Land use and population growth of Primula veris: an experimental demographic approach
Published source details
Ehrlén J., Syrjänen K., Leimu R., Garcia M.B. & Lehtilä K. (2005) Land use and population growth of Primula veris: an experimental demographic approach. Journal of Applied Ecology, 42, 317-326.
Published source details Ehrlén J., Syrjänen K., Leimu R., Garcia M.B. & Lehtilä K. (2005) Land use and population growth of Primula veris: an experimental demographic approach. Journal of Applied Ecology, 42, 317-326.
Cowslips Primula veris (a perennial rosette-leaved herb) mainly occur in temperate European grasslands and woodland pastures. Occurrence in the northern part of its range is linked with grazing and mowing. However, it can persist several decades after management has ceased. To assess how grazing influences population viability of cowslips, four treatments were applied to abandoned grasslands in southeast Sweden (approaching the northern limit of its range) and the demographic response was recorded over 3 years.
Study area: Cowslips were studied in a 1 × 1 km area near the town of Tullgarn, about 60 km south of Stockholm.
Experimental design: In May 1996, six sites within the area were selected where the previous mowing or grazing management had been abandoned. At each, four 2 m × 2 m plots were established. Each plot was randomly assigned one of four treatments:
1) control (left);
2) vegetation removal - vegetation cut to approximately 2 cm height at peak of flowering in June (1996 and 1997), leaving cowslips intact;
3) litter plus vegetation removal - raking in early May 1997 and 1998;
4) soil disturbance plus vegetation removal - in June above-ground parts and as much as possible of below-ground parts removed, leaving cowslips. This treatment also disturbed the upper soil layer.
Treatments were designed to mimic components of management considered important for cowslip performance. Within each plot the locations of all individuals were mapped.
Plant parameters: Each cowslip was monitored twice each year (1996 to 1998). In June, the number of leaves, length of the largest leaf and proportion of leaf area removed by herbivores was recorded, and flower number was counted. In August the numbers of intact and damaged seed capsules were counted. At each visit emerged seedlings were searched for and included in the further study.
Recruitment experiments: Cowslip seedling emergence and survival was examined by seed-sowing experiments. The treatments were as in the main experiment. In 1996 four plots, 10 × 10 cm, per site and treatment (total n= 96) were established and randomly assigned to one of four sowing densities:
1) no seeds, only natural seed rain;
2) 50 seeds sown;
3) 200 seeds sown;
4) 500 seeds sown.
The seeds added (locally collected and sown within 24 h of collection) produced an increase in seed density corresponding to 5,000–50,000/m². The sowing densities were much higher than natural seed rain (average 13.5 seeds/dm², n= 6 permanent plots). In 1997 a second seed-sowing experiment was performed using only 50 seeds/plot. Plots were censused at least once every month in May-July 1996 to 1998.
Modelling: Transition matrix models were constructed to predict the effects of treatments on cowslip population growth rate.
Survival and fitness: Average cowslip survival (1996-1998) did not differ between treatments or site but was positively correlated with the average initial size. The length of the largest leaf was significantly influenced by treatment. On average, established individuals in litter removal + vegetation removal plots and in soil disturbance + vegetation removal plots had smaller leaves than individuals in control plots. Juvenile plants (i.e. the smallest non-seedling category) in 1996, were twice as large in soil disturbance + vegetation removal plots than in the other treatments. After 2 years, juveniles in the soil disturbance + vegetation removal treatment were three times as large as other juveniles. The proportion of individuals that flowered and the number of flowers per flowering individual differed between sites but there was no treatment effect. Fruit set was also unaffected by treatment.
Seed sowing: Sowing increased average seedling number from 0.8 in controls to 75.2 in plots with 500 seeds sown. The proportion of seeds emerging as seedlings the first year after sowing was significantly higher at low sowing densities and after soil disturbance + vegetation removal treatments (e.g. c. 55% at 50 seed rate and 32% at 500 rate). Although a smaller proportion of seeds emerged at high sowing densities, the number of seedlings was still higher than controls. Also, the second year after sowing seedling number was higher in disturbance plots and in plots with higher sowing densities. The size of seedlings in their second year was not affected by sowing density or treatment. In the second sowing experiment in 1997, seedling emergence rates the first year after sowing were again significantly higher in soil disturbance plus vegetation removal plots than in control plots.
Effects on population growth rate: Matrix models showed that cutting surrounding vegetation had no effect on population growth rate (λ). Litter removal + vegetation removal increased λ to 1.46 (0.35 above controls). The increase after soil disturbance + vegetation removal was more pronounced at 1.60 (0.49 above controls), this mainly because of a higher probability that small individuals would reach an intermediate size. The second most important factor was more seedlings in disturbed plots. Differences in the other stages contributed little to differences in population growth rate.
Conclusions: In this study, seedling establishment was identified as a key phase in the cowslip life cycle, and litter accumulation as a key factor acting against small seedling survival and seedling establishment.