|
 Rodent and invertebrate predation on seeds and seedlings may have a strong impact on the population of seedlings surviving to
maturity. Predation has thus far been difficult to characterize at Pothole; while we may think that certain microsites are unsuitable for germination, it may be that a microsite is suitable for a seed predator, reducing the probability of seed
germination at that site. Thus, microsite and seed/seedling predation are easily confounded. A similar problem is the assumption that there are few seeds falling, and that they fall in fairly sparse patches. It may be that dense patches of seedfall
are attractive to predators.
The density of seeds and the effect of predation on them were tested by setting out seeds at densities of 6, 12, and 60 seeds/m2 with some plots excluding
vertebrate predators as a control. The densities were estimated based on examples from
the literature regarding Douglas-fir seed production and from estimates based on basal area and density of trees in the stand. The average density of natural seed fall was estimated to be between 5 and 11 seeds/m2. Fifty-four plots were
installed, twenty-seven of which excluded predators, which gave nine replications of each experimental unit.
The plots were located in the area west of the Pothole Creek permanent sample
plot, in order to isolate the predation study from other research traffic. Approximately 0.9 ha was chosen and the fifty-four plots were located using a
randomized transect system. The starting points for the four transects along a line extending due west from the Pothole Creek
plot were chosen randomly, and the transects extended north. The position of the plots along
each transect and the
order of treatments were also assigned randomly. Because flagging tape may attract seed predators such as birds, the plots were not conspicuously marked, but were carefully mapped. The plots are 0.5 m x 2 m, with the broad side facing south, and
with the southwest and southeast corners marked with galvanized nails. The seeds are placed within the plot in a regular array, with small galvanized steel nails marking the ends of each
seed line. The seeds were handled with gloves to avoid contamination. Care was also taken to avoid steep slopes to reduce the possibility that seeds would be washed away by snowmelt in the spring.
 The
predator exclusion cages were built from flexible hardware cloth
that was draped over rigid wire arcs and extended beyond the margins
of the plot by up to 25 cm on each side. The edges of the hardware
cloth also extended two to five cm diagonally into the soil. Most large
woody debris was removed from the plots to allow for the installation
of predation covers.
The germination of seeds was monitored once monthly throughout
the summer of 1999, and throughout the next growing season as
well, after the seeds were replenished in the spring. The germination
outcomes were tallied and analyzed by logistic regression. In
general, germination rates were very low across the entire
study, and there were no demonstrable density or exclusion cover
effects. However, we recorded much evidence of insect damage and
had to conclude that insect predation is a more significant
source of seed loss than we had anticipated. Research Branch contact: Catherine Bealle Statland.
|
