Overestimated importance of plant-soil feedbacks for Janzen-Connell
effects in natural grasslands: Evidence from the field
Abstract
Negative plant-soil feedbacks can be viewed as Janzen-Connell effects
and influence plant population dynamics in grasslands. However, even
though plant-soil feedbacks are often referred to as a mechanism for
Janzen-Connell effects, for grassland species this is based on pot
experiments and these effects have rarely been examined in the field. We
examined the spatial distribution of a monocarpic perennial Jacobaea
vulgaris to infer whether there is a distance- and/or density dependent
effect and if the pattern is soil-mediated. Replicated plots were
constructed to investigate J. vulgaris populations at two sites.
Rosettes and flowering plants were marked, and their coordinates were
recorded within each plot. For three plots, plants were tracked
repeatedly during a single season to examine temporal distribution
patterns. We then examined distance- and density dependent effects with
spatial point pattern analysis. We also collected soil underneath
flowering plants and 0.5-meter away from each plant. Seed germination,
survival and growth of seedlings were traced in these soils with
bioassays. Further, we measured biomass of J. vulgaris grown in soil
from patches with high densities of J. vulgaris and in soil from outside
of these patches. The density of rosettes was generally lower than
expected from null models at close distances from flowering plants. The
degree of clustering decreased from rosettes to flowering plants
indicating density dependent self-thinning. Both the distance-based
decay in rosette density and life-stage-dependent spacing became
stronger over time in the plots where repeated measures were taken. Seed
germination was higher in soil further away than in soil underneath
flowering plants. However, seedling mortality and biomass did not differ
in soils from different distances, and plants produced similar biomass
in soil from pairwise patches. Our study provides spatial-based evidence
for Janzen-Connell effects of J. vulgaris, and suggests plant-soil
feedbacks play a minor role in mediating Janzen-Connell effects.