Geraldine L. Tierney

Colder soils in a warmer world: A snow manipulation study in a northern hardwood forest ecosystem

In this special section of Biogeochemistry, we present results from asnow manipulation experiment in the northernhardwood forest ecosystem at the Hubbard BrookExperimental Forest in the White Mountains ofNew Hampshire, U.S.A. Snow is important as aninsulator of forest soils. Later developmentof snowpacks, as may occur in a warmer climate,may result in increases in soil freezing (i.e.colder soils in a warmer world) and could causechanges in fine root and microbial mortality,hydrologic and gaseous losses of nitrogen (N),and the acid-base status of drainage water. Inour study, we kept soils snow free by shovelinguntil early February during the mild winters of1997/1998 and 1998/1999. The treatment producedmild, but persistent soil freezing and inducedsurprisingly significant effects on rootmortality, soil nitrate (NO3−) levelsand hydrologic fluxes of C, N and P. In thisspecial section we present four papersaddressing, (1) soil temperature and moistureresponse to our snow manipulation treatment(Hardy et al.), (2) theresponse of fine root dynamics to treatment(Tierney et al.), (3) theresponse of soil inorganic N levels, insitu N mineralization and nitrification,denitrification and microbial biomass to thetreatment (Groffman et al.)and (4) soil solution concentrations and fluxesof C, N and P (Fitzhugh et al.). In this introductory paper we: (1)review the literature on snow effects on forestbiogeochemistry, (2) introduce our manipulationexperiment and (3) summarize the resultspresented in the other papers in this issue.

Soil freezing alters fine root dynamics in a northern hardwood forest

The retention of nutrients within an ecosystem depends on temporal andspatial synchrony between nutrient availability and nutrient uptake, anddisruption of fine root processes can have dramatic impacts on nutrientretention within forest ecosystems. There is increasing evidence thatoverwinter climate can influence biogeochemical cycling belowground,perhaps by disrupting this synchrony. In this study, we experimentallyreduced snow accumulation in northern hardwood forest plots to examinethe effects of soil freezing on the dynamics of fine roots (< 1 mm diameter)measured using minirhizotrons. Snow removal treatment during therelatively mild winters of 1997–1998 and 1998–1999 induced mild freezingtemperatures (to −4 °C) lasting approximately three months atshallow soil depths (to −30 cm) in sugar maple and yellow birch stands.This treatment resulted in elevated overwinter fine root mortality in treatedcompared to reference plots of both species, and led to an earlier peak infine root production during the subsequent growing season. These shiftsin fine root dynamics increased fine root turnover but were not largeenough to significantly alter fine root biomass. No differences inmorality response were found between species. Laboratory tests on pottedtree seedlings exposed to controlled freezing regimes confirmed that mildfreezing temperatures (to −5 °C) were insufficient to directlyinjure winter-hardened fine roots of these species, suggesting that themarked response recorded in our forest plots was caused indirectly bymechanical damage to roots in frozen soil. Elevated fine root necromass intreated plots decomposed quickly, and may have contributed an excess fluxof about 0.5 g N/m2·yr, which is substantial relative tomeasurements of N fluxes from these plots. Our results suggest elevatedoverwinter mortality temporarily reduced fine root length in treatmentplots and reduced plant uptake, thereby disrupting the temporalsynchrony between nutrient availability and uptake and enhancing ratesof nitrification. Increased frequency of soil freezing events, as may occurwith global change, could alter fine root dynamics within the northernhardwood forest disrupting the normally tight coupling between nutrientmineralization and uptake.

Effects of soil freezing disturbance on soil solution nitrogen, phosphorus, and carbon chemistry in a northern hardwood ecosystem

Reductions in snow cover undera warmer climate may cause soil freezing eventsto become more common in northern temperateecosystems. In this experiment, snow cover wasmanipulated to simulate the late development ofsnowpack and to induce soil freezing. Thismanipulation was used to examine the effects ofsoil freezing disturbance on soil solutionnitrogen (N), phosphorus (P), and carbon (C)chemistry in four experimental stands (twosugar maple and two yellow birch) at theHubbard Brook Experimental Forest (HBEF) in theWhite Mountains of New Hampshire. Soilfreezing enhanced soil solution Nconcentrations and transport from the forestfloor. Nitrate (NO3−) was thedominant N species mobilized in the forestfloor of sugar maple stands after soilfreezing, while ammonium (NH4+) anddissolved organic nitrogen (DON) were thedominant forms of N leaching from the forestfloor of treated yellow birch stands. Rates ofN leaching at stands subjected to soil freezingranged from 490 to 4,600 mol ha−1yr−1, significant in comparison to wet Ndeposition (530 mol ha−1 yr−1) andstream NO3− export (25 mol ha−1yr−1) in this northern forest ecosystem. Soil solution fluxes of Pi from the forestfloor of sugar maple stands after soil freezingranged from 15 to 32 mol ha−1 yr−1;this elevated mobilization of Pi coincidedwith heightened NO3− leaching. Elevated leaching of Pi from the forestfloor was coupled with enhanced retention ofPi in the mineral soil Bs horizon. Thequantities of Pi mobilized from the forestfloor were significant relative to theavailable P pool (22 mol ha−1) as well asnet P mineralization rates in the forest floor(180 mol ha−1 yr−1). Increased fineroot mortality was likely an important sourceof mobile N and Pi from the forest floor,but other factors (decreased N and P uptake byroots and increased physical disruption of soilaggregates) may also have contributed to theenhanced leaching of nutrients. Microbialmortality did not contribute to the acceleratedN and P leaching after soil freezing. Resultssuggest that soil freezing events may increaserates of N and P loss, with potential effectson soil N and P availability, ecosystemproductivity, as well as surface wateracidification and eutrophication.

Effects of mild winter freezing on soil nitrogen and carbon dynamics in a northern hardwood forest

Overwinter and snowmelt processes are thought to be critical to controllersof nitrogen (N) cycling and retention in northern forests. However, therehave been few measurements of basic N cycle processes (e.g.mineralization, nitrification, denitrification) during winter and littleanalysis of the influence of winter climate on growing season N dynamics.In this study, we manipulated snow cover to assess the effects of soilfreezing on in situ rates of N mineralization, nitrification and soilrespiration, denitrification (intact core, C2H2 – based method),microbial biomass C and N content and potential net N mineralization andnitrification in two sugar maple and two yellow birch stands with referenceand snow manipulation treatment plots over a two year period at theHubbard Brook Experimental Forest, New Hampshire, U.S.A. The snowmanipulation treatment, which simulated the late development of snowpackas may occur in a warmer climate, induced mild (temperatures >−5 °C) soil freezing that lasted until snowmelt. The treatmentcaused significant increases in soil nitrate (NO3−)concentrations in sugar maple stands, but did not affect mineralization,nitrification, denitrification or microbial biomass, and had no significanteffects in yellow birch stands. Annual N mineralization and nitrificationrates varied significantly from year to year. Net mineralization increasedfrom ∼12.0 g N m−2 y−1 in 1998 to ∼22 g N m−2 y−1 in 1999 and nitrification increased from ∼8 g N m−2 y−1 in 1998 to ∼13 g N m−2 y−1 in 1999.Denitrification rates ranged from 0 to 0.65 g N m−2 y−1. Ourresults suggest that mild soil freezing must increase soil NO3− levels by physical disruption of the soil ecosystem and not by direct stimulation of mineralization and nitrification. Physical disruption canincrease fine root mortality, reduce plant N uptake and reduce competitionfor inorganic N, allowing soil NO3− levels to increase evenwith no increase in net mineralization or nitrification.

Snow depth manipulation and its influence on soil frost and water dynamics in a northern hardwood forest

Climate change will likelyresult in warmer winter temperatures leading toless snowfall in temperate forests. Thesechanges may lead to increases in soil freezingbecause of lack of an insulating snow cover andchanges in soil water dynamics during theimportant snowmelt period. In this study, wemanipulated snow depth by removing snow for twowinters, simulating the late development of thesnowpack as may occur with global warming, toexplore the relationships between snow depth,soil freezing, soil moisture, and infiltration.We established four sites, each with two pairedplots, at the Hubbard Brook Experimental Forest(HBEF) in New Hampshire, U.S.A. and instrumentedall eight plots with soil and snow thermistors,frost tubes, soil moisture probes, and soillysimeters. For two winters, we removed snowfrom the designated treatment plots untilFebruary. Snow in the reference plots wasundisturbed. The treatment winters (1997/1998 and1998/1999) were relatively mild, withtemperatures above the seasonal norm and snowdepths below average. Results show the treatedplots accumulated significantly less snow andhad more extensive soil frost than referenceplots. Snow depth was a strong regulator ofsoil temperature and frost depth at all sites.Soil moisture measured by time domainreflectometry probes and leaching volumescollected in lysimeters were lower in thetreatment plots in March and April compared tothe rest of the year. The ratio of leachatevolumes collected in the treatment plots tothat in the reference plots decreased as thesnow ablation seasons progressed. Our data showthat even mild winters with low snowfall,simulated by snow removal, will result inincreased soil freezing in the forests at theHBEF. Our results suggest that a climate shifttoward less snowfall or a shorter duration ofsnow on the ground will produce increases insoil freezing in northern hardwood forests.Increases in soil freezing will haveimplications for changes in soil biogeochemicalprocesses.

Evaluating minirhizotron estimates of fine root longevity and production in the forest floor of a temperate broadleaf forest

The minirhizotron technique (MR) for in situ measurement of fine root dynamics offers the opportunity to obtain accurate and unbiased estimates of root production in perennial vegetation only if MR tubes do not affect the longevity of fine roots. Assuming fine root biomass is near steady-state, fine root production (g m−2 yr−1) can be estimated as the ratio of fine root biomass (g m−2) to median fine root longevity (yr). This study evaluates the critical question of whether MR access tubes affect the longevity of fine roots, by comparing fine root survivorship obtained using MR with those from a non-intrusive in situ screen method in the forest floor horizons of a northern hardwood forest in New Hampshire, USA. Fine root survivorship was measured in 380 root screens during 1993–1997 and in six horizontal minirhizotron tubes during 1996–1997. No statistically significant difference was found between estimates of survivorship of fine roots (<1 mm dia.) at this site from MR versus from in situ screens, suggesting that MR tubes do not substantially affect fine root longevity in the forest floor of this northern hardwood forest and providing greater confidence in measurements of fine root production using the MR technique. Furthermore, the methodology for estimating fine root production from MR longevity data was evaluated by comparison of fine root longevity and production estimates made using single vs. multiple root cohorts, and using root-number, root-length, and root-mass weighted methods. Our results indicate that fine root-length longevity estimates based on multiple root cohorts throughout the year can be used to approximate fine root biomass production. Using this method, we estimated fine root longevity and production in the forest floor at this site to be 314 days (or 0.86 yr) and 303 g m−2 yr−1, respectively. Fine root production in this northern hardwood forest is approximately equivalent to standing biomass and was previously underestimated by root in-growth cores. We conclude that the use of MR to estimate fine root longevity and production as outlined here may result in improved estimates of fine root production in perennial vegetation.

Exploring patterns of exotic earthworm distribution in a temperate hardwood forest in south-central New York, USA.

Exotic earthworms invading forests in Canada and northeastern United States that were naturally devoid of large detritivores cause major changes in ecosystem function. To assess their long-term impacts, studies are needed to elucidate the factors that control the patterns of earthworm invasion at the landscape level. We analyzed the distribution patterns of exotic earthworms in a northern hardwood forest in south-central New York (USA), as explained by landscape variables thought to be important in determining earthworm distribution. Forest type, slope angle, elevation, and the distance to agricultural clearings and wet refugia were significant predictors of earthworm presence, whereas local wetness index and the distance to streams and roads were not. Forest type and distance to agricultural clearings were the two most significant predictors. Our data suggest that areas close to agricultural clearings, dominated by mixed hardwoods, and located towards valley bottoms or on gentle slopes are very likely to support communities of exotic earthworms. Steeper slopes, areas dominated by American beech or eastern hemlock, and locations in the core of extensive forest landscapes have lower probabilities of invasion by exotic earthworms. When applied to a nearby area, our statistical model correctly predicted earthworm presence for 67% of 377 sampling points. Most of the mistakes were incorrect predictions of earthworm absence, suggesting that our statistical model slightly underestimated earthworm presence, possibly because of the pervasive influence of active agricultural fields adjacent to the test site.