Long-term species responses of subtropical seasonal wetlands to grazing abandonment and prescribed fire

Background/Question/Methods

Wetlands are important but threatened ecosystems worldwide. In central Florida, many wetlands were affected by the conversion of surrounding landscapes to pastures. The drainage of the landscape and ensuing farming practices such as grazing, prescribed fire, fertilization, liming, and planting of non-native species had important consequences on embedded natural wetland plant communities. Here, we aimed to understand the long-term impacts of cattle exclusion and prescribed fires on wetland plant communities in two management contexts (intensively managed pastures = IMP vs. semi-natural pastures = SNP) with a focus on plant diversity, species turnover and community changes through time. The 14-year experiment consists of 40 seasonal depressional located at Archbold Biological Station’s Buck Island Ranch (Florida). Twenty wetlands were embedded in IMP and twenty were in SNP. In each pasture-type, we fenced 10 wetlands (2007) and burned 10 wetlands (every 2-3 years starting in 2008). We obtained a fully factorial design (pasture-type X grazing X fire) with each combination of treatments replicated 5 times. Vegetation was surveyed each year from 2006 to 2016 and every other year after 2016. We analyzed shifts in species diversity, evenness, and floristic quality in response to treatments using GLMM. We also analyzed species turnover, dominance shift through time, and specific species responses to treatments.

Results/Conclusions

We observed 215 plant species throughout the duration of the experiment. Species richness was on average 29.46 species per wetland. Wetlands in SNP were characterized by higher diversity of native species (~32.77 vs. 25.99), lower diversity and abundance of non-native species, higher evenness, and higher floristic quality. Removing cattle decreased native species richness (~31.01 vs. 27.75), while prescribed fires increased evenness, except in fenced wetlands in IMP where these treatments promoted highly clonal grasses (e.g. Panicum hemitomon, Hemarthria altissima, Hymenachne amplexicaulis). Despite cattle exclusion having little effects on diversity, it strongly affected the composition of wetlands. We observed that the rate of compositional change through time was higher in fenced wetlands, but did not change in response to fire and pasture-type. For example, IMP wetlands were dominated by Juncus effusus, but they became dominated by clonal grasses following fencing. Our results show the importance of preventing conversion of semi-natural pastures to intensively managed pastures for maintaining high diversity in wetlands. Despite 14 years of cattle exclusion, SNP and IMP wetlands remained different in composition, emphasizing the need for other interventions to restore wetlands in IMP to reference conditions.