We present a study on the understory vegetation of neighbouring black locust and native tree woodlands, in pairs, across a vast northern Mediterranean lowland region under agricultural land‐use abandonment. We use RLQ analysis to assess co‐correlations between topographical, soil, stand, and land cover variables (R table) and species trait attributes (Q table), constrained by the relative abundances of understory vegetation species (L table). We also perform a partial RLQ to separate the effects of canopy dominance (native vs. alien) on understory trait distributions from those representing environmental conditions. The ordination diagram of the basic RLQ showed the overwhelming influence of canopy dominance, since nearly all of the black locust stands clustered at the top half of the plot (Fig. 1a). In the partial RLQ, where the effect of canopy dominance was removed, this clustering disappeared (Fig. 1b), the overall significance was maintained, but either the significance nor the contribution to total inertia of some variables were lost, namely the C:N ratio and available P in the soil, but also phenols which had a negligible contribution to total inertia in the partial RLQ (Fig. 1c).Our study shows that plant trait variability in the understory of alien vs. native tree canopies is mainly driven by three processes: soil N mineralization, light availability and edge disturbance intensity. The process of soil N mineralization must mostly be driven by the dominance of black locust in the canopy, given that this tree species contributes to an increase in the C:N ratio and available P and a decrease in phenols in the soil. Light availability and edge disturbance intensity were also associated, to some extent, to the canopy dominance by black locust vs. native trees, but less strictly, as is shown by the partial RLQ results. Perspectives for further research and management implications are discussed.
Combined effects on plant trait composition and soil properties by an alien tree species: perspectives from a black locust case study
Tommaso Sitzia;Thomas Campagnaro;S. Nardi;A. Ertani
2019
Abstract
We present a study on the understory vegetation of neighbouring black locust and native tree woodlands, in pairs, across a vast northern Mediterranean lowland region under agricultural land‐use abandonment. We use RLQ analysis to assess co‐correlations between topographical, soil, stand, and land cover variables (R table) and species trait attributes (Q table), constrained by the relative abundances of understory vegetation species (L table). We also perform a partial RLQ to separate the effects of canopy dominance (native vs. alien) on understory trait distributions from those representing environmental conditions. The ordination diagram of the basic RLQ showed the overwhelming influence of canopy dominance, since nearly all of the black locust stands clustered at the top half of the plot (Fig. 1a). In the partial RLQ, where the effect of canopy dominance was removed, this clustering disappeared (Fig. 1b), the overall significance was maintained, but either the significance nor the contribution to total inertia of some variables were lost, namely the C:N ratio and available P in the soil, but also phenols which had a negligible contribution to total inertia in the partial RLQ (Fig. 1c).Our study shows that plant trait variability in the understory of alien vs. native tree canopies is mainly driven by three processes: soil N mineralization, light availability and edge disturbance intensity. The process of soil N mineralization must mostly be driven by the dominance of black locust in the canopy, given that this tree species contributes to an increase in the C:N ratio and available P and a decrease in phenols in the soil. Light availability and edge disturbance intensity were also associated, to some extent, to the canopy dominance by black locust vs. native trees, but less strictly, as is shown by the partial RLQ results. Perspectives for further research and management implications are discussed.
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
