Three-dimensional models represent seed moisture content as a function of relative humidity and temperature

Seed moisture content (MC) has been considered the most important factor controlling physiological reactions in seeds, and NIC: changes with relative humidity (RH) and temperature (T). This relationship is revealed by studying the interaction of RH and T at equilibrium. Cucumber (Cucumis sativus L.), lettuce (Lactuca sativa L.), maize (Zea mays L.), onion (Allium cepa L.), pea (Pisum sativum L.), and watermelon (Citrullus lanatus M. & N.) seeds were equilibrated over sulfuric acid (1% RH) and various saturated salt solutions (5.5% to 93% RH) at temperatures from 5 to 50 degrees C. Best-fit subset models were selected from the complete third-order model MC = beta(0) + beta(1)*RH + beta(2)*T + beta(3)*RH2 + beta(4)*T-2 + beta(5)*RH*T + beta(6)*RH3 + beta(7)*T-3 + beta(8)*RH*T-2 + beta(9)*RH2*T, using Mallows' minimum Cp as the selection criterion. All six best subset models (R-2, 0.98 to 0.99) had the same functional form, MC = beta(0) + beta(1)*RH + beta(2)*T + beta(3)*RH2 + beta(5)*RH*T + beta(6)*RH3 + beta(9)RH(2)*T. Coefficients had essentially the same respective values among all species except onion and pea, for which some coefficients were statistically different from those of the other species (P less than or equal to 0.05). All models indicated that seed MC increased as RH increased and decreased as T increased; but RH had the greater influence. The inverse relationship between seed MC and T, although slight, was evident in the response surfaces. The interaction effect of RH and T on MC was significant at P less than or equal to 0.001. These results suggest that orthodox seed species respond similarly to T and RH. This in turn suggests that a common model could be developed and used for optimizing seed storage environments.