Variations in sampling procedure and frequency affect estimates of recruitment of barnacles

The influence of variations in sampling procedure and frequency on estimates of recruitment of the barnacle Semibalanus balanoides (L.) on an intertidal rocky shore at Sandy Cove, Nova Scotia, Canada was determined from approximately daily monitoring of recently settled individuals. The effect of sampling procedure (removing or not removing barnacles after sampling) on estimates of recruitment varied with intertidal height. There was no effect of sampling procedure in the mid-intertidal zone, but in the low zone, where recruitment was on average 4 times larger, the estimate of recruitment was significantly greater from quadrats where barnacles had been removed after sampling. Estimates of recruitment and post-settlement mortality decreased exponentially as the sampling frequency decreased. In both the mid- and low-intertidal zones, significantly larger estimates of recruitment and post-settlement mortality were obtained when sampling every 1.3 d compared to sampling ca every 2 d. These findings indicate that comparisons of results between studies will be difficult if estimates of recruitment or post-settlement mortality are confounded by variations in sampling frequency. Unless individuals can be tracked over time, estimates of recruitment made by sampling without removal of recruits will tend to underestimate settlement by incorporating some post-settlement mortality, the magnitude of which will be directly related to the interval between samples. Without quantitative data on the effect of sampling frequency on estimates of recruitment and post-settlement mortality, results of tests of hypotheses requiring accurate estimates of recruitment or post-settlement mortality may be compromised if sampling is not done as frequently as possible.