2.3 Measurements of plant flowering phenology
We selected six dominant and common plant species for phenological observation, including three perennial C3 grasses,S. breviflora , C. ammannii and A. tenuissimum , one perennial C3 subshrub A. frigida , one C4 grass, C. songorica , and one perennial C4 subshrub, K. prostrata. Five individuals per plant species were randomly chosen and numbered in each plot, and their reproductive phenology was observed and recorded using the measured scoring method. Here, the phenological observation of a species started once the first flower buds were discovered. The observation frequency was once every 3 to 5 days. For grasses and non-grasses, a 6-point system (Dunne et al., 2003) and a 4-point system (Price & Waser, 1998) were used to record scores, respectively. The phenological phases observed were 0 = spikelet in boot, 1 = spikelet out of boot, 2 = exserted anthers and styles, 3 = developing seed, 4 = all seed dropped (Price & Waser, 1998). For non-grass plants, plant phenology can be divided into seven stages: 0, not flowering; 1. flower buds appear. 2. flowering. 3. flower senescence. 4. the fruit begins to develop. 5. the fruit begins to disperse. 6. The fruit falls off completely. Each phenological event was recorded as the corresponding day of year. The actual phenological sequences of plants were then used to fit Richard’s equation in 1STOPT software (Sherry et al., 2007). The date and duration of each phenological period were calculated based on the optimal simulation curve according to the Richards equation (Price & Waser, 1998).
Y = K/ (1 + a*exp(-b*X)) m
where ”K” is the maximum growth, that is, the maximum value in the phenology score; ”a” is the start-up parameter of the date when observation of the evaluated species began; ”b” is the reproductive phenological rate during the observed period; and ”m” is the curve shape variation parameter. All phenological scores were well described by the Richards equation (R2 > 0.99).
Four parameters can be determined by equation fitting and substituted into the conversion formula as follows.