nena masthead
NENA Home Staff & Editors For Readers For Authors

Snowpack Loss Promotes Soil Freezing and Concrete Frost Formation in a Northeastern Temperate Softwoods Stand
Corianne Tatariw, Kaizad Patel, Jean D. MacRae, and Ivan J. Fernandez

Northeastern Naturalist,Volume 24, Special Issue 7 (2017): B42–B54

Full-text pdf (Accessible only to subscribers.To subscribe click here.)


Access Journal Content

Open access browsing of table of contents and abstract pages. Full text pdfs available for download for subscribers.

Issue-in-Progress: Vol.30 (1) ... early view

Current Issue: Vol. 29 (4)
NENA 29(4)

All Regular Issues


Special Issues






JSTOR logoClarivate logoWeb of science logoBioOne logo EbscoHOST logoProQuest logo

Northeastern Naturalist C. Tatariw, K. Patel, J.D. MacRae, and I.J. Fernandez 2017 42 Vol. 24, Special Issue 7 Snowpack Loss Promotes Soil Freezing and Concrete Frost Formation in a Northeastern Temperate Softwoods Stand Corianne Tatariw1,2,*, Kaizad Patel3, Jean D. MacRae4, and Ivan J. Fernandez3 Abstract - Snowpack produces a thermal layer that protects soil from freezing and provides a pulse of nutrient-rich water in spring. Climate forecasts for Maine indicate 20–60% reduction in snowfall by 2050. In January 2015, we initiated a snow-removal experiment in Old Town, ME to investigate the impact of snow loss on forest soil conditions. Snow removal significantly lowered winter organic horizon temperatures by 2 °C on average. Soils in snow-removal plots were 25% wetter during the vernal transition because precipitation was not intercepted by snow. These rain-on-soil events caused the formation of concrete frost, delaying soil thaw in snow-removal plots. Our results provide evidence that snowpack loss increases soil frost and can also increase soil moisture, potentially altering biotic function within a coniferous forest type. Introduction Winter snowpack is ecologically important in the northeastern United States because it insulates soil and is a source of water and nutrients in the spring (Campbell et al. 2005, Piatek et al. 2005), but climate change has warmed winters and reduced snowfall amount and snowpack duration in the region (United States Environmental Protection Agency 2013). In Maine, state-wide mean annual air temperatures have increased by 1.6 °C, snowfall has decreased by 2.5 cm (6.6%), and snowpack duration has decreased by ~2 weeks since 1895 (Fernandez et al. 2015). By 2050, compared to recent averages, annual snowfall in Maine is expected to decline by 20–40% or more in all but the Northern Interior (i.e., northwestern-most) climate division of the state (Fernandez et al. 2015). However, in spite of these environmental changes for Maine winters, there is little published research on wintertime soil processes in Maine, a state that is 90% forested (Huff and McWilliams 2016). Loss of snowpack fundamentally alters the magnitude of seasonal changes in soil moisture and temperature, 2 factors that regulate biological processes in soil. When snowpack is reduced or absent, soils are more likely to undergo freeze/thaw cycles that physically alter soil structure and negatively impact multiple taxonomic levels of soil biota from bacteria to forest grazers (Chamberlain and Gow 1979, Hinman and Bisal 1968, Williams et al. 2014). Snowpack also provides a pulse of water and nutrients during the vernal transition (i.e., the period of snowmelt during the transition from winter to spring), which is reduced with a diminishing snowpack 1Department of Biological Sciences, Box 870344, University of Alabama, Tuscaloosa, AL 35487. 2 The Dauphin Island Sea Lab, 101 Beinville Boulevard, Dauphin Island, AL 36528. 3School of Forest Resources, 5755 Nutting Hall, University of Maine, Orono, ME 04469. 4Civil and Environmental Engineering, 5711 Boardman Hall, University of Maine, Orono, ME 04469. *Corresponding author - Manuscript Editor: Christoph Geiss Winter Ecology: Insights from Biology and History 2017 Northeastern Naturalist 24(Special Issue 7):B42–B54 Northeastern Naturalist 43 C. Tatariw, K. Patel, J.D. MacRae, and I.J. Fernandez 2017 Vol. 24, Special Issue 7 (Brooks et al. 1998, Campbell et al. 2014). Earlier soil thaw and snow melt caused by reduced snowpack and warmer winters directly affect soil processes by making fresh water available earlier in the year, which could result in drought stress during the growing season (Grimm et al. 2013, Groffman et al. 2012). However, in Maine, this reduction in water availability may be offset by an increase in early spring rain events as climate change reduces the frequency of frozen forms of precipitation (Fernandez et al. 2015). There have been multiple studies on the effects of climate change on deciduous forest soil processes in the northeastern United States (e.g., Campbell et al. 2014; Decker et al. 2003; Groffman et al. 2001, 2011; Melillo et al. 2002; Reinmann and Templer 2015; Sorensen et al. 2016). However, although about 33% of the 7.1 million ha of forested land area in Maine is dominated by conifer (Picea [spruce]/ Abies [fir]) species (Huff and McWilliams 2016), there are few published studies addressing the effects of snowpack loss on coniferous forest soil processes. Unlike deciduous trees, conifers photosynthesize during the winter (Öquist and Huner 2003, Schaberg et al. 1995) and can be important contributors to wintertime soil carbon (C) storage (Waring and Franklin 1979). However, soil freezing and water loss associated with reduced snowpack can have negative impacts on tree photosynthesis and growth. Sub-freezing soil temperatures reduce winter photosynthesis (Hadley 2000, Schaberg 2000), and freeze/thaw cycles may further suppress winter photosynthesis by decreasing plant cold hardiness (Schaberg 2000). Winter freeze/ thaw cycles can also contribute to longer-term reductions in conifer growth by damaging xylem conduits, effectively reducing plant water transport (Pittermann and Sperry 2006), and the loss of springtime snowmelt can also cause drought stress, especially in saplings (Boyce and Lucero 1999). The objective of this study was to evaluate soil microclimate responses to a declining snowpack in a Maine conifer stand in light of climate trends and predictions for less snowpack in a warming climate. Here we describe the impact of snow on soil temperature and moisture in a coniferous forest stand by experimental snow removal in University Forest, Old Town, ME. We predicted that snow removal would (1) enhance soil freezing in the winter, (2) result in more rapid warming during the vernal transition, and (3) decrease soil moisture through the vernal transition into spring due to a loss of snowmelt water. To our knowledge, this is the first such study reported in a northeastern US conifer forest. Site Description The study site was located in the University of Maine’s Dwight B. DeMeritt Forest (44°56'2.5"N, 68°39'51.1"W) in Old Town, ME, which is located in the Southern Interior climate division. Soils at the site were well-drained coarse loamy frigid Typic Haplorthods (Series: Bangor) (USDA Handbook 436), and had an O (organic) horizon 1–5 cm deep and a mineral horizon >90 cm thick. The site was located at the top of a ridge and covered by coniferous stands. Forest stand composition consisted of 38% Pinus strobus L. (Eastern White Pine), 27% Tsuga canadensis (L.) Carrière (Eastern Hemlock), and 14% Picea rubens Sarg. (Red Spruce) (Alden 1998). Northeastern Naturalist C. Tatariw, K. Patel, J.D. MacRae, and I.J. Fernandez 2017 44 Vol. 24, Special Issue 7 Methods Experimental design In December 2014, we established four 5 m x 20 m experimental plots. Each plot was divided into two 5 m x 10 m subplots—a reference plot, and a treatment (snow-removal) plot (Fig. 1). The reference plots were subjected to minimal disturbance and allowed to accumulate snow throughout the winter. The snow-removal plots were kept free of snow by shoveling through the entire winter. A 1-m buffer was established within the perimeter of each subplot (Fig. 1), providing access to the plots and providing a 2-m separation between coupled treatment and control plots. To reduce additional mechanical disturbance to the soil, shoveling was conducted by reaching in from the buffer to the extent possible. No samples were collected from the buffer. We initiated snow removal the first week of February following the establishment of snowpack in late January 2015. For the rest of the winter, snow was removed within 48 hours of a snow event with a total onsite accumulation of 5 cm or more. To minimize damage to the soil, we “swept” snow from the interior of the plot to the buffer area with shovels. The swept snow was then shoveled from the buffer area off of the plot. We left a ~2-cm base layer of snow on the ground to maintain albedo and reduce disruption of the litter layer from shoveling. Temperature We measured soil temperature at 20- or 30-minute intervals using Thermochron iButton temperature loggers (Model DS1921G, Maxim Integrated, San Jose, CA). iButtons were buried at 3 to 5 cm depth within the O horizon in resealable polyethylene bags. We calculated mean daily soil temperature for each subplot from 24 hours of temperature readings and obtained mean daily air temperature by averaging the daily maximum and minimum temperatures from the National Oceanic and Atmospheric Administration (NOAA) weather station located ~2.8 km east of the study site (Station GHCND:USW00094644; Menne et al. 2012a, b). Figure 1. Diagram of the reference (left) and treatment (right) plot layout. The black oval indicates the location of the iButton data loggers and the black X symbols indicate sampling locations on a roughly 1 m x 1 m grid. Northeastern Naturalist 45 C. Tatariw, K. Patel, J.D. MacRae, and I.J. Fernandez 2017 Vol. 24, Special Issue 7 We divided soil data into 4 time periods using a combination of mean daily air temperature data and snowpack observations: pre-treatment (24–31 January 2015) was the period of time prior to the initiation of snow removal when air temperatures were below freezing; winter (01 February–25 March 2015) was the treatment period when snowpack was present in the reference plots (Fig. 2A) and air temperatures were below freezing (Fig. 3A); vernal transition (26 March–15 April 2015), defined as the transition from winter to spring, started with the melting of the protective Figure 2. Pictures of a reference and treatment plot during (A) winter (treatment plot in foreground), (B) vernal transition (treatment plot on right), and (C) spring (reference plot outlined in solid line, treatment in dashed line). Northeastern Naturalist C. Tatariw, K. Patel, J.D. MacRae, and I.J. Fernandez 2017 46 Vol. 24, Special Issue 7 2-cm base layer of snow in the treatment plots, a time marked by fluctuations in the air temperature between above- and below-freezing (Fig. 3A), and ended when the snowpack in the reference plots was completely melted (Fig. 2B); and spring (16 April—11 June 2015) was the period of time following snowmelt (Fig. 2C) during which air temperatures were above freezing (Fig. 3A). Figure 3. (A) Mean daily temperatures for air (dashed gray), reference soil (solid gray), and treatment soil (solid black). The dotted horizontal line indicates 0 °C. Dashed vertical lines delineate our 4 defined time periods. (B) Mean ΔT for each time period. A negative ΔT indicates lower temperatures in treatment plots. Bars are ± standard deviation. Letters indicate significant differences between time periods at α = 0.05. (C) Hourly temperatures for reference (gray) and treatment (black) soil during the final week of the vernal transition/first days of spring. The final day of snow in the reference plots was on April 14. Northeastern Naturalist 47 C. Tatariw, K. Patel, J.D. MacRae, and I.J. Fernandez 2017 Vol. 24, Special Issue 7 Soil moisture We sampled soil at the site during 10 visits over 5 months to capture seasonal dynamics during winter, vernal transition, and spring (Fig. 2). Three collections took place during winter (07 February, 04 March, and 06 March 2015), 3 during the vernal transition (27 March, 08 April, and 15 April 2015), and 4 in spring (22 April, 29 April, 14 May, and 03 June 2015). O horizon soil samples were collected on a roughly 1 m x 1 m grid. We collected 1 sample per plot per sampling event. Soil within a 10 cm x 10 cm template was excavated down to the mineral horizon. Due to the shallow nature of the organic horizon, we collected multiple templates of soil to obtain enough mass for analysis. When the snowpack was present, we carefully shoveled snow from a roughly 0.25-m2 area prior to sampling. Following soil collection, the snow was replaced to minimize temperature and albedo disturbances. When frozen, soil samples were separated from the surrounding soil using a hammer and chisel along the inside perimeter of the sampling frames, then a trowel or hammer and chisel was used to dislodge the soil block at the O-mineral interface. If the soil was not frozen, we collected soil using a trowel with clippers and soil knives to separate the soil block from the surrounding soil. In all cases, loose litter at the surface was removed prior to sampling. Soil samples were placed in resealable polyethylene bags for transport to the laboratory, where they were held overnight in a refrigerator or in coolers with icepacks to allow frozen samples to thaw. We hand-sieved thawed, field-moist soils through a 6-mm mesh to separate fine earth from coarse fragments and roots. We oven-dried subsamples at 65 °C for 24 hours and calculated gravimetric moisture content (i.e., mass of water per unit mass of oven-dry soil) in triplicate for each sample. Data analysis We completed all analyses in R Version 3.2.2 using RStudio Version 0.99.485 (R Core Team 2015, RStudio Team 2015) and determined significant differences at α = 0.05. To determine the effect of snow removal on soil temperature, we calculated the difference in mean daily soil temperature between treatment and reference plots (ΔT). A lower ΔT indicated colder temperatures in treatment plots compared to the reference plots. ΔT was analyzed by one-way analysis of variance (ANOVA) with Welch’s Correction to account for unequal variances with time period (i.e., pretreatment, winter, vernal transition, and spring) as the factor (Moser and Stevens 1992). We used the Games-Howell post-hoc test for unequal variances to determine significant differences between time periods with the R package userfriendlyscience (Peters 2015). We visually evaluated hourly mean soil temperature measurements and identified 10–18 April as the period when soil temperatures began to dramatically fluctuate daily. We then calculated daily temperature amplitudes (difference in daily maximum and minimum temperature) for each plot during that time period. To determine the effect of snow removal on soil warming, we used a Kruskal-Wallis rank sum test to test for significant differences in temperature amplitude between treatments. Northeastern Naturalist C. Tatariw, K. Patel, J.D. MacRae, and I.J. Fernandez 2017 48 Vol. 24, Special Issue 7 Because we were only interested in treatment effects, we did not test for significant differences in amplitude between dates. To determine the effect of snow removal on gravimetric soil moisture over sampling dates, we used a linear mixed-effects model LME, R Package nlme (Pinheiro et al. 2015) with log10-transformed soil moisture as the response variable, treatment and day as interacting fixed factors, and plot as a random factor. The model was interpreted using analysis of variance and Tukey’s highly significant differences (HSD) post-hoc test to determine the effects of treatment and day on soil moisture. Results Temperature Winter pre-treatment soil temperatures were near freezing (0 °C) in the reference and treatment plots, while the air temperature was -5 to -15 °C (Fig. 3A). The effect of snow removal was seen on soil temperature shortly after initiation of the treatment. Winter mean daily soil temperatures ranged from 0 to 1 °C in the reference plots and -4.6 to 0.6° C in the treatment plots. During the vernal transition, mean daily soil temperatures ranged from 0.2 to 0.5 °C in the reference plots and -0.2 to 2.3 °C in the treatment plots (Fig. 3A). Following snowmelt in all plots (spring), reference and treatment plot soil temperatures were similar and generally increased over time from 1 to 22 °C (Fig. 3A). Delta-T varied significantly between time periods (one-way ANOVA: P < 0.001; Fig. 3B). Snow removal lowered soil temperature in winter, as is indicated by ΔT being significantly larger than pre-treatment (Games-Howell: P < 0.001), vernal transition (Games-Howell: P < 0.001), and spring (Games-Howell: P < 0.001 (Fig. 3B). Pre-treatment ΔT was lower than spring ΔT (Games-Howell: P < 0.001). During the vernal transition, mean daily soil temperature for all plots was within ± 1 °C of freezing until 11 April, when soil temperatures began to fluctuate on a diel basis in the snow-removal plots. On that date, soil temperatures in the treatment plots had a significantly larger amplitude of 3.67 °C, compared to 0.16 °C in the reference plots (Kruskal-Wallis: P = 0.020; Fig. 3C). Amplitude continued to be significantly larger in treatment plots (Kruskal-Wallis: P = 0.020 for all dates) until 17 April, 3 days after the snow cover had melted in the reference plots (Kruskal- Wallis: P = 0.149; Fig. 3C). Moisture Soil moisture was 25% higher in treatment than reference plots across all dates (LME: P < 0.001; Fig. 4). The linear mixed-effects model showed a significant interaction between treatment and date (LME: P < 0.001; Fig. 4). Soil moisture in the treatment plots was higher than reference plots during the first 2 vernal transition sampling days (Tukey’s HSD: P < 0.001), but there was no effect of snow removal on soil moisture during winter and spring. Within treatments, soil moisture varied significantly between collections in the snow-removal plots but not the reference plots. Soil moisture in the treatment plots was highest during the vernal transition and the lowest during the spring (Fig. 4). Northeastern Naturalist 49 C. Tatariw, K. Patel, J.D. MacRae, and I.J. Fernandez 2017 Vol. 24, Special Issue 7 Discussion Snow removal lowered soil temperature below freezing during winter and accelerated soil warming during the vernal transition. In the reference plots, soil temperature stayed at or slightly above freezing underneath snowpack, as the subnivean soil was thermally insulated throughout the winter. Snow, which mostly consists of air, has a low thermal conductivity (Zhang 2005). Depending on the density of the snowpack, the thermal conductivity can be 5–20 times lower than that of mineral soil, decoupling air and soil temperatures (Grundstein et al. 2005, Zhang 2005). Our reference plots demonstrated the insulating properties of the snowpack, with winter soil temperature under snowpack consistently at or above 0°C. In contrast, snow removal recoupled soil-air temperatures, as patterns in soil temperature in the treatment plots tracked fluctuations in air temperature. A snowremoval study conducted at the University of Vermont Proctor Maple Research Center (450 km west of our site) found that soil temperatures in snow-free plots tracked inter-annual variations in winter air temperatures; notably, during a warm winter, soils were often colder under snowpack than in snow-removal plots (Decker et al. 2003). Similarly, at the Harvard Forest in Massachusetts (390 km southwest of our site), the loss of insulating snowpack during a warmer winter contributed to soil frost formation in both snow-removal and reference plots (Reinmann and Templer Figure 4. Mean gravimetric moisture (left axis, circles) and daily rainfall (right axis, light gray fill). Frozen precipitation (i.e., snow) is not included. Vertical dashed lines delineate seasons. For moisture, treatment plots are indicated by dark gray circles and reference plots by open circles and bars are standard deviation. Asterisks indicate significant differences between treatments on that date. Letters indicate significant differences between dates for treatment plots (there were no significant dif ferences in soil moisture for reference plots). Northeastern Naturalist C. Tatariw, K. Patel, J.D. MacRae, and I.J. Fernandez 2017 50 Vol. 24, Special Issue 7 2015). Given that climate-change predictions indicate that decreased snowpack will be concomitant with warmer winters (Fernandez et al. 2015), it is important to note that our measurements, which were made in a colder than average winter (1.9 °C colder than the 30-year average, NOAA 2016), do not fully represent the predicted impacts of climate change on soil. However, regardless of the variability of temperature change, this and other studies (e.g., Cleavitt et al. 2008, Comerford et al. 2013) indicate that the loss of snowpack insulation associated with climate change will result in recoupling of air and soil temperatures, increasing wintertime temperature variability in soil. At the Hubbard Brook Experimental Forest (HBEF), which is ~265 km southwest of our site, mild soil freezing (0 to -4 °C) was cold enough to damage fine roots via cellular and physical disruption (Cleavitt et al. 2008, Tierney et al. 2001). Temperatures in our treatment plots were below -4°C on 12 days, with a minimum of -6 °C. Thus, it is probable that snow removal disrupted soil biological processes in treatment plots due to soil freezing. In the interior northeastern climate region, springtime biological activity typically increases once soil temperatures reach 4 °C (Groffman et al. 2012). Although soil temperatures warmed 3 days earlier in the treatment plots, damage to fine roots and microbes could delay springtime nutrient uptake, resulting in increased export to surface waters. Although we only measured O horizon soils (maximum depth = 5 cm), other studies in New England have shown that snow-removal results in colder soil temperatures as deep as 30 cm (Decker et al. 2003, Hardy et al. 2001) and soil frost formation at depths >15 cm (Hardy et al. 2001, Reinmann and Templer 2015) below the surface. While additional research is needed to explore these hypotheses, we believe that snow removal altered soil temperatures enough to impact soil biota, and by extension, ecosystem processes. Snow removal significantly altered temporal patterns in soil moisture. The maximum average snow depth measured in the reference plots was 540 ± 40 mm (equivalent to 79 ± 23 mm of liquid precipitation), so we expected moisture would be lower in the treatment plots due to the absence of snowmelt water as was observed at HBEF (Hardy et al. 2001). However, we found that soils were wettest in the snow removal plots during the vernal transition prior to the thaw. We propose that increases in soil moisture and persistent freezing soil temperatures that occurred during the vernal transition can be attributed to a 9.1-mm rain event in late March that penetrated the bare soil and froze, resulting in the formation of “concrete frost” in treatment plots. Concrete frost occurs when soils of high moisture content freeze (Shanley and Chalmers 1999, Zhang 2005), and we expect that it contributed to the freezing soil temperatures in the treatment plots until warmer air temperatures melted both the soil frost and snowpack at the end of March. Ice and snow require heat energy to melt, and this heat sink can result in freezing soil temperatures during snowmelt (Zhang 2005). In contrast, during rain-on-snow events (such as occurred in our reference plots during the vernal transition), most rainwater is absorbed by snow and does not reach the soil (Huntington et al. 2009). Rain that does percolate through snow forms frost at the snow-soil interface, warming Northeastern Naturalist 51 C. Tatariw, K. Patel, J.D. MacRae, and I.J. Fernandez 2017 Vol. 24, Special Issue 7 soil temperatures farther down in the soil profile through the release of latent heat during freezing (Putkonen and Roe 2003). Soil temperatures in the reference plots stayed above freezing, suggesting that frost did not penetrate below 3–5 cm in the soil profile. Although snow removal recoupled air and soil temperatures in winter, concrete frost formation during the vernal transition caused treatment soil temperatures to mirror those of the reference plots. However, fluctuations in soil temperature indicate that soils in snow-removal plots thawed 6 days earlier than reference plots. At HBEF and Harvard Forest, snowpack was only removed for part of the winter, and soil frost persisted longer in snow-removal plots (Hardy et al. 2001, Reinmann and Templer 2015). However, at the Proctor Maple Research Center, snow was removed throughout the entire winter (as in this study), and soils warmed earlier in snow-removal plots (Decker et al. 2003) due to a combination of the loss of the thermal barrier of snow and reduced albedo. The fact that concrete frost delayed thawing bare soils is important in Maine, where winter non-snow precipitation events are expected to increase due to climate change (Fernandez et al. 2015). To our knowledge, this study marks the first snow-manipulation experiment in a coniferous forest in the northeastern United States. Trends in soil temperature were similar to those in other New England temperate forest stands, with loss of snowpack leading to greater soil frost which may affect springtime microbial, plant, and macroinvertebrate activity (Campbell et al. 2005). Soil moisture was also influenced by snow removal, but not as we had hypothesized, with wetter soils resulting from concrete frost due to rain on soil during the vernal transition. It is interesting to note that our conclusions could have been different had there been no rain on soil during the vernal transition period for this study. These results highlight the importance of interacting factors and a high variability in space and time that is inherent in a comprehensive understanding of the consequences of a changing climate. Since the trajectory of climate will likely result in greater loss of snow cover with time, it is increasingly important to understand the physical, chemical, and biological consequences of these shifts in ecosystem function to manage and model the cascading forest response to a changing climate. Acknowledgments We thank Dr. Andrew Reeve for lending us iButtons. We thank Cheryl Spencer, Nina Caputo, Tyler Coleman, Justin Libby, Christian Oren, and Lindsey White for their assistance in the field and laboratory. Drs. Sarah Nelson and Kevin Simon provided valuable feedback on experimental design. This research was funded by grants from the University of Maine Graduate Student Government Grant and an Ecology and Environmental Sciences Graduate Research Grant. Support for this research was, in part, from the Maine Agricultural and Forest Experiment Station and the National Science Foundation (DEB-1056692). We thank the 3 anonymous reviewers for their constructive feedback. Finally, we are extremely grateful to our team of shoveling volunteers: Alex Bacjz, Kalyn Bickerman-Martens, Kelsey Boeff, Jesse Call, Meaghan Conway, Julia McGuire, Jules Michaud, and Kaitlyn O’Donnell. This project was supported by the USDA National Institute of Food and Agriculture, Hatch project #ME0- 41507. This is Maine Agricultural and Forest Experiment Station Publication No. 3520. Northeastern Naturalist C. Tatariw, K. Patel, J.D. MacRae, and I.J. Fernandez 2017 52 Vol. 24, Special Issue 7 Literature Cited Alden, P. 1998. National Audubon Society Field Guide to New England. Knopf, New York, NY, USA. Boyce, R., and S. Lucero. 1999. Role of roots in winter water relations of Engelmann Spruce saplings. Tree Physiology 19:893–898. Brooks, P.D., M.W. Williams, and S.K. Schmidt. 1998. Inorganic nitrogen and microbial biomass dynamics before and during spring snowmelt. Biogeochemistry 43:1–15. Campbell, J.L., M.J. Mitchell, P.M. Groffman, L.M. Christenson, and J.P. Hardy. 2005. Winter in northeastern North America: A critical period for ecological processes. Frontiers in Ecology and the Environment 3:314–322. Campbell, J.L., A.B. Reinmann, and P.H. Templer. 2014. Soil-freezing effects on sources of nitrogen and carbon leached during snowmelt. Soil Science Society of America Journal 78:297–308. Chamberlain, E.J., and A.J. Gow. 1979. Effect of freezing and thawing on the permeability and structure of soils. Engineering Geology 13:73–92. Cleavitt, N.L., T.J. Fahey, P.M. Groffman, J.P. Hardy, K.S. Henry, and C.T. Driscoll. 2008. Effects of soil freezing on fine roots in a northern hardwood forest. Canadian Journal of Forest Research 38:82–91. Comerford, D.P., P.G. Schaberg, P.H. Templer, A.M. Socci, J.L. Campbell, and K.F. Wallin. 2013. Influence of experimental snow removal on root and canopy physiology of Sugar Maple trees in a northern hardwood forest. Oecologia 171:261–269. Decker, K.L.M., D. Wang, C. Waite, and T. Scherbatskoy. 2003. Snow removal and ambient air temperature effects on forest soil temperatures in northern Vermont. Soil Science Society of America Journal 67:1234–1243. Fernandez, I.J., C.V. Schmitt, S.D. Birkel, E. Stancioff, A.J. Pershing, J.T. Kelley, J.A. Runge, G.L. Jacobson, and P.A. Mayewski. 2015. Maine’s Climate Future: 2015 Update. University of Maine, Climate Change Institute, Orono, ME. Grimm, N.B., F.S. Chapin, B. Bierwagen, P. Gonzalez, P.M. Groffman, Y. Luo, F. Melton, K. Nadelhoffer, A. Pairis, P. a. Raymond, J. Schimel, and C.E. Williamson. 2013. The impacts of climate change on ecosystem structure and function. Frontiers in Ecology and the Environment 11:474–482. Groffman, P.M., C.D.T. Driscoll, T.J. Fahey, J.P. Hardy, R.D. Fitzhugh, and G.L. Tierney. 2001. Effects of mild winter freezing on soil nitrogen and carbon dynamics in a northern hardwood forest. Biogeochemistry 56:191–213. Groffman, P.M., J.P. Hardy, S. Fashu-Kanu, CT. Driscoll, N.L. Cleavitt, T.J. Fahey, and M.C. Fisk. 2011. Snow depth, soil freezing, and nitrogen cycling in a northern hardwood forest landscape. Biogeochemistry 102:223–238. Groffman, P.M., L.E. Rustad, P.H. Templer, J.L. Campbell, L.M. Christenson, N.K. Lany, A.M. Socci, M.A. Vadeboncoeur, P.G. Schaberg, G.F. Wilson, C.T. Driscoll, T.J. Fahey, M. C. Fisk, C.L. Goodale, M.B. Green, S.P. Hamburg, C.E. Johnson, M.J. Mitchell, J.L. Morse, L.H. Pardo, and N.L. Rodenhouse. 2012. Long-term integrated studies show complex and surprising effects of climate change in the northern hardwood forest. Bio- Science 62:1056–1066. Grundstein, A., P. Todhunter, and T. Mote. 2005. Snowpack control over the thermal offset of air and soil temperatures in eastern North Dakota. Geophysical Research Letters 32:2–5. Hadley, J.L. 2000. Effect of daily minimum temperature on photosynthesis in Eastern Hemlock (Tsuga canadensis L.) in autumn and winter. Arctic, Antarctic, and Alpine Research 32:368–374. Northeastern Naturalist 53 C. Tatariw, K. Patel, J.D. MacRae, and I.J. Fernandez 2017 Vol. 24, Special Issue 7 Hardy, J.P., P.M. Groffman, R.D. Fitzhugh, K. . Henry, A.T. Welman, J.D. Demers, T.J. Fahey, C.T. Driscoll, G.L. Tierney, and S. Nolan. 2001. Snow-depth manipulation and its influence on soil frost and water dynamics in a northern hardwood forest. Biogeochemistry 56:151–174. Hinman, W.C., and F. Bisal. 1968. Alterations of soil structure upon freezing and thawing and subsequent drying. Canadian Journal of Soil Science 48:193–197. Huff, E., and W.H. McWilliams. 2016. Forests of Maine, 2015. US Department of Agriculture, Forest Service, Northern Research Station, Newtown Square, PA. Huntington, T.G., A.D. Richardson, K.J. McGuire, and K. Hayhoe. 2009. Climate and hydrological changes in the northeastern United States: Recent trends and implications for forested and aquatic ecosystems. Canadian Journal of Forest Research 39:199–212. Melillo, J.M., P.A. Steudler, J.D. Aber, K. Newkirk, H. Lux, F.P. Bowles, C. Catricala, A. Magill, T. Ahrens, and S. Morrisseau. 2002. Soil warming and carbon-cycle feedbacks to the climate system. Science 298:2173–2176. Menne, M.J., I. Durre, S.M. Korzeniewski, K. Thomas, X. Yin, S. Anthony, R. Ray, R. . Vose, B.E. Gleason, and T.G. Houston. 2012a. Global Historical Climatology Network - Daily (GHCN-Daily), Version 3.22. NOAA National Climatic Data Center. Available online at daily-version-3. Accessed 3 December 2015. Menne, M.J., I. Durre, R.S. Vose, B.E. Gleason, and T.G. Houston. 2012b. An overview of the Global Historical Climatology Network - Daily Database. Journal of atmospheric and oceanic technology 29:897–910. Moser, B.K., and G. R. Stevens. 1992. Homogeniety of variance in the two-sample means test. The American Statistician 46:19-21. National Oceanic and Atmospheric Administration, National Centers for Environmental Information (NOAA). 2016. Climate at a glance: US time series, average temperature. Available From: Accessed 3 December 2015. Öquist, G., and N.P.A. Huner. 2003. Photosynthesis of overwintering evergreen plants. Annual Review of Plant Biology 54:329–355. Peters, G.-J. 2015. userfriendlyscience: Quantitative analysis made accessible. R Foundation for Statistical Computing, Vienna, Austria.. Piatek, K.B., M.J. Mitchell, S.R. Silva, and C. Kendall. 2005. Sources of nitrate in snowmelt discharge: Evidence from water chemistry and stable isotopes of nitrate. Water, Air, and Soil Pollution 165:13–35. Pinheiro, J., D. Bates, S. DebRoy, D. Sarkar, and R.C. Team. 2015. nlme: Linear and Nonlinear Mixed Effects Models. R Foundation for Statistical Computing, Vienna, Austria.. Pittermann, J., and J.S. Sperry. 2006. Analysis of freeze–thaw embolism in conifers. The interaction between cavitation pressure and tracheid size. Plant physiology 140:374–382. Putkonen, J., and G. Roe. 2003. Rain-on-snow events impact soil temperatures and affect ungulate survival. Geophysical Research Letters 30:1–4. R Core Team. 2015. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Reinmann, A. B., and P. H. Templer. 2015. Reduced winter snowpack and greater soil frost reduce live root biomass and stimulate radial growth and stem respiration of Red Maple (Acer rubrum ) trees in a mixed-hardwood forest. Ecosystems 19:129–141. RStudio Team. 2015. RStudio: Integrated Development for R. RStudio, Inc., Boston, MA. Schaberg, P.G. 2000. Winter photosynthesis in Red Spruce (Picea rubens Sarg.): Limitations, potential benefits, and risks. Arctic, Antarctic, and Alpine Research 32:375–380. Northeastern Naturalist C. Tatariw, K. Patel, J.D. MacRae, and I.J. Fernandez 2017 54 Vol. 24, Special Issue 7 Schaberg, P.G., R.C. Wilkinson, J.B. Shane, J.R. Donnelly, and P.F. Cali. 1995. Winter photosynthesis of Red Spruce from three Vermont seed sources. Tree physiology 15:345–50. Shanley, J.B., and A. Chalmers. 1999. The effect of frozen soil on snowmelt runoff at Sleepers River, Vermont. Hydrological Processes 13:1843–1857. Sorensen, P.O., P.H. Templer, and A.C. Finzi. 2016. Contrasting effects of winter snowpack and soil frost on growing-season microbial biomass and enzyme activity in two mixedhardwood forests. Biogeochemistry 128:141–154. Tierney, G.L., T.J. Fahey, M. Peter, J.P. Hardy, R.D. Fitzhugh, and C.T. Driscoll. 2001. Soil freezing alters fine-root dynamics in a northern hardwood forest. Biogeochemistry 56:175–190. United States Environmental Protection Agency. 2013. Climate impacts in the northeast. Available online at Accessed 14 February 2017. Waring, R.H., and J.F. Franklin. 1979. Evergreen coniferous forests of the Pacific Northwest. Science 204:1380–1386. Williams, C.M., H.A.L. Henry, and B.J. Sinclair. 2014. Cold truths: How winter drives responses of terrestrial organisms to climate change. Biological Reviews 90:214–235. Zhang, T. 2005. Influence of seasonal snow cover on the ground thermal regime: An overview. Reviews in Geophysics 43:1–23.