CARLINE: PRODUCTION BY WILD BROOK TROUT 



outlet joined this section at its midpoint. Maxwell 

 outlet (200 m) was sampled in 1969 and 1972. 

 All trout were measured, about 25'7( were 

 weighed, and fall fingerlings and spring yearlings 

 were permanently marked by fin removal. 



Sampling dates in ponds varied from year to 

 year. I estimated mean lengths and weights of 

 each cohort on 15 April and 15 September so that 

 growth rates from different years could be 

 compared. Mean weights of individuals in each 

 year class were determined graphically by 

 assuming constant instantaneous rates of growth. 

 By graphically estimating mean length, I as- 

 sumed length increased linearly between succes- 

 sive estimates. Most of the adjustments in length 

 or weight involved extrapolating over periods 

 <2 wk and size changes were usually <5%. 



Year class biomass was estimated by multiply- 

 ing mean weights of individual trout by year class 

 density. Biomasses in spring and fall were 

 averaged to calculate mean biomass (B ). I followed 

 procedures suggested by Ricker ( 1975) to calculate 

 instantaneous rates of growth by weight (G), total 

 mortality (Z), natural mortality (M), and fishing 

 mortality (F). Production, the product of G and 

 B, was computed semiannually for each cohort. 

 Production by fingerling trout was calculated 

 from emergence (1 March) to time of spring 

 population estimate and from spring to fall. 

 A mean weight of 0.04 g was assigned to emergent 

 fry (Hunt 1966). I assumed that instantaneous 

 growth and mortality rates from emergence to 

 fall were constant. Mean annual biomass of each 

 cohort was calculated by weighting mean bio- 

 masses in the two intervals according to interval 

 lengths. Annual production was calculated by 

 summing production during the two intervals and 

 expressing the sum for 365-day periods. 



Potential egg production for each population 

 was estimated from numbers of mature females 

 in fall and from a relationship between total 

 length of females and number of eggs. Fecundity 

 of trout was determined from 83 females that were 

 collected from two ponds in the same watershed 

 as the study ponds. Trout were collected in early 

 October, about 2 wk prior to spawning. Mature 

 ova could be easily distinguished from recruit- 

 ment eggs on the basis of size and color ( Vladykov 

 1956). Data on trout length, weight, and total 

 number of eggs were fitted to linear, curvilinear, 

 and logarithmic regression models. A linear 

 regression of total trout length and number of 

 eggs yielded the highest correlation coefficient. 



At Clubhouse and Hoglot springs, densities of 

 some year classes increased during sampling 

 intervals because of immigration from outlets or 

 adjoining streams. Numbers of immigrants were 

 estimated by first calculating expected densities 

 at the end of sampling intervals by using mean, 

 age-specific mortality rates; expected densities 

 were then subtracted from actual densities. If the 

 expected number of trout at the end of an interval 

 was within 107 ( of the actual number or the 

 difference was negative (suggesting emigration), 

 it was assumed no immigration had occurred. 

 Age-specific mortality rates for trout in Club- 

 house and Hoglot springs were estimated from 

 permanently marked fish. For some age groups, 

 mortality rates could not be estimated because 

 of insufficient numbers of marked fish. In these 

 instances I used age-specific mortality rates of 

 the population in Maxwell Springs, where immi- 

 gration did not influence year class densities 

 (discussed later). 



Harvest of trout from Clubhouse and Hoglot 

 springs was estimated from partial creel surveys 

 in 1969, 1970, and 1972. State-wide angling 

 regulations included a bag limit of 10 trout/day 

 and minimum length of 154 mm (6 in). Census 

 clerks worked five randomly.chosen days per week 

 during the entire fishing season, mid-May to mid- 

 September. Catch rates were estimated from data 

 collected during interviews of anglers, and fishing 

 pressure was calculated from instantaneous 

 counts of anglers (Lambou 1961). Harvest was 

 estimated monthly from the product of the hours 

 of fishing and numbers of trout caught per hour. 

 Harvested trout were measured, examined for 

 permanent marks, and scales were collected from 

 a sample of the catch. Harvest data from 

 Maxwell Springs were compiled by the owner and 

 others who fished the pond. Ages of harvested 

 trout from Clubhouse and Hoglot springs were 

 determined from scales and size distributions of 

 permanently marked fish. Ages of trout harvested 

 from Maxwell Springs were estimated from 

 comparisons of lengths of harvested trout with 

 lengths of known age fish in spring and fall. 



RESULTS 



Population Densities and Biomass 



Electrofishing was the most efficient method of 

 collecting trout in these shallow ponds. Popula- 

 tion estimates derived from collections with trap 



753 



