Plant–soil relationships of the invasive annual grass Taeniatherum caput-medusae: a reciprocal transplant experiment

Abstract

The annual grass Taeniatherum caput-medusae (L.) Nevski (medusahead) is highly invasive on rangelands of the western United States. It originates from the Mediterranean basin and was introduced accidentally into North America in the late 1800s. To increase our knowledge of the soil factors that may contribute to its invasiveness, we conducted a reciprocal transplant experiment. Seed sources used were from an invasive population in northeastern California, USA and from a native population in southern France. Four soils were used: two from northeastern California, invaded (USI) and non-invaded (USN) by T. caput-medusae; two from southern France, with T. caput-medusae occupied (FR+) and a similar soil presently unoccupied (FR-) by T. caput-medusae. Treatments were control and autoclaved. Seeds were sown in containers (six replicates) for each soil and treatment combination of the reciprocal planting matrix and allowed to grow for 60 days. Following harvest, above-ground and below-ground tissue was dried and weight recorded separately. Above-ground tissue was analyzed for nutrient concentrations and various soil nutrient pools were quantified. Above-ground mass of T. caput-medusae was greatest in the USN soil and least in the FR- soil. Growth was affected by significant seed source x soil and treatment x soil interactions. French seed produced larger plants than did US seed when grown in the USN soil. For the USI soil, plant growth was significantly greater in the autoclaved soil than control. Availability of soil Fe, Mn, and ortho-P were significant predictors of plant growth as shown by stepwise regression. Moreover, the positive effect on growth of T. caput-medusae, due to autoclaving of USI soil, may be partially a consequence of elevated nutrient availability. Our data also suggests that US seed may have evolved a greater ability to uptake Mn from soil than French seed. Elevated soil nutrient availability in our western United States study area is a potential factor explaining the invasiveness of T. caput-medusae.