Forest ecosystems are commonly characterised by a hierarchy of resources. During a disturbance of a forest community, increased light availability in the understorey can support competitive interactions at the expense of facilitation. This may overwhelm the role of belowground resource heterogeneity in maintaining species coexistence and so result in biotic homogenisation of a site. We re-surveyed species composition while estimating interspecific interactions along a microtopographic (moisture) gradient of a temperate swamp forest after increasing the availability of light into the undergrowth. We measured temporal changes in species diversity, compositional dissimilarity measures and functional traits, while belowground resource heterogeneity remained unchanged over time. To explain observed changes in terms of prevailing interspecific interactions, we evaluated temporal changes in the competition-facilitation balance. The productivity of understorey vegetation increased strongly over the five years of the study. This was accompanied by an increase in species diversity and the maintenance of compositional dissimilarity. The relative importance of competitive and facilitative interactions along the microtopographic gradient did not change over time. Rather than an overall biotic homogenisation of a site, we observed a spread of resource acquisitive species within the resource-rich sites and of stress-tolerant species within the resource-limited sites. We have shown that the temporary increase in productivity did not mitigate the effects of belowground resource heterogeneity on plant species turnover, which is likely maintained by contrasting interspecific interactions prevalent in different parts of the resource gradient. This suggests contrasting community assembly processes maintain diversity and productivity even in a local community.

In July 2013, we established 212 microtopographic gradients in all tree hummocks in three stands of the Alnus swamp forest. There were 71, 71 and 70 hummocks in the three stands and the stands were at least 150 m apart. The coordinates of the geographical centre of each hummock were taken using high-accuracy measurements using a Leica GPS System 1200 (Leica Geosystems, Sankt Gallen, Switzerland). Within each microtopographic gradient, we divided a hummock into 2–6 vertical zones (Fig. 1a). The lowest zone of the microtopographic gradient (resource rich) captured the inundated hollow below the common shoreline (< 0 cm) and was defined as a 100-cm-diameter horizontal area surrounding the dry part of the hummock. The other zones, each of 20 cm height, started at the common shoreline of the dry part of the hummock such that resource availability decreased with increasing height above the common shoreline. Hummock height was defined as the sum of the vertical distance between the highest point of each hummock and the shoreline and the mean depth of the inundated area measured on four opposing sides. The hummock area was defined as a horizontal area within the boundaries of an inundated zone estimated using georeferenced aerial photographs taken for each hummock 2.5 m above the ground. Within each hummock we established 2–6 permanent plots each covering the whole surface area of each zone (i.e., zone area). In total we established 728 plots. In July 2013 and 2018, the percentage cover of all vascular plant species in each plot and whole microtopographic gradient was estimated visually by the same senior researchers.

DATAset_DoudaOikos.csv contains table with the following columns:

ID – identification number of each vegetation sample
HummockID - identification number of hummock (microtopographical gradient)
LevelID - identification number of hummock zones 
X_mean – longitude of hummocks in SJTSK projection
Y_mean – latitude of hummocks in SJTSK projection
HummockArea_m2 – vertical area of hummocks (m2)
HummockHeight_cm – the height of the terrestrial part of hummocks (cm)
WaterDepth_mean_cm – depth of water in the bottom zone of hummocks (cm) 
Year – sampling year
Other columns (AgrCan and the others) represent species and their percentage coverage in the hummock zone