Soil seed pathogens: Initial and persistent inhibitors of plant naturalization in Pacific Northwest steppe and forest

Background/Question/Methods

The Enemy Release Hypothesis (ERH) posits that invasive species abundance and spread is attributable, in part, to the absence of biotic interactions such as parasitism.  Soil seed pathogens, in particular, can influence plant establishment and distribution but the extent to which these pathogens attack invasive species compared to their attack on naturalized and native species is little explored.  We asked 1) whether enemy release from soil seed pathogens results in greater survival for invasive cheatgrass (Bromus tectorum) seeds than the survival for naturalized cereal rye (Secale cereale) and native bluebunch wheatgrass (Pseudoroegneria spicata) seeds in eastern Washington steppe and 2) whether cheatgrass seed survival in the soil differs between invaded eastern Washington steppe and adjacent, uninvaded coniferous forest.  We predicted that 1) cheatgrass seed survival would be significantly greater than co-occurring naturalized or common native species and that 2) cheatgrass seed survival would be lower in forest soils than steppe soils.  We conducted seed survival assays in 2012 and 2013, using soil from three steppe and three forest habitat types.  Soil was either autoclave sterilized or left unaltered. Seeds of B. tectorum, S. cereale, or P. spicata were sown in pots, and seed survival was evaluated after six weeks.

Results/Conclusions

Introduced species (S. cereale, B. tectorum) had greater seed survival than native P. spicata (all P-values <0.001), but seed survival did not differ between introduced species (t = 0.96, P = 0.337).  These results suggest that whereas introduced species have greater survival than natives, enemy release from soil pathogens does not contribute strongly to invasive B. tectorum abundance compared to naturalized S. cereale in these steppe habitats.  Concurrent studies suggest that the high relative abundance of cheatgrass is more strongly influenced by other biotic interactions - strong avoidance by post-dispersal seed predators and strong seedling competitive ability compared to the performance of co-occurring naturalized species.  Cheatgrass seed survival was lower in steppe soils vs. forest soils (F_1,10 = 8.65, P = 0.015); lack of the predicted high seed mortality in forests soils suggests seed-soil interactions alone are not main factors for the low to nil occurrence of B. tectorum in regional forests. Soil seed pathogens appear to play a secondary role influencing the abundance and distribution of naturalized and invasive species in these habitats.