CLAUDE E. BOYD and JOHN C. WILLIAMS* D AlA F-ROM ONE SAM l- taken at a single station are generallk used to pot- trav the Water qUa16t ol a pond, ex en though data to x erifv the efi ectix eness of t his prac- tice are ttnaxailable. W\hen multiple sarn- pling is employed, the additional samples XIProtcs~or, IDpatiment ot t-isheries and Allied Aquaculture and Associate F'i ote.or, Res~earchI D~ata \nis.o are usually taken ftrom dillerent depths heloxN a single station. I his piracticee no doUht r esults because ai atlin in x, ater (ILalitN vN ith depth has long been estahlished. but there is little information on \ariation in xx ater Lqualitv ait a single depth at dillerent stati ons. I herd fore. the pr esent research ),\as conducted to obtain data on \ariation in wxater quality at different locations in ponds to serxe as a guide in sampling pond xxaters. LEAFLET 100 FEBRUARY 1981 AGRICULTURAL EXPERIMENT STATION AUBURN UNIVERSITY GALE A. BUCHlANAN, DIRECTOR AUBURN UNIVERSITY, ALABAMA 101w r .z METHODS Seven water quality variables were se- lected for study: dissolved oxygen, tempera- ture, and Secchi disk visibility which were measured in situ, total hardness, filtrable orthophosphate, total particulate matter, and pH which were determined in the labo- ratory. These variables were selected partly because of their importance in fish culture, but the set also included variables expected to exhibit low, moderate, and high varia- tion. Variables were also chosen because they could be measured rapidly. This short- ened the time required for sampling and analysis, reducing variation in water quality because of natural diurnal fluctuations and changes during sample storage. Analytical procedures followed routines given by Boyd.1 Ponds were located on the Fisheries Re- search Unit, Auburn University Agricul- tural Experiment Station, Auburn Univer- sity, Alabama. Fourteen fish ponds ranging from 0. 1 to 25 acres in area and from 3 to 6 ft. in average depth were used. Each pond was sampled once between April 8 and June 9, 1977. Water samples were collected with a 90-cm water column sampler 2 , poured into 2-liter polyethylene bottles, and quickly transported to the laboratory for analysis. Dissolved oxygen, temperature, and Secchi disk visibility measurements were taken within a 3-foot radius of the location where each water sample was collected. Depending upon size, 6 to 24 samples and in situ mea- surements were obtained per pond. The sampling stations were randomly located in each pond. Sampling was initiated between 8 and 10 a.m. and completed within 15 to 45 minutes. RESULTS AND DISCUSSION Means and variances were calculated for data from each pond. However, since pond size had no influence on the magnitude of the variances, variances were pooled from all ponds to obtain a mean variance for each variable. The mean variance values and the ranges of means of variables for the 14 ponds are presented in table 1. Allponds had rela- tively low total hardness concentrations. However, means for the other variables covered the ranges generally expected during the growing season in most fish ponds in Alabama. The mean variance values, table 1, were used in the following equation 3 to calculate sample sizes: n 4s2 L 2 where: n - sample size (number of samples); ItLe -ie errorr: s - variance; allowableerror (95 percent probability level). Sample sizes for selected allowable errors are presented in table 2. The mean variance and the equation given above may be used to calculate sample size for any other desired allowable error. Once the deci- sion on sample size is made, the samples must be collected from randomly located sites in the pond. The requirement for samples was least for pH, temperature, and total hardness, inter- mediate for dissolved oxygen and filtrable orthophosphate, and greatest for Secchi disk visibility and total particulate matter, table 2. The large variability in Secchi disk visibility and total particulate matter reflects the influence of-phytoplankton on these measurements. Phytoplankton forms scums on pond surfaces which drift in response to wind action and form greater accumulations at some locations on pond surfaces than at others. 1 BOYD, C. E. 1979. Water Quality in Warm- water Fish Ponds. Auburn University (Ala.) Agri- cultural Experiment Sta. 359 p. 2 BoYD, C. E. 1973. Summer Algal Communities 3 SNEDECOR, G. W. 1956. Statistical Methods. and Primary Productivity in Fish Ponds. Hydro- The Iowa State University Press, Ames, Iowa. biologia 41: pp. 357-390. 534 p. [21 TABLE I. RANGE OF MEANS FOR WATER QUALITY MEASUREMENTS ON SAMPLES FROM 14 FISH PONDS AND MEANS OF VARIANCES FOR DATA FROM ALL PONDS Variable Dissolved oxygen (mg/liter) ............................ Secchi disk visibility (cm) ................................. Tem perature ( C ) ........................................ p H ...... ............................................... Total hardness (mg/liter as CaCO 3 ) ........................ Particulate organic matter (mg/liter)........................ Filtrable orthophosphate (pg/liter) .......................... TABLE 2. NUMBERS OF WATER SAMPLES R WATER QUALITY VARIABLES WIT WILL NOT EXCEEI Variable No. Dissolved oxygen S0.5 mg/liter ...................... 6 S1.0 mg/liter* ..................... 2 pH + 0.25 unit ......................... 2 ? 0.5 unit* ... ....................... ? 1.0 unit..... ..................... I Secchi disk visibility 2 cm ...........................40 5 cm * ...........................7 10 cm ...........................2 Temperature 0.50C ............................2 S1.00C* .. ........................ I *Sufficiently reliable for routine purposes. Range of Mean means variance 7.3- 15.1 0.37 23 - 143 40.83 20.4- 27.3 0.072 6.4 - 9.6 0.034 8.1 - 27.0 0.11 3.8 - 53.3 14.02 2 - 96 17.25 EQUIRED FROM PONDS TO ESTIMATE THE MEANS OF H A 95 PERCENT CERTAINTY THAT ERRORS D THE SPECIFIED AMOUNTS Variable No. Total hardness + 1.0 m g/liter* ..................... l Total particulate matter ? 1 mg/liter.......................54 ? 2 mg/liter*......................14 + 5 mg/liter........................3 Filtrable orthophosphate + 1 g/liter ... ..................... 69 S5 #g/liter* ........................ 3 + 10 #g/liter ... ..................... I [31 Information contained herein is available to all without regard to race, color, sex, or national origin. 14]