Published online by Cambridge University Press: 20 January 2017
A greater understanding of the factors that regulate weed seed return to and persistence in the soil seedbank is needed for the management of difficult-to-control herbicide-resistant weeds. Studies were conducted in Tifton, GA to (1) evaluate whether glyphosate resistance, burial depth, and burial duration affect the longevity of Palmer amaranth seeds and (2) estimate the potential postdispersal herbivory of seeds. Palmer amaranth seeds from glyphosate-resistant and glyphosate-susceptible populations were buried in nylon bags at four depths ranging from 1 to 40 cm for intervals ranging between 0 and 36 mo, after which the bags were exhumed and seeds evaluated for viability. There were no detectable differences in seed viability between glyphosate-resistant and glyphosate-susceptible Palmer amaranth seeds, but there was a significant burial time by burial depth interaction. Palmer amaranth seed viability for each of the burial depths declined over time and was described by exponential decay regression models. Seed viability at the initiation of the study was ≥ 96%; after 6 mo of burial, viability declined to 65 to 78%. As burial depth increased, so did Palmer amaranth seed viability. By 36 mo, seed viability ranged from 9% (1-cm depth) to 22% (40-cm depth). To evaluate potential herbivory, seed traps with three levels of exclusion were constructed: (1) no exclusion, (2) rodent exclusion, and (3) rodent and large arthropod exclusion. Each seed trap contained 100 Palmer amaranth seeds and were deployed for 7 d at irregular intervals throughout the year, totaling 27 sample times. There were seasonal differences in seed recovery and differences among type of seed trap exclusion, but no interactions. Seed recovery was lower in the summer and early autumn and higher in the late winter and early spring, which may reflect the seasonal fluctuations in herbivore populations or the availability of other food sources. Seed recovery was greatest (44%) from the most restrictive traps, which only allowed access by small arthropods, such as fire ants. Traps that excluded rodents, but allowed access by small and large arthropods, had 34% seed recovery. In the nonexclusion traps, only 25% of seed were recovered, with evidence of rodent activity around these traps. Despite the physically small seed size, Palmer amaranth is targeted for removal from seed traps by seed herbivores, which could signify a reduction in the overall seed density. To be successful, Palmer amaranth management programs will need to reduce soil seedbank population densities. Future studies need to address factors that enhance the depletion of the soil seedbank and evaluate how these interact with other weed control practices.
Current address: Project Scientist, University of California, Davis, CA 95616.