Many ecosystem services provided by wetlands decline if they are managed for agricultural use. Ecological restoration of retired agricultural lands can restore these ecosystem services, yet practitioners require information on where restoration is possible and most likely to succeed. We report trends in the Massachusetts cranberry industry which suggest that cranberry farms located in riparian fens are well suited for ecological restoration that enhance their characteristics and functions as wetlands. We created a classification scheme for cranberry farms based on their: (1) crop status; (2) renovation status; (3) cultivar type; and (4) hydrologic type for the Wareham River watershed in southeastern Massachusetts. We ranked farms for their priority for restoration and extrapolated our results to the total cranberry acreage of Massachusetts. The occurrence of low-yielding native cranberry cultivars in all riparian farms (i.e. those with a direct hydrological connection to an adjacent river or stream), combined with our finding that 100% of the area of the highest-yielding new cultivars were planted in newly renovated but non-riparian farms suggest that riparian farms are not targets for investment but instead have a high likelihood of retirement. We found that 20% of farm area in this watershed had riparian hydrology, a proportion suggesting the existence of over 1,000 ha of high-priority farms for restoration statewide. Restoration of these stream-adjacent riparian farms can provide an exit strategy for some cranberry growers, ease economic pressures on remaining growers, and develop wetlands able to provide ecosystem services such as habitat provision, nitrogen removal, and recreation. Implications for Practice• We propose an aerial imagery-based method for the classification of wetlands cultivated for cranberries based on their: (1) crop status;(2) renovation status; (3) cultivar type; and (4) hydrologic type. • Trends in farm renovations and cultivar distribution suggest that the Massachusetts cranberry farm acreage can be partitioned into three ranks based on their likelihood for retirement and restoration, with the highest priority sites being traditionally constructed farms located adjacent to rivers that are planted with low-yielding native cultivars. • Riparian farms represent the most attractive sites for retirement, permanent land protection, and ecological restoration because of their lower yields, challenging management, and potential to offer ecosystem services such as habitat provision, nitrogen storage and removal, and recreation.

Retirement of cultivated croplands creates potential for ecosystem and wetland restoration, but vegetation and soil legacies of cropping influence the development of post-agriculture vegetation. In low-lying coastal watersheds of southeastern Massachusetts, cranberries (Vaccinium macrocarpon Aiton) are cultivated in commercial farm units that are diked, leveled and sanded beds created from historic wetlands. Current low cranberry prices and expanding cranberry production elsewhere now increase the likelihood of cranberry farmland retirement. We quantified the trajectories of plant species richness and cover, and plant characteristics (life form, native or non-native, wetland or non-wetland) in a chronosequence of cranberry farm beds that ranged from cultivated units to those retired from cropping and revegetated for 90 years with no post-cropping management. Species richness increased from active beds to 10–20 year-old beds and subsequently decreased. Post-retirement species richness was overwhelmingly dominated by native species. Shrub and tree richness and cover increased steadily over time. The richness of wetland, upland and facultative species all increased quickly after retirement and then declined in the oldest retired beds. The basal area and canopy cover of red maple (Acer rubrum) and pitch pine (Pinus rigida) increased over time. Vegetation followed a relatively consistent successional trajectory and the plant community after five to nine decades was predominantly forest and dominated by non-wetland plants. Encouragement of long-term persistence of greater diversity and cover of wetland plant species on retired cranberry farms will likely require active hydrological and soil modifications that decrease sand depth and raise water tables.

Nitrogen (N) pollution is arguably the single greatest threat to coastal water quality in the United States. In the state of Massachusetts, sustainable management of cranberry agriculture requires detailed understanding of potential sources and losses of N, which can contribute to impaired coastal waters. Given the complexity of N transport processes, models are often used to quantify the amount of N delivered from cranberry farms to coastal waters. However, ambiguities in model parameters, inconsistencies in deriving N loading rates, and inaccuracies in spatial coverages leave open questions about the contribution of cranberry agriculture to impaired coastal waters. In this study, we synthesize the methods, assumptions, and results of all N budget studies conducted in cranberry farms from 1995 to 2019. We propose a new analytical framework for measuring N fertilizer export from cranberry farms which includes our best understanding of N retention and transport in cranberry farms. We then apply our results to two watershed land-use models (Massachusetts Estuaries Project Linked Watershed-Embayment Model, LM; the Waquoit Bay Nitrogen Loading Model, NLM) to quantify N loads contributed by cranberry farms to the highly eutrophic Wareham River estuary. We find that the LM and NLM are in close agreement with respect to the amount of N delivered from cranberry farms to the Wareham River estuary. In our application of the LM, we estimated that 7791 kg NÁyr −1 from cranberry farms, or 15% of the total N load, reached the estuary. By comparison, we found that the NLM estimated slightly higher N loads contributed by cranberry farms to the estuary: 8365 kg NÁyr −1 , or 16% of the total N load. However, the models gave contrasting estimates of unattenuated N loads from cranberry farms (9670 kg NÁyr −1 , LM; 12,870 kg NÁyr −1 , NLM) and the amount of N transmitted through the watershed to the estuary (81% LM; 65% NLM).