Summary
Phenotypic plasticity and genotypic variation were studied in Agropyron repens L. (Beauv.) collected from populations in two grassland communities which differed in the length of time since the last major disturbance. Twenty genotypes were collected from each population. Each genotype was vegetatively propagated, and subjected to six different treatments in a greenhouse. Phenotypic plasticity and genotypic variability were measured as across- and within-treatment standardized variances respectively. Patterns of plasticity were measured by genotype correlations across treatments. The results were presented graphically by the regression method of Garbutt and Zangerl (1983). Analysis of variance revealed significant population, genotype and treatment effects. Significant positive correlations between magnitude and variability of performance were found for all characters. Phenotypic plasticity and magnitude of performance were generally greater in plants collected from the older established field. Evidence for greater specialization in the older population was suggested by negative correlations between performance in the most favorable and least favorable treatments and by greater dissimilarity of genotype response across treatments. A more variable phenotypic response across treatments (i.e., higher plasticity) for plants from the older population may therefore be a consequence of specialization and not an adaptive trait per se.
Methodology
The study site was a hayfield at the Queen's University Biology Station in southeastern Ontario, Canada (44°34'02"N, 76°21'52"W). The hayfield was last ploughed in 1974 and sown with a seed mixture comprised of Trifolium pratense L., Phleum pratense, and Medicago sativa L. (G. Hughson, personal communication). Two adjacent 30 m x 100 m plots were selected from a relatively flat, homogeneous portion of the field. One of these plots was ploughed in spring 1984 and resown with the same species used in the 1974 planting. Prior to 1984 the field was mown annually for hay.
In May 1985, 20 plants of A. repens were randomly collected from each population. Collection sites were separated by at least 3 m to reduce the likelihood of sampling the same genotype twice. The plants, therefore, are loosely referred to as different genotypes, but only to imply that they represent the products of different zygotes. Each genotype was transferred to the greenhouse into 8-inch standard plastic pots filled with a standard potting mix. The 40 genotypes were grown for several weeks with periodic fertilizer application and above ground clipping to encourage tillering.
In late September 1985, 18 tillers were removed randomly from each genotype. Above and below ground material was clipped to 3 cm and 1 cm respectively to attain a uniform starting size. Each tiller was planted in a separate pot filled with sterile, grade 16, silicate sand. The 18 tillers of each genotype were subjected to six treatments (3 replicates per treatment). Treatments were chosen to represent a range of important biotic and abiotic conditions that the species may encounter in the field. The treatments were 1) Control--daily watering to keep the sand moist (20 ml); daily fertilization with Plant Prod 20:20:20 fertilizer (0.012 g per pot, per day); 4-inch standard plastic pots (9 cm deep, 10 cm top diameter). 2) Low water--watering every other day; fertilization every other day; pots as in control. After three weeks, this treatment did not appear to cause sufficient water deprivation. At this point, therefore, watering frequency was changed to once every three days. 3) Low nutrients--application of fertilizer once every 3 weeks; watering and pots as in control. 4) Defoliation--each plant was clipped to 5 cm high after weeks 3 and 6; watering, fertilizer and pots as in control. 5) Small pots--plants were grown in 2-inch standard pots (5.5 cm deep, 5.5 cm top diameter) which hold one third the volume of 4-inch standard pots; watering and fertilizer as in control. 6) Competition--pots were sown 4 weeks previously with 100 seeds (estimated by weight) of Phleum pratense (obtained from the same source as used to sow the 1984 plot); watering, fertilizer and pots as in control. Phleum pratense was an abundant species in both the 1974 and 1984 plots.
The experiment ran for 12 weeks in a heat-controlled greenhouse. Pots were rotated weekly to minimize experimental error involved with microenvironmental variability within the greenhouse. After 10 days, any tillers which had not become established were replaced by surplus tillers of the same genotype. If there was not a sufficient number of replacement tillers, dead tillers were recorded as missing values. At the end of the 12 weeks, the following characters were recorded for each plant: 1) shoot (above-ground) biomass, 2) root (below-ground, including rhizomes) biomass, 3) total biomass, 4) root/shoot quotient (root biomass/shoot biomass), 5) number of tillers, 6) height (measured from ground level to the auricle of the highest leaf).