This study assesses the impacts of whole-tree harvesting (WTH) versus stem-only harvest (SOH) on two oak-dominated stands located in Northern Wisconsin. Specifically, our study follows up on an experiment designed to better understand the impacts of WTH and whether stands are at risk of long-term nutrient losses, which could contribute to declines in productivity. The original study assessed: (i) full biomass removal (referred to as WTH), (ii) partial biomass removal (removal to 5 cm diameter bole), (iii) a traditional SOH (removal to 10 cm bole), and (iv) no harvest (control). Our data were collected 93 months after harvest and showed significant decreases (α = 0.05) in the total (soil, woody debris, litter, and tree) Ca between WTH and SOH and for N and K from all harvest types. P and Mg were not significantly different between any of the treatment types. The woody debris was significantly decreased in the WTH in comparison to the SOH across P, K, Ca, Mg, but not N. The differences in the woody debris were initially detected, but a nutrient budget modeling exercise did not show significant declines for the ecosystem. The nutrient budget model assumptions and scale of input terms are discussed. Revisiting the sites at 93 months post-harvest versus the original 1.75 years allowed for the detection of the differences in total N, K, and Ca pool sizes.

Sandy outwash and glacial till soils compose large amounts of public forestland due to historically poor agricultural yields. The outwash soils have low fertility, poor nutrient retention and are restricted from whole-tree harvesting (WTH) in some states, whereas the glacial till has medium nutrient retention and fertility, and is unrestricted from WTH. To assess the long-term sustainability of harvesting, a nutrient budget was constructed from field measurements, the National Cooperative Soil Survey (NCSS) database, and literature values for stem-only harvesting (SOH) and WTH at a 45-year rotation length and 11 rotations were simulated. The budgets showed that SOH and WTH recovery years, or the time necessary for the inputs to match outputs through leaching and one harvest, exceeded common rotation lengths for both soil types under all weathering scenarios, and the average WTH reduced the total available rotations by one harvest. The large variation in soil nutrient pools and harvest removals complicated the ability to identify the difference between SOH and WTH early in the model, but differences became apparent with sequential harvests. The recovery years were 2–20 times the 45-year rotation length under all weathering rates. Taken together, models in this study bridge the gap between short- and long-term studies and bring into question the sustainability of WTH and SOH practices on nutrient-poor soils.

The whole-tree harvest of an oak forest showed significant decreases to total nitrogen and potassium pools, and the calcium pool showed weaker evidence for significant declines in comparison to the stem-only harvest. Simulated whole-tree harvests in aspen decreased the total number of rotations by a full rotation on nutrient-poor outwash soils (Entic Haplorthods and Typic Udipsamments) by depleting the potassium, calcium, and magnesium pools in 180 years from whole-tree harvest in comparison to 225 years in the stem-only harvest. Nitrogen and phosphorus harvest removals did not compose a large percentage of ecosystem pools. Since harvest removals for some elements exceed the inputs at the current rotation length, inputs to the system should be increased if sustained yields at the current volumes are desired. In comparison to the ecosystem nutrient pools observed in this study, there is textural and management history variation across the landscape that could result in different ecosystem nutrient pool sizes located on the outwash soils. Soil complexes (Alfic Haplorthods-Entic Haplorthods/Typic Udipsamments) have differing soil properties and are currently partially restricted at the estimated composition of the complex. Habitat type provides a tool to distinguish between the soil types within the soil complexes. Relative elevation in the form of the deviation from mean elevation (DEV) combined with the moisture stress index (MSI) provide a framework to estimate the soil types remotely. DEV performed best for mapping the landforms at the 1300-meter radius with 10-meter digital elevation models. Landform mapping in the water-worked glacial till context will misclassify coarser textured soils incorporated into the moraine and erosion from the finer textured soils of the moraine to the lower outwash regions. Habitat type will detect the coarser textured soils, and MSI will identify the cases of erosion where trees reside on finer soil textures. The National Cooperative Soil Survey pedon database allowed for the detection of differences in coarse silt between Entic Haplorthods and Typic Udipsamments.