JADE Student Edition 2019 JADE JSLUG 2019 | Page 72

10m sweeps were performed to capture larger trees (>5cm DBH) in the area, which contribute to the then current canopy as well as the canopy before disturbance. Canopy openness and R. ponticum presence were recorded to act as measurements of ecological disturbance. The frequency of R. ponticum, L. kaempferi stumps and canopy openness were compared against one another to form a statistically significant relationship between these proxy measurements and ground flora data. and no larch were felled, in comparison to those of Zone 2 (=0.517) and Zone 3 (=0.489) (see Fig.2) . However, the difference between the biodiversity indices of Zone 1 and Zones 2 and 3 (≤0.242) were not significant from one another (F(2,24)=1.21, N=9, P=0.317). An increased sample size of 45 separate sample sites across the study area would be required to ascertain an 80% chance of finding statistically reliable results according to a power analysis of this test. During data processing, the Shannon diversity index and species evenness for each sample site’s species abundances was calculated using the following formula: H=-i-1spi*ln⁡(pi). pi stands for proportion of vegetative coverage of a species within a quadrate, and s represents species richness. Forest Regrowth Statistics Anderson-Darling tests were performed on all result sets to verify their viability for reliable parametric tests such as the T-test and ANOVA. Kruskal- Wallis tests were used to find the differences in the medians of non-parametric datasets and Mood’s Median tests were only performed on non-parametric datasets whose distributions were not comparable with one another. Discussion Ground Flora Results Biodiversity In terms of species richness, Zone 1 had the greatest mean number of species per quadrat with 4.3, which was at least 1 species more than Zones 2 and 3 (S=3.3, S=3.1 respectively). However, as no more than 8 species were observed within a single sample site, more sophisticated methods of measuring biodiversity were required to discern whether these differences were ecologically significant. The mean Shannon Index of Biodiversity of ground flora was highest in Zone 1 (=0.759) where high densities of rhododendron bushes were removed, 72 When segregating the birch sapling frequency found at each sample site into canopy openness classes, a clear trend is seen (see Fig. 3). A clear peak in median birch sapling frequency (29.0) is visible in sites of 40-60% canopy openness with much lower frequencies in sites above and below this range (medians not surpassing 4.0). This difference is significant (H=11.85, DF=4, N=27, P=0.018), suggesting that areas of intermediate canopy disturbance are faster regenerators of light- demanding tree species, such as birch.  P. ramorum management appeared to have no effect on woodland ground flora biodiversity despite research showing that clear felling of trees can either augment species richness (Kirby et al., 2017) or severely diminish it (Jianwei & Chenqiu, 1999, Keith et al., 2009). Instead, this study either finds support in the theory of increased woodland homogeneity as stated by Staley et al. (2013) or accepts the unlikely possibility that management had little impact on ground flora diversity in the case that Keele Springpool Wood was ecologically homogenous before 2015. As all larch-specific niches were removed from the ecosystem, management almost certainly had some effect on woodland biodiversity suggesting that the theory of management-caused