HETTLER and CHESTER: WINTER TEMPERATURE AND SPRING LANDINGS OF PINK SHRIMP 



is, winter lows were displaced farther from the 

 mean than were summer highs. 



As a group the years 1962-67 were cooler in the 

 summer and warmer in the winter than 1974-81. 

 Lower temperatures during cold months re- 

 flected a series of very cold winters in the mid- 

 1970s (Diaz and Quayle 1980). An analysis of 

 covariance showed that the average yearly tem- 

 perature of the 1962-67 year group was signifi- 

 cantly different (P<0.05) from the 1974-81 

 group. Either a systematic calibration bias was 

 introduced by different observers, thermo- 

 graphs, and recording locations, or tempera- 

 tures were actually more extreme in the latter 

 year group. Calibration bias does not satisfac- 

 torily explain how both high and low tempera- 

 ture extremes could occur, but a climate phe- 

 nomenon can be cited. According to R. G. Quayle, 

 NOAA National Climatic Center, Asheville, 

 N.C., the 1962-67 winters were less variable in 

 daily temperature means than the 1973-81 win- 

 ters. For example, the January and February 

 monthly mean air temperatures at Wilmington, 

 N.C., and Cape Hatteras, N.C., were not differ- 

 ent between the two year-groups, but the stan- 

 dard deviations of the monthly means were sig- 

 nificantly different at both stations between 

 year-groups (Table 2). Thus, our recorded de- 

 pressions in weekly winter temperatures in the 

 1973-81 group probably reflect more extreme 

 actual fluctuations. 



Table 2. — Comparison of means and standard devia- 

 tions of January and February air temperatures at two 

 North Carolina coastal stations for year-groups 1962-67 

 and 1973-81. 



Air- Water Temperature Relation 



Close thermal coupling between air and water 

 has been found in shallow estuaries. Roelofs and 

 Bumpus (1953) reported that water temperature 

 in Pamlico Sound showed a seasonal cycle closely 

 related to air temperature. Lindner and Ander- 

 son (1956), documenting a winter kill of white 

 shrimp in south Atlantic and Gulf of Mexico 

 waters of the United States, also referred to a 



close relationship between air temperature and 

 surface water temperature. Smith and Kierspe 

 (1981) presented a model of air-water heat ex- 

 changes in a shallow estuary and suggested that 

 their model could reduce the need for in situ in- 

 strumentation while providing for close approxi- 

 mation of daily average temperatures. 



For the purpose of using air temperatures as 

 proxy data for missing water temperatures 

 (1968-72, Table 1), we decided to examine the re- 

 lation between local air and water temperatures 

 (Fig. 3). Although air temperature fluctuations 

 were accompanied by a predictable shift in 

 water temperatures over the entire range (r 2 = 

 0.97), at water temperatures below 12°C the rela- 

 tionship was not as useful ( r 2 = 0.68). We believe 

 that water temperatures rather than air temper- 

 atures are required for acceptable predictions of 

 fishery yields in estuaries. 



3 2r 



26 



3 



< 



tr 



20 



■saP' 



.... '-I." 





_L 



_L 



8 14 20 



AIR TEMPERATURE (°C) 



26 



32 



Figure 3.— Correlation of Morehead City, N.C., average 

 weekly air temperatures with Newport River estuary (Pivers 

 Island) average weekly water temperatures over 339 consecu- 

 tive weeks from 1974 to 1981. (Intercept = 2.21. slope = 0.96, 

 r 2 = 0.97.) For points below a water temperature of 12°C. r 2 = 

 0.68; and this relationship is not considered useful for predic- 

 tive purposes during winter. 



Relationship Between Temperature, 

 Rainfall, and Pink Shrimp Landings 



The February through July pink shrimp land- 

 ings were considered a dependent variable to be 

 plotted against various combinations of winter 

 temperature data (Table 1). Landings were re- 

 gressed on average winter water temperature 

 from December through March (r 2 =0.79), aver- 

 age temperature in January and February ( r 2 = 

 0.80), average temperature of the coldest week 



765 



