Print PageKudzu (Pueraria lobata) is an aggressively invasive weed that covers over 30,000 sq km of the American South. In addition to ecological concerns, kudzu control is critical to prevent the vine from overtaking parks, historical sites, and utility infrastructure. In recent years, some municipal land managers in Tennessee and Georgia have begun using goats to control kudzu in an effort to decrease herbicides use. To determine the most efficient use of limited resources, we constructed a two-compartment model of kudzu growth, spread, and dispersal that separates photosynthetic above-ground biomass, which is vulnerable to grazing and frost, from starch-rich below-ground biomass, which is protected from grazing and frost but consumed during above-ground growth. Using this model, we found an optimal goat deployment tactic that for a given cost minimized kudzu biomass in a single patch within a single season. Then, using the growth model and within-season tactics, we simulated kudzu growth, spread, and dispersal on spatially heterogeneous landscapes over five years and compared easily communicated and feasibly implemented management strategies, such as prioritizing removal of small patches or prioritizing patches on favorable habitats, to the optimal strategy found using dynamic programming.
Results/Conclusions
Optimal within-season tatics minimize below-ground biomass at the end of the season, but two qualitatively different tactics exist with optimality dependent on the relative magnitudes of setup versus ongoing costs. Setup costs include fencing the property and transporting goats; ongoing costs include supplemental feed and goat herds. When setup costs are very high compared to the ongoing costs, it is optimal to keep a small number of goats continuously on the property. Otherwise, it is optimal to deploy a herd of goats in a few pulses equally spaced throughout the growing season. Pulsing is optimal for costs comparable to those incurred by Knoxville, TN. Preliminary results show that the feasible, easily communicable management strategies result in 20% to 250% more biomass at the end of five years than the optimal strategy for each landscape. Strategies that exploited heterogeneity in the landscape performed best on landscapes with sharp boundaries. No single management strategy performed best under all spatial arrangements, and two strategies performed both the best on one landscape and the worst on another, suggesting that mixed strategies may reduce variability at a cost to situational performance.
