Abstract:The fern Phyllitis scolopendrium (Aspleniaceae) is listed as a secondary protected plant in China. The species is extremely rare, and its distribution is limited to Changbai Mountain along the border with North Korea. Nearly nothing is known about its ecology, except for scarce studies on its artificial culture. The results of our previous field survey show that the natural population of P. scolopendrium has an aggregated distribution within a very limited area, and only few individuals are found in the Linjiang area of Jilin Province. A total of 70 plant species were found to occur near P. scolopendrium, including 16 species of trees, 16 species of shrubs, and 38 species of herbs. In the present study, we examined the inter-specific associations of this fern with other plant species and their relationship with community stability. On the basis of a 2 × 2 contingency table, we used a set of different methods, namely, X2 and W tests, association coefficient, point correlation coefficient, Ochiai index, and percentage co-occurrence, to analyze the inter-specific associations between P. scolopendrium and co-occurring species. In addition, we tested the level of community stability at elevations of 729 m and 1008 m by using the Godron stability analysis. The results showed that the total community association of variance ratio was greater than 1; the test statistic W was lower than X0.05 (N)2, indicating that the species in the overall community were positively associated (although this association was not significant). An analysis of association indices revealed that P. scolopendrium was less strongly associated with trees and shrubs than with herbs. When species pairs were analyzed, the association of P. scolopendrium with trees was not significant (X2 < 3.841), its association with the shrub Euonymus maximowiczianus was significant (X2 > 3.841), and all other associations analyzed were not significant. For herbs, the association of P. scolopendrium with Polystichum tripteron was strongly and significantly positively correlated (X2 > 6.635), that with Dryopteris crassirhizoma was significant, and the association with Lamium barbatum was significantly negatively correlated (3.841 < X2 < 6.635). These results indicated that there were no significant inter-specific associations between P. scolopendrium and most co-occurring plant species. With the exceptions of a few shrubs and herbs, inter-specific correlations tended to be independent and random. Godron stability analysis showed that the ratios of cumulative inverse of species number to cumulated relative frequency of two plant communities at 1008 m and 729 m of elevation were 35/65 and 34/66, respectively, distinctly different from 20/80. The ratios of the arbor layer, shrub layer, and herb layer in each community were also far from the stability point, indicating that the whole community was unstable. The stability ranking within each community was herb layer > shrub layer > tree layer. These results indicated that the P. scolopendrium community is not stable, and stronger protective measures are required. For species protection and ecological recovery, species with strong positive associations should be chosen for protection in order to establish a suitable environment for P. scolopendrium, stabilize its population, and effectively protect this rare species.