FISHERY BULLETIN: VOL. 80, NO. 4 



(r 2 = 0.72), and average temperature of the two 

 coldest consecutive weeks (r 2 = 0.82). All rela- 

 tionships were significant (P<0.01). If the aver- 

 age temperature of the two coldest consecutive 

 weeks was <6°C, then shrimp landings were be- 

 low average (Fig. 4). 



Although average winter temperatures (De- 

 cember-March) and average monthly tempera- 

 tures (January plus February) each accounted 

 for significant portions of the variance in annual 

 landings, the strongest relationship was found 

 between landings and the average of the two 

 coldest consecutive weeks. This may arise be- 

 cause, as Williams (1969b) stated, averages do 

 not adequately represent extremes since they 

 dampen the duration and intensity of the cold. 

 Using a process of expressing temperature in 

 heating degree days, Williams (1969b) postu- 

 lated that the catch of all species of penaeid 

 shrimps of a given year in North Carolina may 

 depend on net heating degree days during the 

 coldest preceding 6 mo (November-April). He 

 found the poorest catches in cold years (1958, 

 1961, and 1963) for all three species combined 

 and further suggested that warm years may be 

 as beneficial as cold years are deleterious. 



The role of temperature on activity and osmo- 

 regulation of pink shrimp has been documented 

 (Williams 1955a, 1960). The lower temperature 

 for activity under experimental conditions was 

 about 14°-16°C; complete cessation of activity 

 was noted below about 10°C. Below 8.8°C, osmo- 

 regulatory ability was impaired. Pink shrimp 

 may survive periods of winter cold by burying 

 deeply into the substrate, and Fuss and Ogren 

 (1966) reported that below 14°C, shrimp remain 

 buried, abandoning the usual pattern of noctur- 

 nal emergence. Laboratory experiments showed 

 pink shrimp to be more tolerant to combinations 

 of low salinity and low temperature than brown 

 shrimp, and this may explain the occurrence of 

 pink shrimp in North Carolina estuaries during 

 the winter (Williams 1960). In contrast, fall and 

 midwinter brown shrimp immigrants do not 

 survive cold weather as well. The usual recruit- 

 ment period for brown shrimp postlarvae is Feb- 

 ruary and March; for white shrimp it is June 

 through September (Williams 1965). 



Because osmoregulation is impaired at low 

 temperatures, we considered that low salinity 

 could increase mortality caused by low tempera- 

 tures. Although salinity records were not avail- 

 able, we compared local rainfall measurements 

 with pink shrimp landings from 1962 to 1981. 



landings (k fl )= 75648(T| - 213044 

 r 2 =0 82 



s - 



1 2 3 4 5 6 7 8 9 10 11 12 



AVERAGE OF THE TWO COLDEST CONSECUTIVE WEEKS 

 TEMPERATURE (°C) 



Figure 4.— Relation of pink shrimp landings in North Caro- 

 lina to the average water temperature of the two coldest con- 

 secutive weeks in the Newport River estuary. Numbers by dots 

 represent year of landings. (Actual landings of 198,000 kg 

 were predicted to be 165,200 kg, based on the coldest average 

 biweekly period temperature in 1982 of 5.0°C.) 



We found no correlation between rainfall and 

 landings (r 2 = 0.003) (Table 1). Further, rainfall 

 added no significant contribution to the explana- 

 tion of variance in landings data when it was in- 

 cluded with temperature as a predictive variable 

 (multiple R 2 = 0.826). Both the driest (18.0 cm in 

 1965) and wettest (49.3 cm in 1974) winters oc- 

 curred in years when landings were very large, 

 approximately 500 metric tons. Williams ( 1969a) 

 also found no significant relationship between 

 rainfall and total catch of all shrimp species. 

 However, Hunt et al. 3 reported that salinities 

 >10%o and temperatures >20°C during April 

 and May are necessary for good brown shrimp 

 harvests in North Carolina. Similarly, Gunter 

 and Hildebrand (1954) found a strong correla- 

 tion of total rainfall and white shrimp catch in 

 Texas. 



Deviations from the shrimp landings-temper- 

 ature relationship may, in part, be due to the 

 process of estimating landings. Errors may in- 

 clude improper species identification by fish 

 dealers, lack of accuracy in estimated weight 

 landed, and incomplete landing coverage. The 

 direct trading of shrimp to private individuals 

 by numerous part-time fishermen, plus the rec- 

 reational landings, neither of which is reported, 

 undoubtedly causes an underestimate of total 



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

 berg. 1980. Relationship between environmental factors 

 and brown shrimp production in Pamlico Sound, North Caro- 

 lina. N.C. Dep. Nat. Resour. Community Dev. Div. Mar. Fish. 

 Spec. Sci. Rep. 33, 29 p. 



766 



