|Harrison, S. Kent|
Submitted to: Agriculture, Ecosystems and Environment
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/4/2010
Publication Date: 9/1/2010
Citation: Schutte, B.J., Liu, J., Davis, A.S., Harrison, S., Regnier, E.E. 2010. Environmental Factors That Influence a Mutualism Between the Earthworm Lumbricus terrestris L. and the Annual Weed Ambrosia trifida L. Agriculture, Ecosystems and Environment. 138:197-205. Interpretive Summary: Ambrosia trifida (giant ragweed) is one of the most competitive and problematic summer annual weeds in corn and soybean production in the eastern U.S. Corn Belt. Because A. trifida infestations produce few viable seeds, and because A. trifida seeds are short-lived in the soil seedbank, agricultural practices that focus on the destruction of seeds by agroecosystem seed predators are promising low-input control tactics. The efficacy of seed predator-based control tactics is reduced by the seed caching activities of the earthworm Lumbricus terrestris. Thus, to successfully develop and deploy seed predator-based control tactics for A. trifida, we must understand factors that influence L. terrestris seed caching. In this investigation, we utilized naturally occurring variation in L. terrestris caching of A. trifida seeds to identify environmental factors that influence this interspecific association. Across 2007, 2008 and 2009, both L. terrestris and A. trifida were present in all no-till agricultural fields surveyed across the eastern U.S. Corn Belt (Illinois, Indiana and Ohio); however, L. terrestris caching of A. trifida seeds occurred in only 53% of surveyed fields (n = 30). Whether the association between L. terrestris and A. trifida (hereafter abbreviated "LtAt association") was present or absent was not a function of soil physical characteristics (percent clay, percent sand, percent silt, pH, bulk density, cation exchange capacity, taxonomic class). To the contrary, variability in LtAt association presence corresponded with differences in overwinter climate. Site-years characterized by warmer and wetter winters were more likely to exhibit the LtAt association compared to site-years characterized by colder and drier winters. This indicates that A. trifida seeds can be buried by L. terrestris in no-till agricultural fields throughout the eastern U.S. Corn Belt, provided the overwinter climate is conducive to L. terrestris activity on the soil surface. Because climate cannot be controlled, seed predator-based weed control tactics for the eastern U.S. Corn Belt will need to account for the antagonistic effects of L. terrestris-mediated seed burial. This points to a need for multitactic approaches for weed seed destruction, that is, combinations of pre-dispersal predation, post-dispersal predation, and microbial degradation.
Technical Abstract: The earthworm Lumbricus terrestris L. can improve Ambrosia trifida L. seed survival and seedling recruitment in agroecosystems with high risks of post-dispersal seed predation. In a previous 1-yr survey of no-till agricultural fields in the eastern U.S. Corn Belt, both L. terrestris and A. trifida were universally present but the interspecific association (abbreviated “LtAt association”) was present only in the east region. Here, we expand the previous survey to 3 years to: 1) determine if regional trends in LtAt association presence among no-till agricultural fields in the eastern U.S. Corn Belt are consistent across years, and 2) to conduct exploratory analyses to identify environmental factors that contribute to LtAt association presence in these fields. Field survey data were collected from 30 sites during 2007, 2008 and 2009. Following field data collection, geographic coordinates were used to gather soil and climate data from internet resources. Regional trends in LtAt association presence were not consistent across years, and the annual percent of fields that exhibited the LtAt association ranged from 50 to 100% in the east, 0 to 100% in the west. LtAt association presence was not influenced by soil-related factors but was affected by climate variables. Classification tree models indicated that LtAt association presence was primarily influenced by the frequency of nonfreezing days (days in which minimum air temperate > 0 °C) from 1 October to 31 March. Site-years with 57 or more nonfreezing days were more likely to exhibit the LtAt association compared to site-years with less than 57 nonfreezing days (model prediction accuracy rate = 87.5%, n = 30). These results indicate that L. terrestris can affect A. trifida seedling recruitment wherever these species co-occur, provided the overwinter climate is conducive to L. terrestris soil surface activity.