Search

Janzen-Connell effects state that the accumulation of host-specific natural enemies near parent plants can negatively affect their offspring. Negative plant-soil feedbacks can produce patterns of seedling performance predicted by Janzen-Connell effects and influence plant populations, but their relevance in field conditions remains unclear. Here, using spatial point-pattern analysis, we examine the spatial distribution of Jacobaea vulgaris to assess whether distance- and density-dependent predictions of Janzen-Connell effects are evident in the field. We established 27 replicated 64 m2 plots at two grassland sites and mapped positions of rosette-bearing and flowering J. vulgaris plants within each plot. To investigate temporal distribution patterns, we tracked plant positions repeatedly in three plots during a single season. Additionally, we tested whether these patterns are soil-mediated. Soil samples were collected underneath flowering plants and at a distance of 0.5 meter, and used to compare seed germination, seedling survival, and growth under controlled conditions. Furthermore, we measured J. vulgaris growth in soil from patches with high J. vulgaris densities and in soil from areas outside these patches. The density of rosette-bearing plants was lower at close distances from flowering plants than expected from null models, suggesting negative distance-dependent plant recruitment. The degree of clustering decreased over time from rosette-bearing to flowering plants, indicating density-dependent self-thinning. Seed germination was higher in soil further away from flowering J. vulgaris plants than in soil underneath plants at one site, but soil distance was not an overall significant factor in explaining seed germination. However, seedling mortality and biomass did not differ between soils collected at the two distances, and plants produced similar biomass in soil collected from inside and outside J. vulgaris patches. Our study demonstrates conspecific distance- and density-dependent plant recruitment in J. vulgaris in the field, but we found no evidence this depends on belowground natural enemies.

Current and legacy effects can greatly affect the growth of a focal plant and its interactions with herbivores and such effects can be mediated by above- and belowground effects. However, determining the relative importance of current and legacy above- and belowground effects in natural conditions is a major challenge. In a long-term grassland experiment, we examined the relative importance of the current and legacy above- and belowground effects of plant communities on the growth and aboveground herbivore damage on a focal plant, Leucanthemum vulgare. Focal plants were planted into tubes with soil collected from different plant communities and placed back into the plant communities. Weekly, plant growth and damage were recorded and after 12 weeks plant biomass was measured. We analyzed how well aboveground and belowground characteristics of the current and legacy plots explained plant growth and herbivory. We found both current plant communities and legacy plant communities significantly affected plant growth (shoot biomass and the number of leaves) and herbivory. Root biomass of the focal plants was influenced by current plant communities only. Current and legacy above- and belowground characteristics explained 12% and 11% of the variation in shoot biomass. Root biomass was mainly explained by current above- and belowground characteristics with a total explained variation of 10%, while legacy effects explained 3%. Legacy effects explained most variation in the number of leaves during the first two weeks of measurements, and the effect remained present during the growth season. In contrast, characteristics of the current community explained most of the variation in herbivory throughout the growth period, with on average 6% explained variance aboveground vs. 5% belowground. Our grassland field study highlights that both current and legacy effects influence plant growth, but herbivory on focal plants is caused by current neighborhood effects only and not by legacy effects.