Data from: Beyond plant-soil feedbacks: mechanisms driving plant community shifts due to land-use legacies in post-agricultural forests

Although biotic legacies of past agricultural practices are widespread and increasing in contemporary ecosystems, our understanding of the mechanisms driving such legacies is still poor. Forest understories on former agricultural land show low frequencies and abundance of typical woodland species when compared with ancient forests. These community shifts have been ascribed to the effects of dispersal limitation. A rarely considered mechanism is that post-dispersal processes driven by plant-associated communities determine the poor performance and recruitment of woodland indicators. Given the strong alterations in soil conditions due to former agricultural practices, we hypothesized that (abiotic) plant–soil feedbacks could be a major factor in community shifts. We addressed this hypothesis by comparing plant-associated communities in the soil and above the ground in ancient and post-agricultural alluvial forests; then, we experimentally tested whether the changes in biotic and abiotic soil properties could affect above-ground herbivore abundance and pressure and plant performance. Ancient and post-agricultural communities clearly differed in composition at different levels of the food web. Besides the plant community, we also observed the differences in the microbial and nematode community with increased abundance of root-feeding nematodes in post-agricultural soils. The composition of the above-ground invertebrate community did not differ in ancient and post-agricultural forest parcels; however, plants growing in post-agricultural sites showed higher abundance of invertebrate herbivores and suffered more herbivory. Nutrient analyses of soil and plants showed that increased levels of phosphorus (and to a lesser extent, nitrogen) made plants more nutritious for insect herbivores. Laboratory experiments further pointed to this mechanism as an explanation of the poorer performance of woodland indicators in post-agricultural woodlands. Our results point to biotic and abiotic plant–soil feedbacks coupled with herbivory as a new mechanism to explain the legacy effects in temperate forests. The modification of the below-ground community and soil abiotic characteristics by previous agricultural activity affects not only the plant growth but also the plant nutrient content in the compared understorey species, making them more susceptible to above-ground herbivory. Our results provide one of the first examples of integrating plant–soil feedback and above- and below-ground interactions to explain land-use legacies. Data_Functional_Ecology_All_data_de_la_Pena_et_al_2016Each sheet contains a different data set: Sheet 1: Abiotic fators Data for abiotic soil factors i.e. pH-KCl, total nitrogen, percentage of ash rest, K, Mg, Ca, Al, P, Olsen-P, N/P ratio, percentage of organic matter OM, percentage of carbon, C/N ratio of soil samples collected at three forest sites in Flanders (Belgium): Aelmoeseneie, Doode Bemde and Muizen forest. In all sites, we compared soil conditions in ancient and postagricultural forest parcels. Sheet 2: Nematode community Nematode composition in soil samples collected in three different forest sites in Flanders (Belgium); at each site ancient parcels and post-agricultural parcels were sampled. Sheet 3: Data info of characteristics of species in the introduction experiment Data on basic plant traits of plants reintroduced in post-agricultural and ancient parcels in the Muizen forest (Belgium). For each plant we meassured: length in cm, number of stems, number of holes in leaves, number of leaves, number of leaves with signs of herbivory, proportion of leaves showing herbivory marks, herbivory index. Sheet 4: Invertebrate abundance on surveyed plants along transects in the Muizen forest, Belgium Invertebrate abundance was assessed for all plants present along 10m x 1m transects in the Muizen forest in Belgium. We compared invertebrate abundance in post-agricultural forest parcels and ancient parcels. Sheet 5: Vegetation plots Data on vegetation plots. Plots 10 x 10m. For each forest site i.e. Muizen forest, Aelmoeseneie(ALM) and Doode Bemde six parcels were sampled; 3 on ancient forest parcels and 3 on postagricultural. For each plot the understory vegetation was recorded. Sheet 6: Vegetation transects Comparison of plant species in the understory along transects in the Muizen forest. 10 transect surveys were conducted in ancient forest plots and 10 in post-agricultural. Sheet 7: Ecoplates For the characterization of the soil microbial community we used a method that measures by spectrometric quantification the utilization by microbes of different carbon substrates in microtiter plates (EcoPlates®). Here we compare mean values for soil samples taken in post-agricultural and ancient forest sites.   Sheet 8: Data experiment Urtica dioica Data on plant traits and analysis of population build-up of Aphis urticata on Urtica dioica plants growing in soil collected in post-agricultural forest parcels or in ancient forest parcels. The experiment also compared the effect of soil sterilization and provenance on plant performance by measuring plant growth (biomass, no. of runners and flowering). Sheet 9: Plant nutrient analysis Data on nitrogen and phosphorus content of harvested plants from a re-introduction experiment in ancient and post-agricultural forest parcels (in the Muizen forest, Belgium). There were four species compared i.e. Geum urbanum, Circaea lutetiana, Primula elatior and Urtica dioica. Plants were weighed after drying to constant weight at 70 °C for 48 h. Sheet 10: Data Deschampsia cespitosa experiment Data on the effect of soil sterilization (sterilized vs. non- sterile) and provenance (i.e. ancient and post-agricultural) on plant growth of Deschampsia cespitosa. Sheet 12 and Sheet 13: Population build-up of aphids on Urtica dioica and Deschampsia cespitosa