Skip to main content
Dryad

Wildfire impacts on seedbank and vegetation dynamics in Calluna heath

Cite this dataset

Naszarkowski, Noemi et al. (2023). Wildfire impacts on seedbank and vegetation dynamics in Calluna heath [Dataset]. Dryad. https://doi.org/10.5061/dryad.fbg79cp0f

Abstract

Moorlands dominated by Calluna vulgaris (hereafter Calluna) are globally rare and under increasing threat of wildfires due to climate change. The soil seedbank is important for community resilience, but research on its contribution to regenerating vegetation after moorland fires has, to date, focused on prescribed fire or on the short-term (≤ 2 year) impacts of wildfire. To address the role of the seedbank in long-term vegetation regeneration, we studied a chronosequence of six wildfire sites within a Scottish moorland catchment, ranging from 2 to 64 years since burning. We recorded vegetation composition and sampled the seedbank. Calluna and Erica spp. comprised 66 % and 27 % of germinated seeds, respectively, and the majority of vegetation species were not represented in the seedbank. Canonical Correspondence Analysis (CCA) revealed that time since fire was a strong predictor of vegetation species composition, but it had no effect on seedbank composition, suggesting that the same species dominate the seedbank regardless of burning. Whilst total seed density was not affected by time since wildfire, the effect on the non-Calluna seedbank was more notable. The proportion of non-Calluna species decreased with time in both vegetation and seedbank as Calluna cover was re-established and the seedbank of other species became gradually depleted. Whilst relative non-Calluna cover in the vegetation was at a maximum in recently burnt areas and then declined with time since fire, the species richness and proportion of non-Calluna in the seedbank followed a unimodal pattern after fire and reached a peak after approximately 35 years. Our results contribute to the knowledge of how moorland ecosystems regenerate after severe disturbances, which are likely to become more common as the climate changes.

Methods

Raw seedbank and vegetation data originating from 1x1 m randomly placed quadrats in and around historical wildfire sites in Glen Tanar, Scotland. Vegetation species are given in percentage cover by visual estimate. Seedbank species are given in number of germinated seeds per sample (two seedbank samples were collected in different locations in each quadrat using a soil core of 6 cm diameter and 4 cm depth, resulting in a total soil volume of 226.2 cm3 collected from a total area of 56.5 cm2 in each quadrat.). We have also included species number, the number of total seeds and non-Calluna vulgaris seeds per square metre, and proportion of non-Calluna seeds. The seedbank dataset additionally includes environmental data of quadrats: elevation, aspect, slope, soil pH, soil bulk density, soil N, soil C, ratio of C to N, National Vegetation Classification (NVC) community and a broad habitat type (wet or dry moorland). Both datasets include years since fire and burning status, where B = burnt, U = unburnt.

Funding

Macaulay Development Trust