Comparing the invasibility of experimental "reefs" with field observations of natural reefs and artificial structures.

Natural systems are increasingly being modified by the addition of artificial habitats which may facilitate invasion. Where invaders are able to disperse from artificial habitats, their impact may spread to surrounding natural communities and therefore it is important to investigate potential factors that reduce or enhance invasibility. We surveyed the distribution of non-indigenous and native invertebrates and algae between artificial habitats and natural reefs in a marine subtidal system. We also deployed sandstone plates as experimental 'reefs' and manipulated the orientation, starting assemblage and degree of shading. Invertebrates (non-indigenous and native) appeared to be responding to similar environmental factors (e.g. orientation) and occupied most space on artificial structures and to a lesser extent reef walls. Non-indigenous invertebrates are less successful than native invertebrates on horizontal reefs despite functional similarities. Manipulative experiments revealed that even when non-indigenous invertebrates invade vertical "reefs", they are unlikely to gain a foothold and never exceed covers of native invertebrates (regardless of space availability). Community ecology suggests that invertebrates will dominate reef walls and algae horizontal reefs due to functional differences, however our surveys revealed that native algae dominate both vertical and horizontal reefs in shallow estuarine systems. Few non-indigenous algae were sampled in the study, however where invasive algal species are present in a system, they may present a threat to reef communities. Our findings suggest that non-indigenous species are less successful at occupying space on reef compared to artificial structures, and manipulations of biotic and abiotic conditions (primarily orientation and to a lesser extent biotic resistance) on experimental "reefs" explained a large portion of this variation, however they could not fully explain the magnitude of differences.