PS 16-155 - Effect of cattle exclusion and pasture management on seedbank composition in subtropical seasonal wetlands

Background/Question/Methods

Wetlands embedded in agroecosystems are a high priority for restoration efforts. However, wetland potential for recovery might be impeded by availability of propagules, whether through dispersal or presence in the seedbank. Here, we examined how cattle exclusion and pasture type affected soil seedbank in seasonal depressional wetlands commonly found in Florida grazinglands. We compared the aboveground vegetation and seedbank composition of fenced (since 2007) and grazed wetlands embedded within two different pasture types (highly managed pastures = HMP vs. semi-natural pastures = SNP). Wetlands within HMP were fertilized and experienced higher stocking density, whereas wetlands in SNP were never fertilized and experienced lower stocking density.

In October 2018, we surveyed the vegetation of 12 seasonal wetlands (3 grazed in HMP, 3 fenced in HMP, 3 grazed in SNP and 3 fenced in SNP) by recording species’ cover in 10 randomly located 1-m² plots. We also collected soils at each of these 10 locations and recorded species germinating from each of these soils until November 2019. We calculated species richness in both seedbank and aboveground vegetation and tested the effect of cattle exclusion and pasture type using linear models. Finally, we compared seedbank composition to aboveground vegetation composition using ordination.

Results/Conclusions

99 different species germinated during the length of the experiment, whereas we observed 109 species observed in these wetlands during vegetation surveys. We observed slightly higher species richness in the seedbank of wetlands within semi-natural pastures (SNP = 47.2 [43.3, 51] vs. HMP = 40.6 [36.8, 44.5]), but no effect of cattle exclusion. We observed similar patterns in the aboveground vegetation. However, species rich wetlands did not necessarily have higher seedbank diversity.

Overall, we did not observe a strong similarity between seedbank composition and aboveground vegetation. However, the similarity between seedbank composition and vegetation was higher in wetland within highly managed pastures, especially when grazed. This was due to dominant species (e.g., Juncus effusus, Persicaria punctata) in grazed HMP wetlands forming an important seedbank and dominant species in SNP wetlands not germinating from the seedbank. Surprisingly, some key wetland species (Panicum hemitomon, Rhynchospora inundata) seldom germinated compared to what would be expected based on their abundances in the vegetation.

Our study shows that grazing alters seedbank composition, likely influencing how wetland vegetation may respond to restoration efforts. Species missing from seedbanks may require active planting if dispersal pathways are limited.