Along rivers in the Southwestern U.S., non-native tamarisk (Tamarix spp.) has gradually replaced native cottonwood (Populus spp.) and willow (Salix. spp.). Our research aimed to develop a mechanistic and predictive understanding of cottonwood and tamarisk seedling damage caused by flooding, in an effort to understand ecogeomorphic feedbacks, differential responses among species, and implications for riparian restoration. We tested the effect of controlled flood events on seedlings in an outdoor experimental meandering stream channel. To simulate the interflood colonization of low-lying bar elevations, seedlings were installed in a 1.5-m-wide sand bar within the stream during low flow conditions. Stream discharge was then increased to a constant flood level for approximately 8 hours, after which seedling response was assessed. There were 8 experimental trials with different combinations of species, density, and sediment feed rate. We hypothesized that plant damage (burial, scour, or permanent pronation) would be influenced by individual plants’ height, species-specific morphology, vegetation community structure (density and species configurations), hydraulic forces (as a function of location on the bar), sediment transport rate, and the interactions between those parameters. We analyzed these processes within a logistic regression framework that predicted the probability of damage given those explanatory variables.
Results/Conclusions
We found that tamarisk had a higher probability of damage regardless of height, density, and community configuration (monospecific or mixed). In dense patches, the odds of damage for tamarisk were 5.3 times higher than for cottonwood, and 8.8 times higher than cottonwood in monospecific patches. However, for every centimeter increase in plant height, the odds of damage for tamarisk decreased by 8%, but decreased only 4% for cottonwood, suggesting that tamarisk develop flood resilience quickly as they grow taller. Additionally, with every meter closer to the thalweg, the odds of plant damage for both species increased by a factor of 6.3 with high sediment feed and a factor of 48.4 with no sediment feed. The results suggest that both environmental variables and plant morphological traits are important to consider when designing flow-based restoration efforts on regulated rivers. We expect the highest probability of damage for tamarisk to be on sandbars with high densities of young seedlings. In order to maximize the establishment of native species such as cottonwood, floods should be released in the early growing season when tamarisk are still short and most vulnerable.