Tradeoffs between production and perennial vegetation in dairy farming systems vary among counties in the northeastern U.S.

Dairy farms that grow more perennial vegetation as grazing pastures or conserved forages can offer many environmental benefits but may show reduced milk production relative to farms feeding higher amounts of grain and corn silage. Because yields of annual and perennial crops vary with soil type, an accurate comparison of the productive potential of these systems over county or regional scales may require taking into account spatial variation in soil quality. In this study, we present a novel approach to calculate the production from dairy systems that adjusts average crop yields to the productive potential of local soils using the National Commodity Crop Productivity Index (NCCPI). We used on-farm survey data to define confinement and grazing systems with varying amounts of perennial forage and applied our method to a sample of five counties in the northeast United States. High corn silage farm systems produced 21 to 168% more milk per hectare of farmland than grazing-based farm systems, but variation among counties was greater than variation among systems, with the best (Lancaster, PA) producing as much as 5.3 times more than the least (Orange, VT). Adjusting yields for soil productivity had smaller effects on milk production than differences in farm system or county. On average, grazing farm systems generally produced slightly more milk when yields were adjusted using the NCCPI (8%) while high corn silage systems showed a moderate decrease (13%). Compared to scenarios of all local crop production, scenarios with unlimited corn and soybean imports often more than doubled county-scale milk production. Restricting grain imports to prevent excess phosphorus resulted in a 3-15% decrease in milk production relative to unlimited imports, but still produced far more milk than in the all local production scenarios. Sensitivity analysis of the model showed that milk production in each county was very responsive to changes in perennial forage yields (especially for grazing systems), responsive to changes in average daily milk production per cow, and generally not responsive to changes in the productive lifetime of lactating cows. This study demonstrates a persistent tradeoff between perenniality and production in dairy systems, but suggests that opportunities may exist to maintain current milk production levels in the Northeast while also expanding land cover in perennial vegetation.