I estimated variance-mean (s2-x ) relationships for gill net catches of 8 forage and sportfish species plus a composite group composed of all species combined. These relationships explained 75%-95% of the variation in loge(s2). Predictive equations for loge(s2) were back-transformed to a linear scale, adjusted to correct for transformation bias, and substituted into a standard equation for estimating sample size requirements as a function of the desired level of precision and expected sample mean. Sample size requirements for all species increased with an increase in the desired level of precision or with a decrease in the expected mean. Based on statewide mean catches, all species studied can be sampled with a precision ≤0.3 with a sample size of 25 gill nets and most could be sampled with a precision of ≤0.2 with 50 gill nets; these results represent approximate 95% confidence intervals about estimates of mean catch of x ± 0.6x (precision= 0.3) or x ± 0.4x (precision = 0.2). Equations for predicting sample size requirement presented in this paper are specific to Texas fisheries and a specific gill net configuration; however, they can be used to provide preliminary estimates of sample size requirements elsewhere.
