Kathryn L. Hofmeister’s research while affiliated with University of Wisconsin–Oshkosh and other places

Despite decades of progress, much remains unknown about successional trajectories of carbon (C) cycling in north temperate forests. Drivers and mechanisms of these changes, including the role of different types of disturbances, are particularly elusive. To address this gap, we synthesized decades of data from experimental chronosequences and long‐term monitoring at a well‐studied, regionally representative field site in northern Michigan, USA. Our study provides a comprehensive assessment of changes in above‐ and belowground ecosystem components over two centuries of succession, links temporal dynamics in C pools and fluxes with underlying drivers, and offers several conceptual insights to the field of forest ecology. Our first advance shows how temporal dynamics in some ecosystem components are consistent across severe disturbances that reset succession and partial disturbances that slightly modify it: both of these disturbance types increase soil N availability, alter fungal community composition, and alter growth and competitive interactions between short‐lived pioneer and longer‐lived tree taxa. These changes in turn affect soil C stocks, respiratory emissions, and other belowground processes. Second, we show that some other ecosystem components have effects on C cycling that are not consistent over the course of succession. For example, canopy structure does not influence C uptake early in succession but becomes important as stands develop, and the importance of individual structural properties changes over the course of two centuries of stand development. Third, we show that in recent decades, climate change is masking or overriding the influence of community composition on C uptake, while respiratory emissions are sensitive to both climatic and compositional change. In synthesis, we emphasize that time is not a driver of C cycling; it is a dimension within which ecosystem drivers such as canopy structure, tree and microbial community composition change. Changes in those drivers, not in forest age, are what control forest C trajectories, and those changes can happen quickly or slowly, through natural processes or deliberate intervention. Stemming from this view and a whole‐ecosystem perspective on forest succession, we offer management applications from this work and assess its broader relevance to understanding long‐term change in other north temperate forest ecosystems.

Soil organic matter (SOM) influences a wide range of ecosystem processes, including nutrient cycling, water movement, plant productivity, and biodiversity. In agricultural landscapes, adjacent land uses often differ in SOM contents and related soil properties, such as soil organic carbon (SOC) stocks, but the direction and magnitude of these effects are inconsistent across studies. We assessed how land uses differed in SOM and related properties in a representative US Midwest agricultural–forest landscape to support land‐use and management decisions by local landowners and producers. We measured SOM, bulk density (Db), root biomass, and pH, and estimated SOC stocks, in a Typic Hapludalf under four adjacent land uses (permanent forest, pasture, restored prairie on former pasture, and spruce plantation on former pasture). Surface SOM concentrations and stocks were higher under permanent forest (89 g kg⁻¹ and 85 Mg ha⁻¹, respectively) and pasture (63 g kg⁻¹ and 81 Mg ha⁻¹, respectively) than under restored prairie (49 g kg⁻¹ and 58 Mg ha⁻¹, respectively) and spruce plantation (46 g kg⁻¹ and 46 Mg ha⁻¹, respectively). Land uses also differed in Db, root biomass, and pH, with permanent forest and spruce plantation soils having generally lower Db, more root biomass, and more acidic pH than pasture and restored prairie soils. Specific statistically significant differences depended upon depth in the soil profile. Overall, our results suggest that each land use differentially impacts a unique set of soil properties, precluding any single explanation or management recommendation aimed at improving soil health as a whole.

The decision to establish a network of researchers centers on identifying shared research goals. Ecologically specific regions, such as the USA’s National Ecological Observatory Network’s (NEON’s) eco-climatic domains, are ideal locations by which to assemble researchers with a diverse range of expertise but focused on the same set of ecological challenges. The recently established Great Lakes User Group (GLUG) is NEON’s first domain specific ensemble of researchers, whose goal is to address scientific and technical issues specific to the Great Lakes Domain 5 (D05) by using NEON data to enable advancement of ecosystem science. Here, we report on GLUG’s kick off workshop, which comprised lightning talks, keynote presentations, breakout brainstorming sessions and field site visits. Together, these activities created an environment to foster and strengthen GLUG and NEON user engagement. The tangible outcomes of the workshop exceeded initial expectations and include plans for (i) two journal articles (in addition to this one), (ii) two potential funding proposals, (iii) an assignable assets request and (iv) development of classroom activities using NEON datasets. The success of this 2.5-day event was due to a combination of factors, including establishment of clear objectives, adopting engaging activities and providing opportunities for active participation and inclusive collaboration with diverse participants. Given the success of this approach we encourage others, wanting to organize similar groups of researchers, to adopt the workshop framework presented here which will strengthen existing collaborations and foster new ones, together with raising greater awareness and promotion of use of NEON datasets. Establishing domain specific user groups will help bridge the scale gap between site level data collection and addressing regional and larger ecological challenges.

Soils with low permeability horizons (e.g., claypans) are vulnerable to loss of nutrients through surface runoff along with preferential flow paths through the restrictive horizon to deeper aquifers. Partitioning between these hydrologic pathways is important to determine transport processes and develop strategies that mitigate stream contamination. Our objective was to investigate controls on groundwater chemistry and recharge pathways using natural geochemical tracers in the Goodwater Creek Experimental Watershed in Missouri, U.S. Groundwater samples were collected during 2011–2017 from 32 piezometers ranging from 0.13 to 16 m deep along with stream water and precipitation. Diagnostic tools of mixing models indicated that chemistry of perched water directly above the claypan and shallow groundwater immediately below was controlled primarily by chemical equilibrium. Five solutes behaved conservatively in most deep piezometers (>5 m), reflecting mixing of two end members and the lack of significant denitrification processes. End member mixing analysis showed that the deeper groundwater originated primarily from groundwater at similar depths, often upslope or from strata directly above, with small contributions from perched water, highlighting the importance of both horizontal and vertical preferential recharge pathways. Vertical pathways are likely dictated by soil heterogeneity throughout the critical zone and do not occur synchronously with precipitation events or simultaneously over all piezometer locations. The complex recharge pathways provide stochastic conduits for nitrate transport to deeper aquifers where legacy stores accumulate, presenting a significant challenge for water quality management in watersheds with restrictive soil horizons and spatially and temporally heterogeneous preferential flow pathways, including the Mississippi River Basin.

Seasonal ponds are small, isolated wetlands with variable hydrology, often occurring embedded in upland forests, which provide habitat for amphibians and invertebrates uniquely adapted to fishless waters. Seasonal ponds are challenging to identify due to their small size, ephemeral hydrology, diverse vegetation, and occurrence across a range of settings, yet in order to inform their conservation and management, it is essential to understand their distribution and how management impacts them. We conducted a systematic review to define and quantify attributes of seasonal ponds, summarize mapping and inventory methods, and synthesize forest harvesting impacts on ponds in the western Great Lakes and northeastern United States. Definitions of seasonal ponds differ regionally and for scientific vs. regulatory purposes; the necessity of documenting pond-dependent indicator species (e.g., fairy shrimp) is one of the most vexing inconsistencies. Seasonal ponds are most effectively mapped in the spring, using a combination of aerial photographs or radar imagery and topographic information, especially in settings with small ponds or heavy canopies. Combining these mapping efforts with carefully stratified field validation is essential for developing a regional inventory of seasonal ponds. Most guidelines intended to reduce impacts of forest harvesting on pond ecosystems rely on buffers, which most effectively minimize physical or biological impacts when most lightly treated, although some impacts (particularly water levels) appear unavoidable when any harvesting occurs adjacent to seasonal ponds. Overall, distinct physical and biological impacts of harvesting differ in magnitude and direction, though most appear to subside over multi-decadal timescales.

This study coupled a landscape‐scale metagenomic survey of denitrification gene abundance in soils with in‐situ denitrification measurements to show how environmental factors shape distinct denitrification communities that exhibit varying denitrification activity. Across a hydrologic gradient, the distribution of total denitrification genes (nap/nar + nirK/nirS + cNor/qNor + nosZ) inferred from metagenomic read abundance exhibited no consistent patterns. However, when genes were considered independently, nirS, cNor, and nosZ read abundance was positively associated with areas of higher soil moisture, higher nitrate, and higher annual denitrification rates, while nirK and qNor read abundance was negatively associated with these factors. These results suggest that environmental conditions, in particular soil moisture and nitrate, select for distinct denitrification communities that are characterized by differential abundance of genes encoding apparently functionally redundant proteins. In contrast, taxonomic analysis did not identify notable variability in denitrifying community composition across sites. While the capacity to denitrify was ubiquitous across sites, denitrification genes with higher energetic costs, like nirS and cNor, appear to confer a selective advantage in microbial communities experiencing more frequent soil saturation and greater nitrate inputs. This study suggests metagenomics can help identify denitrification hotspots that could be protected or enhanced to treat nonpoint source nitrogen pollution. This article is protected by copyright. All rights reserved.

Core Ideas Soil moisture and water table position vary across ecosystems and parent materials. Ecosystem classifications and topographic wetness indices represent moisture status. Topographic wetness indices perform best in till rather than outwash soils. Including soil properties with topographic information improves moisture estimates. In forest ecosystems, soil water availability is an important indicator and driver of biogeochemical transformations, pedogenesis, and surface water–groundwater linkages. Given the importance of soil moisture and shallow groundwater to ecosystem processes, field measurements are critical. We investigated the spatial patterns and temporal dynamics of moisture conditions in a forested first‐order watershed in northern Michigan. We measured soil volumetric water content (VWC, %) and water table levels across a range of glacial parent materials and landforms. We also assessed the utility of using two different spatial frameworks (Landscape Ecosystem classification, Topographic Wetness Index) to represent spatial patterns of soil moisture and water table levels across the watershed. At the lowest landscape position (outwash‐lake plain swamp), saturation was perennial, with a median soil VWC of 53% and the water table 8 cm below ground surface. Among upland ecosystems, outwash landforms had consistently low VWC (16%) and showed no evidence of groundwater within 4 m of the surface; moraine ecosystems (till parent material) possessed mixed hydrologic conditions, with VWC ranging from 18 to 25% and water table levels 6 to 65 cm below the surface. In low‐variability dry or wet areas with relatively homogenous soils, larger ecosystem classification map units provide good representations of moisture conditions. In the more heterogeneous till soils, a finer‐scale spatial framework that accounts for local soil variation is optimal. A combination of both spatial frameworks is most appropriate for estimating moisture conditions in this glaciated landscape and can be used to identify biogeochemically important sites and to inform land management decisions.

Headwater streams are critical components of drainage systems, directly connecting terrestrial and downstream aquatic ecosystems. The amount of water in a stream can alter hydrologic connectivity between the stream and surrounding landscape, and is ultimately an important driver of what constituents headwater streams transport. There is a shortage of studies that explore concentration‐discharge (C‐Q) relationships in headwater systems, especially forested watersheds, where the hydrological and ecological processes that control the processing and export of solutes can be directly investigated. We sought to identify the temporal dynamics and spatial patterns of stream chemistry at three points along a forested headwater stream in northern Michigan and utilize concentration‐discharge (C‐Q) relationships to explore transport dynamics and potential sources of solutes in the stream. Along the stream, surface flow was seasonal in the main stem and perennial flow was spatially discontinuous for all but the lowest reaches. Spring snowmelt was the dominant hydrological event in the year with peak flows an order of magnitude larger at the mouth and upper reaches than annual mean discharge. All three C‐Q shapes (positive, negative, flat) were observed at all locations along the stream, with a higher proportion of the analytes showing significant relationships at the mouth than at the mid or upper flumes. At the mouth, positive (flushing) C‐Q shapes were observed for dissolved organic carbon and total suspended solids, while negative (dilution) C‐Q shapes were observed for most cations (Na+, Mg2+, Ca2+) and biologically cycled anions (NO3‐, PO43‐, SO42‐). Most analytes displayed significant C‐Q relationships at the mouth, indicating that discharge is a significant driving factor controlling stream chemistry. However, the importance of discharge appeared to decrease moving upstream to the headwaters where more localized or temporally‐dynamic factors may become more important controls on stream solute patterns.

In the United States (U.S.), the maintenance of forest cover is a legal mandate for federally managed forest lands. More broadly, reforestation following harvesting, recent or historic disturbances can enhance numerous carbon (C)-based ecosystem services and functions. These include production of woody biomass for forest products, and mitigation of atmospheric CO2 pollution and climate change by sequestering C into ecosystem pools where it can be stored for long timescales. Nonetheless, a range of assessments and analyses indicate that reforestation in the U.S. lags behind its potential, with the continuation of ecosystem services and functions at risk if reforestation is not increased. In this context, there is need for multiple independent analyses that quantify the role of reforestation in C sequestration, from ecosystems up to regional and national levels. Here, we describe the methods and report the findings of a large-scale data synthesis aimed at four objectives: (1) estimate C storage in major ecosystem pools in forest and other land cover types; (2) quantify sources of variation in ecosystem C pools; (3) compare the impacts of reforestation and afforestation on C pools; (4) assess whether these results hold or diverge across ecoregions. The results of our synthesis support four overarching inferences regarding reforestation and other land use impacts on C sequestration. First, in the bigger picture, soils are the dominant C pool in all ecosystems and land cover types in the U.S., and soil C pool sizes vary less by land cover than by other factors, such as spatial variation or soil wetness. Second, where historically cultivated lands are being reforested, topsoils are sequestering significant amounts of C, with the majority of reforested lands yet to reach their capacity relative to the potential indicated by natural forest soils. Third, the establishment of woody vegetation delivers immediate to multi-decadal C sequestration benefits in aboveground woody biomass and coarse woody debris pools, with two- to three-fold C sequestration benefits in biomass during the first several decades following planting. Fourth, opportunities to enhance C sequestration through reforestation vary among the ecoregions, according to current levels of planting, typical forest growth rates, and past land uses (especially cultivation). Altogether, our results suggest that an immediate, but phased and spatially targeted approach to reforestation can enhance C sequestration in forest biomass and soils in the U.S. for decades to centuries to come.

Significance Forestland in the United States is a carbon (C) sink, offsetting ∼10% of annual greenhouse gas emissions and mitigating climate change. Most of the C in forests is held in soils, and the capacity of forest soils to sequester C makes them a major component of the US forest C sink. Where reforestation is presently occurring, either through deliberate replanting after forestland is disturbed (e.g., burned), or where previously nonforested lands (e.g., cultivated) are converting to forestland, topsoils are accumulating C. However, these C accumulation rates are poorly constrained; quantifying them with empirical data are critical to accurately represent the role of reforestation in the US C budget and forecast the longevity of the US forest C sink.