Soil moisture regulates the response of soil respiration to long-term warming in a southern boreal forest

doi:10.21203/rs.3.rs-2860101/v1

Download PDF

Article

Soil moisture regulates the response of soil respiration to long-term warming in a southern boreal forest

https://doi.org/10.21203/rs.3.rs-2860101/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 20 Aug, 2024

Read the published version in Nature Geoscience →

Version 1

posted

You are reading this latest preprint version

The effects of long-term climate warming on soil respiration and its drivers remain unclear in forests, which store approximately 40% of global soil carbon (C). This uncertainty may hinder realistic predictions of the global C cycling feedbacks under future climate warming. To address this knowledge gap, we conducted a long-term (13 years) climate change experiment, B4WarmED, at two southern boreal forest sites in northern Minnesota, USA. Treatments included simultaneous above- and belowground warming (ambient, + 1.7°C, and + 3.3°C) under different rainfall scenarios (100% and 60% of summer rainfall) and contrasting canopy conditions (open and closed). Across all treatments and years, soil respiration increased by 7% and 17% under + 1.7°C and + 3.3°C, respectively, compared to the ambient temperature plots. In all contexts (all warming levels, rainfall amounts, and canopy conditions), warming increased soil respiration more when soils were moist but less (or even decreased soil respiration) when soils were dry. Furthermore, these responses occurred regardless of whether soil moisture deficits were driven by low ambient rainfall, experimental rainfall reduction, canopy removal, or experimental warming. Our results suggest that factors affecting soil moisture should be accounted for when predicting or modeling the response of soil respiration to warming. Otherwise, warming effects on soil respiration, and thus the positive climate feedback, may be overestimated in forests that periodically experience low soil moisture, and especially those that will do so more frequently under future climate.

Earth and environmental sciences/Ecology/Biogeochemistry/Carbon cycle

Earth and environmental sciences/Ecology/Climate-change ecology

Earth and environmental sciences/Ecology/Forest ecology

Soil respiration

climate warming

drought

canopy condition

soil moisture

boreal forests

Soils store more carbon (C) than plants and the atmosphere combined¹. Therefore, a small increase in soil respiration caused by climate warming could significantly elevate the atmospheric CO₂ concentration, and hence accelerate future global warming^2–5. Although climate warming increases both above- and below-ground temperature in the real world⁶, no long-term (e.g. > 10 years) studies have been conducted to determine the effects of simultaneous aboveground and belowground warming on soil respiration and associated drivers in forests. The lack of relevant long-term results from ecologically realistic experiments could impact the accuracy of predictions of global C cycling under future climate change. Yet, such predictions are necessary for future C cycle projections, given that forests store approximately 40% of global soil C⁷. Herein we reported on a field experiment conducted from 2009 to 2021 at two southern boreal forest sites in northern Minnesota, USA that included combined manipulation of warming, canopy condition, and rainfall amount.

Given that global warming can increase evapotranspiration more than precipitation⁸, soil water deficits during the growing season (e.g. rainfall reduction during summer) in boreal forests are likely in the future^9–11. Rainfall reduction can significantly reduce soil microbial activity and plant growth (and thus plant inputs to soil) by decreasing soil moisture^12–14, and thus potentially reduce warming effects on soil respiration^15,16. Canopy condition, which varies temporally and spatially in forests because of phenology, regeneration, and natural and anthropogenic disturbances¹⁷, also can significantly affect microclimate in forests, including soil temperature and moisture,¹⁸ which are important in determining soil respiration rates¹⁹. Because climate warming is happening across forests differing in rainfall and canopy conditions, how warming effects on soil respiration are altered by rainfall amount or canopy condition and what controls these combined effects are important to understand, especially in the long term.

When compared to treatments with ambient rainfall and closed canopy conditions, rainfall reduction and open canopy may lead to lower soil moisture by decreasing water input and increasing evapotranspiration, respectively. Given the significant effect of soil moisture on soil respiration, we hypothesized that when compared to ambient rainfall amount and closed canopy condition, the positive long-term average effect of warming on soil respiration should be smaller under rainfall reduction and open canopy conditions, respectively (Hypothesis 1). We also predicted a significantly less-than-additive interaction between long-term warming and rainfall reduction treatments.

Varying temporal patterns of soil moisture will also influence temporal patterns of warming impacts on soil CO₂ flux. Generally speaking, warming can affect soil respiration by altering soil microbial activity^4,20 and community²¹, root biomass²², and litterfall amount²³. Although the positive effect of warming on soil respiration has been widely found in terrestrial ecosystems²², ambient soil water conditions should regulate short-term fluctuations in warming effects on soil respiration. Specifically, when soil moisture is higher than a threshold that significantly inhibits root growth and soil microbial activity, the positive direct effect of warming on soil respiration (e.g. via increasing soil temperature) likely exceeds the negative indirect effect of warming (e.g. via warming-induced soil drying). Therefore, we expected that warming would increase soil respiration whenever ambient soil moisture is high (Hypothesis 2a). On the other hand, when soil moisture is low, warming would further decrease soil moisture and then inhibit root growth and soil microbial activity, ultimately exacerbating drought effects. As a result, we expected to find insignificant or even negative effects of warming on soil respiration when ambient soil moisture is low (Hypothesis 2b).

Long-term studies are crucial for understanding warming effects on soil carbon cycling because the results observed in the short term may not predict long-term effects well²⁴. For example, in a 26-year experiment (soil warming only) in a humid temperate oak-maple forest, at Harvard Forest, Massachusetts, USA, warming of + 5C initially stimulated rates of soil respiration by as much as 20%; but over a decade this stimulation decreased to near zero, increased again to near 20% over the next 8–9 years, and then over another 6 years decreased to near zero again²¹. This interannual temporal pattern was mainly attributed to changes in soil C pools and microorganisms, but was not explained by variations in annual precipitation. As there are few other long-term studies reporting soil respiration responses to warming in forests, it is unknown whether similar oscillating results would be found in other contexts, including mid-continent southern boreal forests, which are sensitive to even modest climate change (e.g. warming and rainfall reduction)²⁵. The only other long-term forest warming experiment we are aware of, the 16-year Flakaliden warming experiment in spruce forest in northern Sweden²⁶, had data from two periods (years 1–6 and 14–16); although long-term oscillations were impossible to identify, they did also show differences with time – in the first six years soil respiration increased by 21% due to + 5C warming, whereas after ~ 15 years it increased negligibly (by 2%).

It is unknown whether results at Harvard Forest or Flakaliden would have differed if both soils and aboveground plants had been warmed. For example, the potentially greater stimulation of evapotranspiration caused by both above- and below-ground warming in our study site (~ 765 mm mean annual precipitation, MAP)¹⁸ than at either the slightly warmer Harvard Forest (~ 1200 mm MAP)²¹ or slightly cooler Flakaliden sites (~ 600 mm MAP) may have resulted in larger soil water deficits. Therefore, we hypothesized that interannual temporal patterns of warming effects on soil respiration in our study should be largely explained by the variation in annual precipitation and/or soil moisture (Hypothesis 3).

To test the three abovementioned hypotheses, we measured soil CO₂ flux under different warming levels (ambient, + 1.7°C, and + 3.3°C), rainfall amounts (100% vs. ~60% rainfall amount during summer), and canopy conditions (open and closed) from 2009 to 2021. Because rainfall reduction only occurred in the open canopy condition, the study was an incomplete factorial of warming x site x canopy x rainfall, which included two experiments (e.g. warming and canopy condition; warming and rainfall amount). The large number of soil respiration measurements (22,386) across 72 plots at two sites and range of ambient soil moisture conditions (0.13–0.27 m³ H₂O per m³ soil [10th to 90th percentile] from April to November) in our 13-year study provides an unparalleled opportunity to advance understanding of the role of soil moisture in modulating warming effects on soil respiration and their interannual temporal pattern in boreal forests.

Main and interactive effects of warming and canopy or rainfall on soil respiration

Warming generally increased long-term rates of soil respiration (P < 0.001, Fig. 1, Table 1, and Extended Data Fig. 1 and Fig. 2). Specifically, on average across all observations in all years in both experiments, soil respiration increased by 6.5% and 17.2% under + 1.7°C and + 3.3°C, respectively (Extended Data Fig. 3a). The stimulation varied among years from lows of near zero to a high of 17.7% and 34.3% in 2019 for + 1.7°C and + 3.3°C, respectively (Extended Data Fig. 4). On average across all years, there was no difference in soil respiration in open and closed canopies (Table 1); soil respiration was lower by 4.7% in rainfall reduction than ambient rainfall treatment (Extended Data Fig. 3b).

Table 1

Repeated measures linear mixed-effects model analyses of interactive effects of warming, canopy condition, and rainfall reduction on soil respiration.
Warming and canopy experiment				Warming and rainfall experiment
Effect	DF	F	P > F	Effect	DF	F	P > F
Warming	1	159.2	< 0.001	Warming	1	40.7	< 0.001
Canopy	1	0.9	0.37	Rainfall	1	12.8	< 0.001
Year (yr)	12	9.9	< 0.001	Year (yr)	9	19.9	< 0.001
Site	1	4.1	0.07	Site	1	3.2	0.14
Warming × Canopy	1	10.1	< 0.01	Warming × Rainfall	1	4.2	0.04
Warming × yr	12	5.4	< 0.001	Warming × yr	9	4.1	< 0.001
Canopy × yr	11	8.4	< 0.001	Rainfall × yr	9	3.6	< 0.001
Warming × Site	1	1.0	0.32	Warming × Site	1	16.4	< 0.001
Canopy × Site	1	0.0	0.88	Rainfall × Site	1	0.1	0.72
yr × Site	12	3.3	< 0.001	yr × Site	9	2.9	< 0.01
Warming × Canopy × yr	11	4.0	< 0.001	Warming × Rainfall × yr	9	3.3	< 0.001
Warming × Canopy × Site	1	12.9	< 0.001	Warming × Rainfall × Site	1	5.1	0.02
Warming × yr × Site	12	1.9	0.03	Warming × yr × Site	9	3.9	< 0.001
Canopy × yr × Site	11	1.0	0.41	Rainfall × yr × Site	9	2.1	0.03
Warming × Canopy × yr × Site	11	6.0	< 0.001	Warming × Rainfall × yr × Site	9	0.9	0.50
Warming: experimental warming level; Canopy: canopy condition; Rainfall: experimental rainfall amount; Site: field site; F: F ratio. Significant effects (P < 0.05) are highlighted in bold. The measurements of soil respiration from all campaigns were used in the analyses. The values of increased temperature were used to represent the experimental warming level in the analysis (0, 1.7, and 3.3 for ambient, + 1.7°C, and + 3.3°C treatment, respectively); and year, canopy condition, and rainfall amount were nominal factors in the mixed model.

In the “warming and canopy condition” experiment, warming was a main factor affecting soil respiration (Table 1), and the effect of warming on soil respiration varied among years, canopies, and sites (significant interactions, Table 1 and Fig. 1). Similarly, in the “warming and rainfall amount” experiment, warming and rainfall reduction individually and interactively affected soil respiration, and their effects varied with site and year (significant interactions, Table 1).

Whether warming increases or decreases soil respiration may depend in part on the net effect of the direct stimulatory effect and the potential negative effect of soil drying. Our long-term data showed that warming treatment increased soil temperature (Fig. 2 and Extended Data Fig. 5 and decreased soil moisture (Fig. 2 and Extended Data Fig. 6) regardless of canopy condition, rainfall reduction treatment, and site (Extended Data Fig. 7). Since soil respiration showed a greater response to realized soil temperature than soil moisture (Fig. 2), we found overall positive effects of warming on soil respiration (Extended Data Fig. 3a).

As two of the only other long-term studies of soil respiration response to warming in forests, the 26-year warming experiment in the Harvard Forest and 16-year Flakaliden warming experiment in spruce forest in northern Sweden, provide useful context. Those studies, which elevated soil temperature to 5°C above ambient, but did not warm aboveground, found that warming increased soil respiration on average at Harvard Forest by 8.5%²¹ and by 2% in years 14–16 of the experiment at Flakaliden²⁶. In contrast, combined plant and soil warming of + 3.3°C in the present B4WarmED study over all years stimulated approximately two and eight times more soil respiration than + 5.0°C warming of only soils in the Harvard Forest and long-term results from Flakaliden. Moreover, warming of + 1.7°C in B4WarmED resulted in almost as large, and much larger, respectively, average stimulation of soil respiration as + 5.0°C soil warming at Harvard Forest and in years 14–16 at Flakaliden. Moreover, we observed no tendency for warming stimulation of soil respiration averaged across sites and sub-experiment to diminish over time (Extended Data Fig. 4).

Two potential explanations for the much greater effect per unit C warming of combined above and belowground warming are as follows: first, when compared to belowground warming alone, simultaneous above- and belowground warming might further increase plant growth including root biomass, and thus enhance autotrophic root respiration, an important component of soil respiration²⁷; and second, combined above- and belowground warming could increase plant aboveground and belowground litter and exudate inputs, increasing heterotrophic use of plant-derived C. Additionally, although respiration at our site being more responsive per degree warming than at the warmer Harvard Forest site²⁸ is consistent with past evidence that soil organic C decomposition is more sensitive to warming in colder regions, the lesser sensitivity at colder Flakiliden²⁶ is not.

Mid-term (2–6 years) studies in forests also found warming generally increased soil respiration in those early years by more than the longer-term studies after 10 years. In studies in tropical forest in Panama (+ 4°C for 2 years)²⁹, deciduous forest in Japan (+ 3°C for 4 years)³⁰, boreal pine forest in Finland (+ 2 to 4°C for 4 years)³¹, temperate oak forest in China (+ 1.6 to 2.1°C for 5 years)³², mixed coniferous forest in California (+ 4°C for 5 years)³³, and Norway spruce forest in Austria (+ 4°C for 6 years)¹⁶, warming increased soil respiration by 55%, 16%, 35%, 29%, 30%, and 37%, respectively. It should be noted that the differences in climate zone (e.g. boreal, temperate, and tropical zones), warming method (e.g. soil surface warming, whole-soil warming, only aboveground warming, and both above- and below-ground warming) and level, and longevity among these mid-term studies make it difficult to compare their results, and/or their results in relation to the three longer-term studies.

Pathways regulating effects of warming, canopy condition, and rainfall amount on soil respiration

The effects of warming, rainfall reduction, and canopy removal logically might occur at least in part through changes in soil temperature and soil moisture. Thus we used structural equation modeling to assess direct and indirect effects of treatments including via these two pathways (Fig. 2). Warming and open canopy conditions increased soil temperature, ultimately increasing soil respiration (Fig. 2a and 2b); warming and rainfall reduction decreased soil moisture, ultimately decreasing soil respiration (Fig. 2c and 2d). Significant three-way interactions between warming, site, and canopy condition or rainfall reduction (Table 1) were mainly explained by the change in soil moisture (Fig. 2), which was partly consistent with Hypothesis 1. It should be noted, however, that the interactions between warming and canopy condition or rainfall reduction were more complex than proposed in Hypothesis 1, in that warming and canopy condition interacted with site in affecting soil respiration for the “warming and canopy condition” experiment (P < 0.001, Table 1). Generally speaking, soil respiration increased with increasing temperature except for in the open canopy condition at HWRC site, which showed insignificant differences in soil respiration between + 1.7°C and + 3.3°C (Extended Data Fig. 8a). This lack of temperature effect is likely because plots in open canopy condition at HWRC had significantly lower soil moisture (Extended Data Fig. 7a). The negative effects of both warming and open canopy condition on soil moisture led to the lowest soil moisture under + 3.3°C, and thus, the negative effects of soil drying offset the positive effect of warming on soil respiration. However, at CFC, canopy condition did not significantly affect soil moisture (Extended Data Fig. 7a). Therefore, soil respiration was stimulated at both levels of warming in both canopy treatments at CFC. Overall, because of the different effects of canopy condition on soil moisture at the two sites (insignificant difference in soil moisture between open and closed canopy condition at CFC but significantly higher soil moisture under closed than open canopy condition at HWRC, Extended Data Fig. 7a), soil respiration increased with increasing warming level regardless of canopy condition at CFC, but not at HWRC (Extended Data Fig. 8a), which was partly consistent with Hypothesis 1.

A similar mechanism (e.g. warming effects on soil respiration mediated by rainfall reduction effects on soil moisture) appeared to be at work for the “warming and rainfall amount” experiment. For example, significant interactions between warming, rainfall reduction, and site were found (P < 0.05, Table 1). Specifically, when compared to CFC, smaller differences in soil respiration between warming levels were found at HWRC (Extended Data Fig. 8b). The absolute decline in soil moisture caused by rainfall reduction was similar between the two sites (Extended Data Fig. 7b). However, because HWRC had lower ambient soil moisture overall than CFC, rainfall reduction led to much lower levels of soil moisture at HWRC, likely inhibiting soil respiration response to warming more at that site. This interpretation is further supported by the results of SEM analysis, which showed a stronger correlation between soil moisture and respiration at HWRC than CFC (Fig. 2c and 2d). Overall, our findings suggest that canopy condition and rainfall reduction can significantly change warming effects on soil respiration by altering soil moisture.

Role of ambient soil moisture in warming effects on soil respiration

When compared to how warming affects soil respiration by altering soil temperature and moisture, little is known about how fluctuations in ambient soil temperature and moisture (e.g. soil temperature and moisture under ambient warming treatment) modulate warming effects on soil respiration, partly due to the few measurements from long-term studies. Across 13 years of our study, ambient soil moisture and temperature from 22,386 observations ranged from 0.05 m³ m^− 3 to 0.44 m³ m^− 3 and from − 0.14°C to 27.74°C during times when soil respiration was measured, respectively. The large variation in ambient soil moisture and temperature could help us advance our understanding of their roles in mediating warming effects on soil respiration in boreal forests. Generally speaking, when the soil was extremely dry (e.g. 0th − 10th percentile ambient soil moisture), warming effects on soil respiration were minor or even negative, but became positive when ambient soil moisture was higher (Fig. 3), which supports Hypothesis 2. In addition, warming effects on soil respiration remained the same or decreased when the soil was extremely moist (e.g. 90th − 100th percentile) compared to intermediate moisture levels (e.g. 10th − 90th percentile). When soils are very dry, the growth of both plants and soil microorganisms is significantly inhibited^13,14,22. The decrease in soil moisture induced by warming can further exacerbate the water limitation of plants and microorganisms. Therefore, insignificant or even negative effects of warming on soil respiration occurred when ambient soil moisture was extremely low. When soils are very wet, soil microorganisms are oxygen limited^4,34, thus weakening the positive effect of warming on soil respiration. Given that extreme precipitation events (e.g. flooding and prolonged drought) are projected to become more frequent and severe in the future^35,36, our results suggest that warming effects on soil respiration may fluctuate from positive to negative as ambient soil moisture levels vary. By overlooking the negative effect of warming on soil respiration in extremely dry soils, we would overestimate the overall positive effect of warming on soil respiration and thus the positive climate feedback, especially in regions that are projected to experience more drought events in the future.

Long-term interannual temporal pattern of warming effects on soil respiration

Significant interannual variability in soil respiration was found for all treatments at both sites (Figs. 1 and 4). Overall, the response of soil respiration to warming in our experiment exhibited either two- (decrease and then increase) or three-phase (decrease-increase-decrease) patterns (Fig. 4 and Extended Data Fig. 9), similar to the results from the study at Harvard Forest, MA, USA²¹. For example, in closed canopy plots, the effects of both warming levels on soil respiration decreased for the first four years at both sites, then increased in years five and six, and then either increased continuously (at CFC) or showed another decline (at HWRC). Under open canopy condition and ambient rainfall at both sites, the effects of both warming levels were initially positive, generally peaked in year 2, then decreased to neutral or negative with a trough after 4–6 years, then increased to positive values again. For open canopy plots with rainfall reduction, warming effects started out positive and declined at CFC but were small and did not change a lot over time at HWRC. Overall, our findings indicate that the temporal pattern of long-term warming effects on soil respiration in boreal forests is complex, varying significantly with canopy condition, rainfall amount, and site.

To test Hypothesis 3, we determined the relationships between annual ambient soil moisture, temperature, precipitation, and the changes in soil respiration caused by warming. Unlike short-term fluctuations in soil moisture, which were significantly related to the effects of warming on soil respiration across individual measurements, the annual average change in soil respiration with warming treatment was not significantly correlated with annual precipitation, ambient soil moisture, or temperature, under almost all situations. These findings indicate that the long-term interannual patterns of warming effects on soil respiration in boreal forests were not explained by the interannual variations of either precipitation, soil moisture, or temperature, which is inconsistent with Hypothesis 3. Thus, interannual changes in soil biochemical or microbial variables, rather than in weather, likely contributed to the interannual pattern of warming effects on soil respiration²¹.

For the “warming and canopy condition” experiment, warming effects on soil respiration declined for all treatment combinations over the first few years (Fig. 4). This decline might be driven by the depletion of more easily decomposable C under warming over time²¹. The duration of the decrease in warming effects on soil respiration was much shorter under closed than open canopy conditions especially at HWRC, possibly due to the relatively higher soil moisture but lower soil temperature under closed than open canopy conditions (Extended Data Fig. 7a and 7c). The same increase in temperature has been shown to stimulate decomposition rates more in cold than warm regions²⁸; furthermore, higher soil moisture may promote labile C decomposition. Therefore, colder and wetter soils under closed than open canopy conditions might have resulted in a faster labile C decomposition rate, shortening the duration of declining warming effects on soil respiration. Despite no direct evidence from the present study, changes in the soil microbial community as were found in the Harvard Forest²¹ might explain why warming effects on soil respiration oscillated over time in our study. Lower litterfall amount and sparse grass growth under closed canopy condition at HWRC might explain why warming effects on soil respiration decreased again after the decrease-increase pattern under this condition. Overall, our results indicate that the interannual pattern of warming effects on soil respiration varies with canopy condition.

For the “warming and rainfall amount” experiment, the relatively modest temporal changes in warming effects on soil respiration under rainfall reduction than ambient rainfall treatment (Fig. 4) suggest that by causing soil water deficits, rainfall reduction might slow the depletion of soil labile C and changes in soil microbial community caused by long-term warming. Since global climate change is projected to encompass both warming and changes in rainfall amount in the future⁸, the role of rainfall reduction in mediating the temporal pattern of warming effects on soil C cycling should be considered in future experimental and modeling studies.

Global warming is projected to increase both above- and below-ground temperature⁶. To the best of our knowledge, however, no prior studies have reported long-term (e.g. > 10 years) effects of simultaneous above- and below-ground warming on soil respiration in forests. Since forests store approximately 40% of global soil C⁷, this knowledge gap could significantly inhibit understanding of terrestrial C cycling under future climate change. Herein, we conducted a 13-year field experiment in southern boreal forests and found that soil moisture played a large role in mediating warming effects on soil respiration under different canopy conditions and experimentally manipulated temperature and summer rainfall conditions. For example, similar to rainfall reduction, warming also decreased soil moisture and thus dampened the stimulation of soil respiration, ultimately offsetting some of the positive effects of warming on soil respiration. Although the effect of warming on soil respiration was generally positive and high under the intermediate to moist soil water conditions, it became modest or even negative in extremely dry soils, since warming further aggravated soil water limitation. Overlooking the significant role of soil moisture in mediating global warming effects on soil respiration will lead to inaccurate predictions of forest C cycling under future climate change, especially in regions that often experience periodic drought and/or excessive moisture conditions.

Site description and experimental design

The Boreal Forest Warming at an Ecotone in Danger (B4WarmED) experiment was conducted at two field stations of the University of Minnesota: the Cloquet Forestry Center (CFC), Cloquet, MN, USA (46° 40’ 46’’ N, 92° 31’ 12’’ W, 382 m a.s.l.) and approximately 150 km further north, the Hubachek Wilderness Research Center (HWRC), Ely, MN, USA (47° 56’ 46’’ N, 91° 45’ 29’’ W, 415 m a.s.l.). Both sites are located in the ecotone that spans the transition from temperate to boreal biomes, and the dominant overstory species at the time of experimental establishment were 40- to 60-year-old aspen (Populus tremuloides), birch (Betula papyrifera), and fir (Abies balsamea) with intermixed individual old red and white pine (Pinus resinosa and P. strobus). Both two sites have post-glacial coarse-textured alluvial soils, and the soil texture is sandy loam with 10% clay content. The experiment was an incomplete factorial of warming x site x canopy x rainfall, since rainfall reduction only occurred in the open canopy condition. Thus, we analyzed our data as two distinct full factorial experiments as follows.

At both sites, simultaneous above- and below-ground warming treatments at three levels (ambient, + 1.7°C and + 3.3°C; with the latter two with respect to unwarmed plots in the same canopy and rainfall treatments) were accomplished through a chamberless infrastructure (infrared lamp heaters and buried soil heating cables)^18,37. Warming was implemented approximately from April to November between 2009 and 2021 except for 2014 under closed canopy condition and 2017 under open canopy condition. Across all treatments (e.g. canopy condition, rainfall reduction, and sites) and years, above- and below-ground temperature was increased by 1.7 ± 0.2°C and 3.3 ± 0.3°C under + 1.7°C and + 3.3°C treatments, respectively, compared to the ambient temperature treatment. Initially, the experiment had a 2 site (CFC and HWRC) × 2 canopy condition (open and closed) × 3 warming (ambient, + 1.7°C, and + 3.3°C) factorial design with 3 blocks, each containing 2 replicate plots of each treatment combination, for a total of 72 research plots until 2012. In 2012, rainout shelters were installed in half of the open canopy plots (18 plots), enabling two rainfall levels (100% rainfall [ambient] vs. ~ 60% rainfall from June to September [rainfall reduction]). The simulated summer rainfall reduction was achieved by manually extending rainout shelters mounted 4 m above the plots before scheduled rain events^37,38. Therefore, from 2012 onward, we consider a second factorial experiment that contrasts two rainfall levels (e.g. 2 sites × 3 warming levels × 2 rainfall amounts x 3 blocks for a total of 36 plots). For this experiment, to show the interannual patterns of soil respiration, soil moisture, soil temperature, and the change in soil respiration caused by warming under different warming levels, the results from all plots before 2012 were used for both ambient rainfall and rainfall reduction treatments (Fig. 1, Extended Data Figs. 5, 6, and 9). Therefore, the results before 2012 were identical to each other between ambient rainfall and rainfall reduction treatments. Rainfall reduction treatment was not implemented in 2017 at either site.

Prior to initiating the warming treatments in 2008, blocks designated as open canopy condition were harvested, and stumps were cut to less than 30 cm height. In addition, all blocks were brush cut by hand to remove tree saplings and shrubs, and 11,616 seedlings of 11 species including six native broadleaf (Acer rubrum, Acer saccharum, Betula papyrifera, Populus tremuloides, Quercus macrocarpa, and Quercus rubra), four native needle-leaved (Abies balsamea, Picea glauca, Pinus banksiana, and Pinus strobus), and one naturalized broadleaf (Rhamnus cathartica) species were planted. Over the years of the experiment, we have grown multiple cohorts of juveniles of 21 tree species in different mixtures (Supplementary Table 1). For each cohort, we sourced approximately 2-3-year-old bare root stock seedlings of all native species from the Minnesota Department of Natural Resources (MN DNR), while invasive species were sourced from the local MN DNR nursery and/or parks and planted intermixed with the naturally occurring shrub, herb, and fern vegetation in each plot.

Measurements and analyses of soil respiration, temperature, and moisture

In 2008, two PVC collars (10.2 cm diameter) were randomly inserted approximately 2 cm into the soil in each plot to account for heterogeneity in tree species locations. From 2009, in situ soil respiration was measured approximately every two weeks when warming treatments were operational (roughly from April to November) using a LICOR 6400 with a soil chamber attached (LI-COR Biosciences, Inc., Lincoln, NE, USA). Before each measurement of soil respiration, living plants and litterfall inside the PVC collar were removed. In each plot, two soil thermocouple probes were installed at 10 cm depth to continuously monitor soil temperature, and one time-domain reflectometer probe (CS-616, Campbell Scientific Inc., Logan, UT, USA) was installed at depth 0–22.5 cm to continuously monitor volumetric soil moisture. We used soil temperature logged at 15-minute intervals and soil moisture measured once per hour in all plots throughout the whole year. Soil respiration, temperature, and moisture were not measured under closed canopy condition in 2014 because warming was not implemented. For statistical analyses, we matched the timing of the soil respiration measurements to the nearest interval for which soil moisture and temperature measurements were logged. To measure rainfall, a tipping bucket rain gauge (MET-One model 385AC, Campbell Scientific Inc., Logan, UT, USA) was installed at each site.

Statistical analysis

Repeated-measures linear mixed-effects models were used to test the responses of soil respiration to treatment combinations, year, and site. Year and warming treatment were used as continuous variables, and block was added to each model as a random effect. More specifically, for the “warming and canopy condition” experiment, measurements of soil respiration from 2009 to 2021 were included in the analysis (from 2012, only data in ambient rainfall plots were used). For the “warming and rainfall amount” experiment, measurements of soil respiration in open canopy plots from 2012 to 2021 were included in the analysis. A structural equation model (SEM) was performed to determine how warming, canopy removal, and rainfall reduction affect soil respiration by altering soil temperature and moisture. We did not include the measured soil respiration data during years when warming or rainfall reduction was not implemented (1.9% of total observations) to calculate the overall change in soil respiration caused by warming or rainfall reduction over the entire experiment (Extended Data Figs. 3 and 4). For all other analyses, we used the measured soil respiration data during years when warming or rainfall reduction was not implemented. Overall, whether including the measured soil respiration data during years when warming or rainfall reduction was not implemented in the analyses did not significantly affect the results.

Acknowledgements

We acknowledge support from the many field assistants who were involved in implementing and maintaining the experimental facility, the experimental planting, and the measurements of soil respiration presented in this paper. This research was supported by the US Department of Energy, Office of Science, and Office of Biological and Environmental Research award number DE-FG02-07ER64456; the National Science Foundation, Biological Integration Institutes grant NSF-DBI-2021898; DOE National Institute for Climate Change Research award; DOE Global Change Education Program; Minnesota Agricultural Experiment Station MN-42-030 and MN-42-060; and the College of Food, Agricultural and Natural Resources Sciences and Wilderness Research Foundation, University of Minnesota.

Author contributions

P.B.R., R.A.M., R.L.R, and S.E.H. conceived and designed the original experiment and the experimental warming and rainfall manipulations, and P.B.R., A.S., and R.L.R. managed the experiment over time. A.S., R.B., and R. L. R. collectively implemented the experiment; A.S., R.B., and W.C.E. supervised or performed the acquisition of all soil respiration data; and A.S., R.B., and R. L. R. acquired all soil temperature and moisture and rainfall data; and curated all data. G.L. and P.B.R. developed the idea for this study. G.L. carried out all analyses with suggestions from P.B.R. and S.E.H. and constructed the Figures and Tables, and wrote the first draft. All authors contributed to the interpretation of the results and were involved in writing and editing subsequent drafts. P.B.R. and R.A.M. were responsible for acquiring the funding for the project, and P.B.R. was responsible for all project supervision and administration.

Competing interests

The authors declare no competing interests.

Scharlemann, J. P. W., Tanner, E. V. J., Hiederer, R. & Kapos, V. Global soil carbon: Understanding and managing the largest terrestrial carbon pool. Carbon Management 5, 81–91 (2014).
Crowther, T. W. et al. Quantifying global soil carbon losses in response to warming. Nature 540, 104–108 (2016).
Bond-Lamberty, B. & Thomson, A. Temperature-associated increases in the global soil respiration record. Nature 464, 579–582 (2010).
Davidson, E. A. & Janssens, I. A. Temperature sensitivity of soil carbon decomposition and feedbacks to climate change. Nature 440, 165–173 (2006).
Schlesinger, W. & Andrews, J. Soil Respiration and the Global Carbon Cycle. Biogeochemistry 48, 7–20 (2000).
Soong, J. L., Phillips, C. L., Ledna, C., Koven, C. D. & Torn, M. S. CMIP5 Models Predict Rapid and Deep Soil Warming Over the 21st Century. J. Geophys. Res. Biogeosci. 125, (2020).
Pan, Y. et al. A Large and Persistent Carbon Sink in the World’s Forests. Science 333, 988–993 (2011).
Sherwood, S. & Fu, Q. A Drier Future? Science 343, 737–739 (2014).
Seager, R. et al. Dynamical and Thermodynamical Causes of Large-Scale Changes in the Hydrological Cycle over North America in Response to Global Warming*. Journal of Climate 27, 7921–7948 (2014).
Wang, Y., Hogg, E. H., Price, D. T., Edwards, J. & Williamson, T. Past and projected future changes in moisture conditions in the Canadian boreal forest. The Forestry Chronicle 90, 678–691 (2014).
Way, D. A., Crawley, C. & Sage, R. F. A hot and dry future: warming effects on boreal tree drought tolerance. Tree Physiology 33, 1003–1005 (2013).
Ma, Z. et al. Regional drought-induced reduction in the biomass carbon sink of Canada’s boreal forests. Proc. Natl. Acad. Sci. U.S.A. 109, 2423–2427 (2012).
Schimel, J. P. Life in dry soils: Effects of drought on soil microbial communities and processes. Annual Review of Ecology, Evolution, and Systematics 49, 409–432 (2018).
Zhou, Z., Wang, C. & Luo, Y. Response of soil microbial communities to altered precipitation: A global synthesis. Global Ecology and Biogeography 1–16 (2018) doi:10.1111/geb.12761.
Zhou, L. et al. Interactive effects of global change factors on soil respiration and its components: a meta-analysis. Global change biology 22, 3157–3169 (2016).
Schindlbacher, A. et al. Soil respiration under climate change: prolonged summer drought offsets soil warming effects. Glob Change Biol 18, 2270–2279 (2012).
Frelich, L. E. & Reich, P. B. Spatial Patterns and Succession in a Minnesota Southern-Boreal Forest. Ecological Monographs 65, 325–346 (1995).
Rich, R. L. et al. Design and performance of combined infrared canopy and belowground warming in the B4WarmED (Boreal Forest Warming at an Ecotone in Danger) experiment. Global Change Biology 21, 2334–2348 (2015).
Hursh, A. et al. The sensitivity of soil respiration to soil temperature, moisture, and carbon supply at the global scale. Glob Change Biol 23, 2090–2103 (2017).
Chen, J. et al. Differential responses of carbon-degrading enzyme activities to warming: Implications for soil respiration. Glob Change Biol 24, 4816–4826 (2018).
Melillo, J. M. et al. Long-term pattern and magnitude of soil carbon feedback to the climate system in a warming world. Science 358, 101–105 (2017).
Song, J. et al. A meta-analysis of 1,119 manipulative experiments on terrestrial carbon-cycling responses to global change. Nature Ecology and Evolution 3, 1309–1320 (2019).
Giardina, C. P., Litton, C. M., Crow, S. E. & Asner, G. P. Warming-related increases in soil CO2 efflux are explained by increased below-ground carbon flux. Nature Clim Change 4, 822–827 (2014).
Leuzinger, S. et al. Do global change experiments overestimate impacts on terrestrial ecosystems? Trends in Ecology and Evolution 26, 236–241 (2011).
Reich, P. B. et al. Even modest climate change may lead to major transitions in boreal forests. Nature 608, 540–545 (2022).
Lim, H. et al. Boreal forest biomass accumulation is not increased by two decades of soil warming. Nat. Clim. Chang. 9, 49–52 (2019).
Bond-Lamberty, B., Wang, C. & Gower, S. T. A global relationship between the heterotrophic and autotrophic components of soil respiration? Global Change Biology 10, 1756–1766 (2004).
Carey, J. C. et al. Temperature response of soil respiration largely unaltered with experimental warming. Proc. Natl. Acad. Sci. U.S.A. 113, 13797–13802 (2016).
Nottingham, A. T., Meir, P., Velasquez, E. & Turner, B. L. Soil carbon loss by experimental warming in a tropical forest. Nature 584, 234–237 (2020).
Noh, N. J., Kuribayashi, M., Saitoh, T. M. & Muraoka, H. Different responses of soil, heterotrophic and autotrophic respirations to a 4-year soil warming experiment in a cool-temperate deciduous broadleaved forest in central Japan. Agricultural and Forest Meteorology 247, 560–570 (2017).
Niinistö, S. M., Silvola, J. & Kellomäki, S. Soil CO2 efflux in a boreal pine forest under atmospheric CO2 enrichment and air warming. Global Change Biology 10, 1363–1376 (2004).
Wang, Y. et al. Microbe-mediated attenuation of soil respiration in response to soil warming in a temperate oak forest. Science of The Total Environment 711, 134563 (2020).
Soong, J. L. et al. Five years of whole-soil warming led to loss of subsoil carbon stocks and increased CO2 efflux. Sci. Adv. 7, eabd1343 (2021).
Yan, Z. et al. A moisture function of soil heterotrophic respiration that incorporates microscale processes. Nat Commun 9, 2562 (2018).
Kirchmeier-Young, M. C. & Zhang, X. Human influence has intensified extreme precipitation in North America. Proc. Natl. Acad. Sci. U.S.A. 117, 13308–13313 (2020).
Ummenhofer, C. C. & Meehl, G. A. Extreme weather and climate events with ecological relevance: a review. Phil. Trans. R. Soc. B 372, 20160135 (2017).
Stefanski, A., Bermudez, R., Sendall, K. M., Montgomery, R. A. & Reich, P. B. Surprising lack of sensitivity of biochemical limitation of photosynthesis of nine tree species to open-air experimental warming and reduced rainfall in a southern boreal forest. Global Change Biology 26, 746–759 (2020).
Bermudez, R., Stefanski, A., Montgomery, R. A. & Reich, P. B. Short- and long-term responses of photosynthetic capacity to temperature in four boreal tree species in a free-air warming and rainfall manipulation experiment. Tree Physiology 41, 89–102 (2021).

There is NO Competing Interest.

Download PDF

Journal Publication

published 20 Aug, 2024

Read the published version in Nature Geoscience →

Version 1

posted

You are reading this latest preprint version

Soil moisture regulates the response of soil respiration to long-term warming in a southern boreal forest

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Results and Discussion

Main and interactive effects of warming and canopy or rainfall on soil respiration

Pathways regulating effects of warming, canopy condition, and rainfall amount on soil respiration

Role of ambient soil moisture in warming effects on soil respiration

Long-term interannual temporal pattern of warming effects on soil respiration

Conclusions

Methods

Site description and experimental design

Measurements and analyses of soil respiration, temperature, and moisture

Statistical analysis

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1