not be completely deter- 

 ministic, since landings will also be influenced 

 by socioeconomic factors (e.g., prices and costs as 

 they affect effort) as well as biological factors un- 

 related to exploitation (Ricker 1978). Despite 



•Contribution 1235, Center for Environmental and Estua- 

 rine Studies of the University of Maryland. 



2 University of Maryland, Chesapeake Biological Labora- 

 tory, Solomons, MD 20688. 



3 University of Maryland, Chesapeake Biological Labora- 

 tory, Solomons, Md.; present address: Senior Scientific Officer, 

 Fisheries Campus Chandpur, P.O. Baburhat, Dist Comilla, 

 Bangladesh. 



4 University of Maryland, Chesapeake Biological Labora- 

 tory, Solomons, Md.; present address: 517 Flag Harbor Drive, 

 St. Leonard, MD 20685. 



5 University of Maryland, Chesapeake Biological Labora- 

 tory, Solomons, Md.; present address: CH2M Hill, 1500 114th 

 Ave. SE., Bellevue, WA 98004. 



6 Martin Marietta Corporation, Environmental Center, Bal- 

 timore, MD 21227. 



these many complicating factors, significant cor- 

 relations between various environmental vari- 

 ables and commercial landings of various species 

 have been found in a number of fisheries. Dow 

 (1977), for example, showed that temperature 

 correlates well with the landings of 24 species of 

 finfish, Crustacea, and mollusks off the coast of 

 Maine. Sutcliffe( 1972) found freshwater input to 

 St. Margaret's Bay to be a good indicator of fish- 

 eries production, possibly because of the stimula- 

 tion of production caused by the nutrients in the 

 runoff water. However, in neither case were the 

 observed relationships demonstrated to help in 

 predicting harvests, nor were the specific mech- 

 anisms responsible for the observed relation- 

 ships rigorously delineated. In contrast, a regres- 

 sion model of brown shrimp landings off North 

 Carolina, using temperature and salinity as in- 

 dependent variables, was found to be a reason- 

 ably accurate predictor of future landings (Hunt 

 et al. 7 ). Hunt's model has proven to be a useful 

 management tool for this fishery, helping fisher- 

 men to decide how to gear up for the coming sea- 

 son. 8 Multiple linear regression has likewise 

 been employed to explain variations in catch 

 (e.g., Flowers and Saila 1972; Driver 1976). Only 



Manuscript accepted January 1982. 

 FISHERY BULLETIN: VOL. 80, NO. 3, 1982. 



7 Hunt, J. H., R. J. Carroll, V. Chinchilli. and D. Franken- 

 burg. 1979. Relationship between environmental factors 

 and brown shrimp production in Pamlico Sound, North Caro- 

 lina. Completion Report for Project 2-315-R, North Carolina 

 Department of Natural Resources, Morehead City, N.C., 37 p. 



8 M. W. Street, Chief, Fisheries Management Section, Divi- 

 sion of Marine Fisheries, North Carolina Department of Nat- 

 ural Resources and Community Development, P.O. Box 769, 

 Morehead City, NC 28557, pers. commun. April 1981. 



611 



