Facilitation by nurse plants regulates community invasibility in harsh environments

Question The biotic resistance hypothesis states that species diversity is a barrier for plant biological invasions because alien species are more likely to find strong competitors as diversity of native plants increases. In stressful environments, however, these negative relationships could be diluted because increased physical harshness might reduce the importance of competition as a structuring force in plant communities. Nevertheless, if nurse plants facilitate other species and create more diverse vegetation patches than those found in their surroundings, the performance of invaders should be negatively related to diversity within these patches. Conversely, diversity would not regulate invasibility in the surroundings of nurses because these habitats are subjected to higher levels of environmental harshness. Location A high‐Andean landscape of central C hile, dominated by cushion plants that facilitate native species and field chickweed ( C erastium arvense ), an invasive exotic species. Methods To test our hypotheses we used observational data taken in two consecutive growing seasons by sampling plant assemblages within and outside cushion‐protected patches. Additionally, a field experiment with artificially assembled plant communities was conducted to assess why the abundance and performance of the invasive species was related to native plant diversity. Results Contrary to our expectations, observational data indicated that the abundance of the invader species was positively related to native plant diversity within cushion patches, while negative relationships were found in the surrounding habitat. Similarly, the field experiment indicated that the performance of C . arvense increased with plant diversity within cushion patches but, in this case, neutral relationships were found in the surrounding open areas. Conclusions Our findings indicate that nurse plants can modulate invasibility–diversity relationships in harsh environments. Overall, the results suggest that the biotic resistance hypothesis may fail to predict the direction of these relationships within nurse‐protected patches. Instead, these results suggest that the alternative biotic acceptance hypothesis could operate at small spatial scales.