Premium
Sampling requirements and the implications of reduced sampling effort for the estimation of annual zooplankton population and community dynamics in north temperate lakes
Author(s) -
Rusak James A.,
Montz Pamela K.
Publication year - 2009
Publication title -
limnology and oceanography: methods
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.898
H-Index - 72
ISSN - 1541-5856
DOI - 10.4319/lom.2009.7.535
Subject(s) - zooplankton , sampling (signal processing) , abundance (ecology) , community structure , ecology , population , temperate climate , ordination , sampling design , biology , environmental science , statistics , mathematics , demography , sociology , filter (signal processing) , computer science , computer vision
Detecting and understanding long‐term changes in zooplankton populations and communities requires sound sampling strategies and reliable estimates of annual abundance. We use a long‐term, highly resolved data set from four north‐temperate lakes to investigate the sampling requirements necessary to precisely quantify common zooplankton species and taxonomic aggregates as well as the influence that a reduced sampling regime has on our ability to investigate changes in community structure. Sample sizes required for estimating mean annual abundance (precision of 20%) ranged from 4–7 samples per year for total crustacean zooplankton abundance to as many as 34–86 samples per year for individual rotifer species, depending on the lake. Despite these often onerous sampling requirements for some taxa, patterns of population abundance were typically quite similar when compared using reduced (five samples per year) and extended (14 samples per year on average) data sets. The same 3‐fold reduction in sampling produced taxa‐specific shifts in community structure ranging between 6% to 20% for two‐dimensional solutions and between 9% to 21% in three dimensions, depending on the lake. Although community structure was not significantly different between reduced and extended sampling scenarios, we found that the correlations between taxa and ordination axes could change and that these differences increased with the dimensionality of the ordination. Although individual investigators need to determine acceptable boundaries of sampling precision based on the objectives of their study, it appears that enumerating 5–10 sample events per year may adequately document long‐term changes in populations and communities on an annual basis in many lakes.