Quantifying the Hydrological Niche of Upland Swamp Plant Communities Using Indicator Species

ABSTRACT Coastal Upland Swamps of the Sydney Basin Bioregion, Australia, are highly biodiverse upland mires. Swamp vegetation organises into five communities, thought to occupy distinct positions along a hydrological gradient. Longwall mining reduces the duration of root zone saturation, but lack of quantitative understanding of hydrological niche dimensions limits our ability to predict the impacts of drawdown in terms of observed vegetation communities. Here, we use indicator species to quantify the hydrological niche of their corresponding plant communities and predict the impacts of mining‐related drawdown. We first modelled the relative occurrence frequency of 20 vascular plant indicator species (four per community) from 11 unmined sites across four swamps and then used this model to predict their frequencies at a mine‐impacted swamp. We quantified the hydrological niche for each community using the average number of days per year of root zone saturation (≤ 30 cm below surface) from time series soil moisture data. Inferred community hydrological niches were broadly consistent with qualitative ecological understanding, with estimated optimal mean days per year saturated ranging from Restioid heath at 17 [≤ 16, 49] (mean ± [95% uncertainty]) days to Ti‐tree thicket at 348 [315, ≥ 353] days, with Banksia thicket 45 [16, 111], Sedgeland 187 [139, 264] and Cyperoid heath 243 [198, 290] communities intermediate. At the impacted Swamp, the model underpredicted wetter‐ and overpredicted drier‐adapted community indicator species, suggesting vegetation has not yet reached hydrological equilibrium. Results suggest potential for using the modelled hydrological niche of wetland plant communities to predict community‐level impacts of hydrological change in upland swamps.