Abstract
Seeds interaction and germination experiments replication.—It has been studied if the significant heterogeneity between replicates often found in germination experiments could be caused by seeds interaction.
Two cultivar of Triticum durum Desf. have been tested with three experiments series: I) germination of n = 5, 10, 15, 20, 25 caryopses per germinator with N germinators per treatment and nN = 500; II) germination of n = 5 and 25 caryopses per germinator with nN = 2500 per treatment; III) monthly germination experiments from 3 to 15 months after harvest, with n = 5 and 25 caryopses per germinator and nN = 500 per treatment.
The results show differences between rilevators and for the same rilevator change regulary along time of observation. Keeping time and rilevator both equal: if n = 25 heterogeneity between N replicates is statistically significant for about 60% of the experiments, and for the 15 % when n = 5; the germination percentage p is fitted by p = po + bn; for one cultivar b changes from positive to negative with seeds age.
The observed interaction is referred to the allelopathic autostimulations and autotoxicities as a case of autoblastism. The quantitative interpretation assumes that each seed releases on the substrate a quantity x i of a substance which has a linear effect on all seeds within the same germinator. Then the number of germinated seeds is g = np o + k[xbar]n 2 and the difference between the associations of interacting individuals is proportional to the one between the squares of the numbers of their components. The significant heterogeneity between replicates is interpreted as caused by heterogeneity of the substance distribution on the substrate. The inversion of the effect during postmaturation relates this interaction to the dormancy physiology. The occurrence of interaction calls for particular care during germination experiments.