Journal of Vegetation Science 23 (2012) 483–494 Patterns of plant community assembly in invaded and non-invaded communities along a natural environmental gradient Riccardo Santoro, Tommaso Jucker, Marta Carboni & Alicia T.R. Acosta Keywords Alien plants; Carpobrotus aff. acinaciformis; Competition; Facilitation; Iceplant; Stress gradient hypothesis Nomenclature Celesti-Grapow et al. (2009) = for alien species; Conti et al. (2005) = for native species Received 2 August 2011 Accepted 27 October 2011 Co-ordinating Editor: Peter Adler Jucker, T. (corresponding author, tommasojucker@gmail.com), Santoro, R. (rsantoro@uniroma3.it), Carboni, M. (mcarboni@uniroma3.it) & Acosta, A.T.R. (acosta@uniroma3.it): Dipartimento di Biologia Ambientale, Universita ` degli Studi di Roma Tre, V.le Marconi, 446 – 00146, Roma, Italy Abstract Questions: Is the stress gradient hypothesis (SGH) effective in predicting patterns of community assembly in coastal dune plant communities along the sea–inland environmental gradient? Does the introduction of invasive plant spe- cies disrupt these patterns, leading to a collapse in community structure? Location: Sandy coastal dunes of the Lazio region (Central Italy). Methods: We randomly sampled coastal dune plant species in 2 m 9 2 m plots (4 m 2 ). Multivariate techniques were used to classify these plots and allowed identification of four plant communities along the sea–inland environmental gradient, three of which were invaded by Carpobrotus aff. acinaciformis (iceplant). For each community, we computed two different indices of co-occurrence that take into account species abundance, and then used Monte Carlo permutations alongside appropriate null models to determine whether overlap in the use of space was greater (aggregated community structure) or smaller (segregated community structure) than what would be expected by chance alone. Results: For the four non-invaded communities, the analysis highlighted how community assemblage patterns were strongly tied to the sea–inland environ- mental gradient. In the two foredune communities, overlap in the use of space was greater than expected by chance, indicating an aggregated assemblage. In contrast, progressing along the zonation, assemblage patterns first shifted to ran- dom in the transition dune community and then became segregated in the fixed-dune community. As for the three communities that were also found in an invaded state, the presence of iceplant was associated with a random struc- ture in community assemblage. Conclusions: Taken together, our results are consistent with the SGH, which predicts that along an environmental gradient, facilitation drives the assembly of plant communities where conditions are harsh, whereas competition is predom- inant where conditions are less severe. The comparison of the three invaded communities with their non-invaded counterparts revealed how iceplant inva- sion has led to a shift to randomness in community structure, as has been docu- mented for wildfires and other disturbances. Introduction Invasive species are considered among the leading culprits of the ongoing biodiversity crisis (Sala et al. 2000; Levine et al. 2003), and in recent years this has contributed to a growing interest in the study of invasion biology (Lamb- don et al. 2008; Davis et al. 2011; Wardle et al. 2011). Research in this field of ecology has covered numerous aspects, such as the development of theoretical frame- works to understand why and where certain species become invasive (Rejma ´nek & Richardson 1996; van Kle- unen et al. 2010; Speek et al. 2011), documenting the impacts of invasives at a multitude of ecological scales (Vitousek et al. 1997; Pawson et al. 2010; Ellis et al. 2011; Holmquist et al. 2011; de Moura Queiro ´s et al. 2011), as well as the design and implementation of conservation Journal of Vegetation Science Doi: 10.1111/j.1654-1103.2011.01372.x © 2011 International Association for Vegetation Science 483