3910:
Calculating Average pH in Substrate Research: Should pH or [H+] Data Values Be Used?

Wednesday, August 4, 2010
Springs F & G
Eugene K. Blythe , Coastal Research & Extension Center, Mississippi State University, Poplarville, MS
Donald J. Merhaut , Department of Botany and Plant Sciences, University of California, Riverside, CA
Data on pH is often collected during research with container-growing substrates using the pour-through method or other extraction methods. A measure of acidity/basicity, pH is approximately equal to the negative of the base 10 logarithm of the molar hydrogen ion concentration ([H+]) of a solution. When analyzing pH data, a question that often arises among researchers is: Should the pH data values be converted to [H+] before calculating a mean value, then calculating the mean pH from the mean [H+]? However, a more general question may be preferable: What are appropriate measures of average acidity for substrate extract? When pH was measured from pour-through extract collected weekly over 12 weeks from 100 2.8-L containers of a compositionally uniform and accurately irrigated pine bark/peat moss/sand substrate, the distribution of the pH values was shown to be symmetrical (near-normal) from week to week, whereas the distribution the [H+] values tended to be asymmetrical (skewed to the right). When a distribution is symmetrical, the (arithmetic) mean and the median are both acceptable measures of location, or average value. (With a perfectly symmetrical distribution, the mean and median coincide.) When a distribution is skewed, the median is typically a preferable measure of location since the median is less sensitive to extreme values than is the mean. Although the skewness of the [H+] values caused the distributions of the weekly data to show statistically significant deviations from a normal distribution, the deviations were not extreme. When values of mean pH calculated from pH values were compared with median pH calculated from pH values, mean pH calculated from mean [H+], and median pH calculated from mean [H+], all differences were within 0.02 pH units. Therefore, mean pH calculated directly from the pH data values is shown to be a valid summary measure of average acidity/basicity for container-substrate extract, with no data conversion from pH values to [H+] values being necessary. Median pH calculated directly from the pH data values was also a suitable summary measure of average extract acidity/basicity.