Next Article in Journal
Forest Inventory Attribute Prediction Using Lightweight Aerial Scanner Data in a Selected Type of Multilayered Deciduous Forest
Next Article in Special Issue
Effect of Timber Harvest Intensities and Fertilizer Application on Stocks of Soil C, N, P, and S
Previous Article in Journal
Forest Restoration Using Variable Density Thinning: Lessons from Douglas-Fir Stands in Western Oregon
Previous Article in Special Issue
Evaluation of Whole Tree Growth Increment Derived from Tree-Ring Series for Use in Assessments of Changes in Forest Productivity across Various Spatial Scales
Article Menu
Issue 12 (December) cover image

Export Article

Forests 2016, 7(12), 308; doi:10.3390/f7120308

Article
The Effect of Harvest on Forest Soil Carbon: A Meta-Analysis
School of Environmental and Forest Sciences, University of Washington, Box 352100, Seattle, WA 98195, USA
*
Author to whom correspondence should be addressed.
Academic Editors: Scott X. Chang and Xiangyang Sun
Received: 28 September 2016 / Accepted: 2 December 2016 / Published: 7 December 2016

Abstract

:
Forest soils represent a substantial portion of the terrestrial carbon (C) pool, and changes to soil C cycling are globally significant not only for C sequestration but also for sustaining forest productivity and ecosystem services. To quantify the effect of harvesting on soil C, we used meta-analysis to examine a database of 945 responses to harvesting collected from 112 publications from around the world. Harvesting reduced soil C, on average, by 11.2% with 95% CI [14.1%, 8.5%]. There was substantial variation between responses in different soil depths, with greatest losses occurring in the O horizon (−30.2%). Much smaller but still significant losses (−3.3%) occurred in top soil C pools (0–15 cm depth). In very deep soil (60–100+ cm), a significant loss of 17.7% of soil C in was observed after harvest. However, only 21 of the 945 total responses examined this depth, indicating a substantial need for more research in this area. The response of soil C to harvesting varies substantially between soil orders, with greater losses in Spodosol and Ultisol orders and less substantial losses in Alfisols and Andisols. Soil C takes several decades to recover following harvest, with Spodosol and Ultisol C recovering only after at least 75 years. The publications in this analysis were highly skewed toward surface sampling, with a maximum sampling depth of 36 cm, on average. Sampling deep soil represents one of the best opportunities to reduce uncertainty in the understanding of the response of soil C to forest harvest.
Keywords:
forest management; harvest; soil carbon; soil order; deep soil; meta-analysis

1. Introduction

Forest ecosystems contain 1240 Pg C [1,2], which represents as much as 80% of aboveground terrestrial C and 70% of all soil organic C [3,4,5]. The relative proportion of forest C found in soils varies among biomes, ranging from roughly 85% of the terrestrial C pool in boreal forests, to 60% in temperate forests, to 50% in tropical rainforests [1,6]. The net balance of soil C in forests relies upon large rates of detrital inputs (61.4 Pg C year−1) and respiratory losses (60 Pg C year−1), which together represent substantial yearly turnover in the soil C pool [7]. By altering the rates of detrital inputs and respiratory outputs in soils, the extent and intensity of forest harvest can have substantial impacts not only on ecosystem function but also on atmospheric chemistry and global climate [6,8,9].
C is one of the principal components of soil organic matter (SOM), a key component of soil that plays an important role in many biological, chemical, and physical properties [10,11,12]. SOM provides a crucial source of energy and nutrients for soil microbes, buffers soil pH, and helps to stabilize soil structure [12,13]. Along with nitrogen and phosphorus, SOM is considered a critical indicator for soil health and quality.
Thus, soil C is an essential component of forest C accounting, yet many models assume that only surface soil responds to forest management and that soil C returns to equilibrium within 20 years after harvest [14]. Recent national or global assessments of forest C lack any mention of mineral soil C [15,16,17], implicitly assuming that soil C remains constant after forest harvest. Furthermore, carbon monitoring programs include soil C inconsistently. For example, the American Carbon Registry [18] and the Verified Carbon Standard [19] do not require or specify protocols for soil C measurements. The Intergovernmental Panel on Climate Change (IPCC) inventory standards [20] assume constant mineral soil C in Tier 1, with an option for inclusion of national soil C inventories only if preferred by a particular agency, and the U.S. Forest Service Inventory and Analysis Program [21] specifically limits soil sampling to 20 cm depth. The inclusion of soil in models of ecosystem C following harvest can have significant effects. For example, in a model of the forest C pool change following intensive bioenergy harvest, Zanchi et al. [22] show that the inclusion of soil increases the C payback period by approximately 25 years when substituting forest bioenergy for coal. Thus, the inclusion or exclusion of soil in ecosystem C models and ecological monitoring programs can have a major impact on forest policy when attempting to mitigate climate change through forest management [14].
Ambiguity about the effect of forest harvesting on soil C has persisted in the literature, likely exacerbated by the inherent spatial and temporal variability in soil measurements that can obscure the results of even the most well-designed studies [23]. By gathering the results from many studies that apply similar treatments, meta-analysis can overcome the high levels of spatial and temporal variability to provide cumulative answers that may not have been evident within individual sites [23,24]. Previous meta-analyses on the effect of harvesting on soil C have found either minimal effects on soil C pools [25] or substantial (30%) loss to O horizon pools with little change to mineral soil C [9]. Variation in soil C response has been shown to significantly differ among soil types and different harvesting strategies [9].
Studies of soil C change due to harvest have historically been strongly biased toward surface sampling [26]. Nave et al. [9] reported a mixed response to harvest in deeper soil (20–100 cm depth), ranging from a slight average decrease (−5%) in studies that reported C pools to a large average increase (+20%) in studies that reported only C concentration. Several recent reviews have highlighted the need for greater sampling of deep soil [26,27,28], especially as the shifting paradigm of SOM research has come to reject the assumption that deep soil C cannot not change on timescales relevant to anthropogenic C emissions [29,30,31]. Resolving the response of deep soil horizons to harvesting is important because these horizons occupy a much greater volume than surface O and A horizons. Even small changes in subsurface C can exacerbate or compensate for changes in surface soil C, and neither the magnitude nor direction of subsoil C change is clear from previous research.
The process of meta-analysis is necessarily cumulative, with each iteration updating previous analyses to further constrain the error in effect size estimates and to extend the scope of analysis. Thus, the objective of our meta-analysis is to update and extend the findings of Nave et al. [9] with respect to five major research questions:
(1)
What is the overall effect of forest harvesting on soil C pools?
(2)
How does the effect of forest harvest on soil C change with soil depth?
(3)
To what extent does the effect of harvesting differ among soil orders?
(4)
Do site pretreatment strategies or increasing harvesting intensity (i.e., whole tree harvest) moderate or accentuate harvesting impacts on soil C?
(5)
How long does soil C take to recover from harvest across different soil types?

2. Materials and Methods

Meta-analysis is a cumulative activity which builds upon previous research and meta-analyses on similar research questions. Our meta-analysis builds upon the work of Nave et al. [9] and Johnson and Curtis [25] by updating their results with studies published between 2008 and 2016. The database published by Nave et al. [9] was independently recreated from each of 75 references. Metadata for each study was verified, and additional metadata such as the sampling depth of each response ratio was gathered. A total of 8 effect sizes differed in our dataset from the Nave et al. [9] database, all of which were either additional data for mineral soils or a split of one effect size into two based upon sampling depth.
To add to this database with studies published between 2008 and 2016, we searched the peer-reviewed literature for relevant studies using the online database Web of Science with combinations of the terms: forest, timber, harvest, logging, soil C, soil organic matter, and management. No climate criteria was used to screen studies. To be included in the meta-analysis, publications had to report both a control as well as harvested treatments. Both pretreatment soil C and unharvested reference stands were considered acceptable controls, and measurements of reference stands were considered the superior control. For forest chronosequence studies, soil C data from the oldest stand was used as the control. A minimum stand age of 30 years was considered acceptable for control stands, although most studies used controls of considerably greater age. Nave et al. [9] found that studies reporting only soil C concentration data yielded different conclusions about the direction of harvest effects than those studies reporting soil C pool data. Consequently, soil C pool data was used in our meta-analysis when both concentration and pool data were available.
We collected potentially useful predictor variables of soil C response from each publication, including soil order, geographic region, and time since harvest (Table 1). Binning of continuous predictor variables (such as precipitation) was carried out in the same intervals as Nave et al. [9] for ease of comparison. Each study was categorized by harvest, residue management, and site preparation strategies. Harvesting technique was categorized as sawlog when only the merchantable bole (stem) was removed from the site or whole tree harvest (WTH) when the tops, limbs, and foliage were removed in addition to the bole. To test the response of soil C at different depths, data from each study was separated into one of five groups: O horizon, top soil (0–15 cm), mid soil (15–30 cm), deep soil (30–60 cm), and very deep soil (60–100+ cm). A sixth group called whole soil was assigned to studies that aggregated mineral soil samples instead of reporting results at separate depths. Several studies aggregated soil data from 0–100 cm, which reduced the number of unique deep and very deep soil observations even though these depths were separately sampled.
Our meta-analysis estimates the magnitude of change in soil C using the ln-transformed response ratio R, which is defined as
ln ( R ) = ln ( X T ¯ X C ¯ )
where X T ¯ is the mean soil C value of treatment (harvested) observations, and X C ¯ is the mean soil C value of control observations for a given set of experimental conditions at a specific site and depth. Multiple response ratios were recorded for each publication, with the number of response ratios (k) depending upon the number of experimental conditions imposed and the number of samples taken by depth. For example, a publication that reports the results of two thinning treatments and two clear-cut treatments at three depth increments (forest floor, top soil, and mid soil) versus a control would yield 12 response ratios. R is a unit-less measure of effect size, which allows comparison among studies that report data in different units [24]. By back transforming ln(R), [( e ( ln ( R ) ) 1 ) × 100 ], mean response ratios can be interpreted as the percentage change in soil C relative to the control. Estimates of the standard deviation and sample size for each X T ¯ and X C ¯ were not available in several publications. Consequently, an accurate estimate of total heterogeneity (QT) for the dataset was not possible. Subsequent partitioning of QT into within- and among-group heterogeneity (QW and QA, respectively) for random and mixed effect models (as is customary for meta-analyses) was not possible [24]. Instead, we used nonparametric resampling techniques (bootstrapping) to estimate confidence intervals around mean effect sizes in an unweighted meta-analysis [9]. Adams et al. [32] recommend bootstrapping confidence intervals for ecological meta-analyses, and show that confidence bounds based on this method are more conservative than standard meta-analyses. Bootstrapping was implemented using the bootES package [33] in R [34]. For all statistical tests in our analysis, α = 0.05.
Although not exhaustive, the database we compiled from the literature search contained 945 soil C response ratios from 112 publications published between 1979 and 2016. Roughly half the dataset was comprised of response ratios analyzed by Nave et al. [9]. The full dataset is available as Supplementary Material, including maximum sampling depth and the number of response ratios from each paper (Appendix A).

3. Results

3.1. Overall Effect and Change with Depth

Across all studies, harvesting led to a significant average decrease in soil C of 11.2% relative to control (Figure 1). Whether the response to harvest was reported as pools or concentrations had a large impact on the estimated effect of harvest on soil C, with mean response for studies reporting C concentration units (%, mg·g−1, etc.) 16.2% higher (with a 95% CI [20.9%, 11.8%]) than studies reporting C pool units (Mg·ha−1, tons·ha−1, etc.). Concentration responses are higher than pool responses at all soil depths, except for very deep and whole mineral soil, which did not have enough concentration response ratios to construct separate confidence intervals (Figure 1). Consequently, all subsequent analyses focused on the subset of data reporting soil C pools.
Several different soil layers show significant losses of C due to harvesting. Overall, O horizons lost 30.2% of their carbon as a result of harvesting. Losses from top soil were much smaller, although the estimated loss when reported in pool units was significant (−3.3%). In mid (15–30 cm) and deep soil (30–60 cm), the average loss of soil C was greater than topsoil, although the smaller number of response ratios for these depths resulted in more poorly constrained estimates. Studies only reporting C concentration observed a 14.5% increase in deep soil (30–60 cm), although the sample size was relatively small. The overall effect in very deep soil (60–100+ cm) was significant, with an average loss of 17.7%. Unfortunately, this region of the profile was not frequently sampled (21 response ratios out of 945 total), and consequently the 95% confidence interval is quite wide.

3.2. Effect of Harvesting across Soil Orders

The effect of harvesting on soil C differs between soil orders (Figure 2). For the Alfisols and Inceptisols, there are significant losses in O horizon C pools (−12.0% and −45.4%, respectively), but no significant loss in the mineral soil. Mollisols lost an average of 17.7%, although neither O horizon nor mineral soil responses were significantly different from 0. In several cases, small samples sizes made separate testing of organic and mineral soil impossible within a single order (Andisols, Entisols, Oxisols). However, in each of these cases the overall effect was significant. Soil C increased by 24.5% on average in Andisols, but decreased by 18.8% in Entisols and 30.9% in Oxisols. The number of response ratios was more concentrated in the Alfisol, Inceptisol, Spodosol, and Ultisol orders, although a large number of publications did not report information on soil classification. The response to harvesting in Spodosols is substantial (−19.0% overall), with significant losses in both the O horizon (−36.4%) and moderately less in the mineral soil (−9.1%). Likewise, Ultisols lost significant soil C in response to harvesting (−24.7% overall), with the most substantial losses occurring in the O horizon (−66.0%) rather than in the mineral soil (−11.9%).

3.3. Differences in Response to Harvest between Forest Types

The response of soil C to harvest differs between hardwood and coniferous/mixed forest types (Figure 3). The decline in O horizon C pools is significantly greater in conifer/mixed forests (−38.1%) compared to hardwood forests (−25.4%). Differences between forest types were not significant for any mineral soil layer. However, the decline in soil C after harvest was significant for hardwood forests but not conifer/mixed forests in deep soil (30–60 cm) and in studies reporting whole mineral soil C pools. Also in these studies, the difference between hardwood and conifer/mixed forest response to harvest is marginally significant (p < 0.1). The number of observations are highly concentrated in O horizon and top soil, consequently limiting the precision of mean effect size estimates in deeper layers. No observations for hardwood forest were made in very deep soil (60–100+ cm).

3.4. Harvest Intensity, Residue Management and Site Pretreatment

Differences in forest management strategies can significantly impact the response of soil C to harvesting (Figure 4). While there was no significant overall difference observed between thinning and clear-cut harvesting, less C was lost from mineral soils under clear-cut harvesting compared to thinning (+9.3%). Likewise, harvest intensity significantly changed the response of mineral soil C, with soils undergoing whole tree harvesting losing 13.3% less C than bole-only harvesting. Possible mechanisms for these counter-intuitive results are considered in Discussion Section 4.5.
The practice of broadcast burning sites in preparation for planting after a harvest leads to significant additional losses of soil C, with burned soils losing 15.2% more C than soils with no pretreatment. This effect is especially severe in the O horizon (40.9% additional loss than if sites were not burned), and somewhat curtailed in the mineral soil (8.3% additional loss). The wide 95% CI for the estimate of differences in O horizon responses due to burning reflects disparities in burn severity and treatment implementation among different studies.
Spreading of residual materials across harvested sites (by chipping tops and limbs or other methods) resulted in significant additional loss of soil C (−10.9%), with these extra losses occurring mostly in the mineral soil (−17.5%). On the other hand, residue removal resulted in no significant additional losses to soil C.
Tillage is sometimes used to prepare soils for planting after harvest, either to create raised planting beds or to prepare the soil seed bed. This intensive style of site preparation did not result in significant losses in soil C, especially in the mineral soil. However, very large losses were reported in the O horizon (mean effect = −37.1%) with a very wide confidence interval due to a small number of observations. Additional study of the effect of tillage would help to reduce this error.

3.5. Recovery of Soil C after Harvest

The recovery time for soil C following harvest differs among soil orders (Figure 5). Only 4 soil orders contained enough observations over time to model recovery times: Alfisols, Inceptisols, Spodosols, and Ultisols. We modeled time as a second degree polynomial (Time + Time2) separately for O horizons and mineral soils for each soil order (Table 2).

4. Discussion

4.1. Overall Effect of Harvesting on Soil C

Our results reveal that across many publications in the literature there is a significant loss of soil C in response to harvest (−11.2% overall, −14.4% for studies reporting C pools). This estimate is slightly greater than that found by Nave et al. [9], who reported −8% change relative to control. The difference between these estimates derives from additional losses reported in mineral soil, since the effect of harvesting on O horizon C is identical between this study and Nave et al. [9] (−30%). Indeed, while no significant loss of soil C due to harvesting was reported in previous meta-analyses on the subject [9,25], this analysis reveals significant if small losses in various mineral soil layers. Our meta-analysis has roughly double the number of responses than previous meta-analyses on the subject, and consequently has greater statistical power. In particular, this has allowed us to break down the response of mineral soil C to harvest into more depth increments to better characterize how response is moderated or accentuated by depth.

4.2. Depth Distribution of Soil C Response to Harvest

The response of soil C to harvest differs among depths in the soil. O horizons show the most substantial declines (by percentage), although the O horizon is typically a smaller pool of C than mineral soil horizons. Consequently, smaller absolute declines in O horizon C pools can lead to larger response ratios. Forest type significantly alters the response of O horizons to harvesting, with hardwood forests undergoing less drastic losses than conifer and mixed forests (Figure 3). This result is in contrast with Nave et al. [9], who found that conifer O horizon soil C declines significantly less than hardwood forest floors. Coniferous forest litter is thought to be more chemically recalcitrant to decomposition because of higher C/N and lignin/N, as well as slower N mineralization rates [35,36]. The trend for more soil C loss from coniferous forest floors could be due to differences in the harvesting techniques utilized for each forest type. On the other hand, less change in soil C in coniferous forests in deeper mineral soils could suggest that some of the additional loss in O horizon C pools is the result of translocation of C into mineral soil rather than mineralization to CO2. Whatever the case, the mechanism for this difference is not clear and warrants additional study.
In mineral soils, the relative response to harvest is typically less than the O horizon, but this small relative loss might correspond to a larger absolute loss of C in the mineral soil in many forests. The major exception to this pattern are Spodosols, which can contain larger proportions of total soil C in deep, acidic O horizons. Declines in top soil C pools were modest (−3.3%) but still significant (Figure 1). Mean effect size estimates become more negative with soil depth, although these estimates are not significant. The overall estimate of change in very deep soil (60–100+ cm) shows substantial and significant loss of C (−17.7%). This estimate, however, only covers a small number of observations (21) from Spodosol, Ultisol, Alfisol, and Inceptisol soil orders and completely excludes hardwood forests. The lack of observations in deeper soil horizons leads to very wide confidence intervals.
On average, the maximum depth of soil sampled by the publications in this meta-analysis was 35.9 cm (Figure 6). The average depth of sampling for each response ratio in the database is even more surface-skewed at 21.3 cm. Many of the observations down to 100 cm in the literature only report treatment differences for the whole mineral soil profile (0–100 cm), which eliminates any possibility of understanding the relative response of different horizons or depths. The scarcity of observations in deep soil is incongruous with the increasing loss of soil C with depth relative to control observed in this analysis. More important than the magnitude or significance of the harvest response in very deep soil is the conclusion that much greater attention should be paid to deep soil C pools in both individual forest manipulation experiments and broad-scale C inventory.
While soil C in deep soil is much less concentrated than in O and A horizons, subsurface soil represents a much greater volume of soil than surface soil, especially in older/more well developed soil orders like Alfisols, Ultisols, and Oxisols. Some major regions for forestry contain substantial portions of total soil C in deep horizons. For example, 38% of total soil C was below 50 cm and 24.1% below 1 m in production forest soils in the Pacific Northwest [37]. The imprint of biological activity extends many meters into soil, even into the C horizon [38]. Globally, the average maximum rooting depth for trees is ~7 m [39], far outreaching even the deepest observations in this database. Harvesting disrupts the continued growth and turnover of roots extended deep into soil by mature trees, which in turn disturbs the steady state of C cycling in deep soil by changing environmental conditions (temperature, moisture) as well as the type and rate of C inputs. Furthermore, the flush of nitrate and dissolved organic matter that frequently follows harvest [40,41] could prime the breakdown of older, subsurface C by providing a spike in nutrient availability and labile energy sources [31,42,43]. Alteration of aboveground ecosystems can cause changes in subsurface soils. For example, Mobley et al. [44] observed that, over a period of several decades following afforestation of agricultural land, modest C gains in surface soil were more than offset by large losses in soil C below 30 cm. Neither the response of deep soil C to harvest nor the mechanisms for that change have been sufficiently resolved in the literature, and future work to address these questions are necessary.

4.3. Differences in Soil C Response to Harvest among Soil Orders

Substantial variation in response to harvest was observed among soil orders. Several soil orders had very few response ratios (Andisols, Entisols, Mollisols, and Oxisols), which greatly widens confidence intervals. Nonetheless, significant changes in soil C in response to harvest were observed for all four of these orders. Andisols were the only order to show a significant average increase in soil C in response to harvest. This likely stems from Andisols particular mineralogy, which is often characterized by short-range-order minerals like allophane and imogolite [45,46,47]. The capacity for these types of minerals to adsorb organic matter makes Andisol soil C especially resistant to loss after harvesting. Alfisols also appear to be resistant to loss of soil C after harvesting, with relatively small loss in O horizons (−12.0%) being the only significant effect. All other soil orders have significant overall losses in soil C, roughly −20% for Entisols, Inceptisols, Mollisols, Spodosols, and Ultisols. The uneven distribution of observations among soil orders (most response ratios in the database are from Alfisols, Inceptisols, Spodosols, and Ultisols) results in substantial differences in the size of confidence intervals among different orders. Unfortunately, many studies did not report soil taxonomic information, and thus 115 response ratios could not be assigned to a soil order. The lack of studies on Andisols is curious given the importance of these soils to forestry in several regions such as the Pacific Northwest, USA and New Zealand. Several studies on Andisol and other under-represented soil orders were excluded from this analysis because of a lack of appropriate controls.

4.4. Recovery of Soil C after Harvest

Recovery of soil C after harvesting can take several decades [9]. O horizon pools decline more severely than mineral soil pools, especially in the first several decades (Figure 5). In Spodosols, O horizons recovered from harvesting after 60–85 years, while mineral soil recovered over a longer period of 75–100+ years. While the response to harvest was less severe in mineral soils, the longer recovery period implies either lagged response time between forest floor and mineral soils or differences in the decay rate constants leading to longer-term changes in mineral soil C compared to the forest floor. In the case of Alfisols and Inceptisols, soil C in mineral pools increased or stayed the same after harvest, while O horizons declined. However, the observations of harvest effects on Alfisols, Inceptisols, and Ultisols were largely confined to within the first 50 years post-harvest. Consequently, an estimate for the recovery period of soil C pools in these soil orders cannot be assessed with much confidence. Continued observation of existing harvesting experiments in other soil orders must be made to better characterize changes in soil C over time. For Andisols, Entisols, Mollisols, and Oxisols, only a few time points have been documented, and much further study will be necessary to understand recovery of soil C after harvest.
The modeled recovery time has a fairly low adjusted R2 (0.1) and thus a low predictive capacity. Substantial variation in the response to harvest exists within each soil order, reflecting differences in tree species, harvest intensity, and pretreatment strategies, among other factors. Moreover, soil orders are hardly homogeneous, and differences in the response of soil C among lower levels of classification within each order could be as important as order-level differences. Nonetheless, the substantial and significant differences between orders considered in the model suggest that both the resistance of soil C to change and the recovery period of soil C following harvest (resilience) consistently varies among soil types. Compared with 20-year recovery periods assumed by many models [14], our results indicate that soil C recovery takes place over at least triple that time frame for both O horizons and mineral soil in many cases.
While forests >30 years of age were considered acceptable controls for this analysis, the preponderance of data in this meta-analysis show decreases in soil C relative to control at time = 30 years. Consequently, studies that use mature second growth stands barely over this threshold for experimental controls likely underestimate the response of soil C due to harvesting treatments. Depending upon the site conditions and soil order, control stands of at least 50–75+ years since harvest would be recommended, with older stands being more accurate controls.

4.5. The Effect of Harvest Strategies on Soil C

Differences in harvesting and soil pretreatment strategies significantly impact the loss of soil C after harvest. Curiously, despite the greater relative losses of soil C in O horizons, significant differences between harvest intensities and pretreatment strategies were only found in the mineral soil with the exception of broadcast burning (Figure 4). The reduced loss of soil C from mineral soil observed in treatments with greater harvest intensity (+9.3% for clearcut, +13.3% for whole tree harvest) runs counter to the intended effect of these experimental treatments on soil C. One possibility is that increased harvest intensity reduces the quantity of dissolved organic matter and inorganic nutrients leached into the mineral soil, thus reducing the priming [42,48,49] of mineral soil C mineralization through less addition of energy-rich substrates and nutrients. Another possibility is that response of soil C to increased harvest intensity is soil-type specific, and thus an aggregate analysis such as this is subject to bias by unequal sampling of different soil orders. Whatever the case, this dataset cannot identify the specific mechanism(s) driving this difference, and further study is warranted.
Tilling of forest soils prior to planting should intuitively disrupt O horizons to a greater extent than less intensive practices. However, due to the very small number of observations of this practice in the dataset, the large mean treatment effect on soil C was could not be differentiated from 0. By mixing organic material into the surface mineral soil, tilling could increase top soil C in the short term and possibly prime additional breakdown of C over time. In regions where this practice is used, additional research could help to reveal the mechanisms driving change in the soil C of O horizons and mineral following tillage.
Broadcast burning led consistently to additional loss of soil C in both O horizons and mineral soil. The large additional reduction in O horizon C (−40.9%) is expected given that such a treatment is intended to reduce slash on site to facilitate planting. The loss of carbon after harvest extends into deep soil, especially following slash burning (Figure 7). Although there are few observations in very deep soil (60–100+ cm), burning appears to especially exacerbate C losses in this layer. This result is despite the direct effects of fire (such as soil heating and nutrient volatilization) being highly attenuated with depth [50,51]. Levels of mineralized nitrogen (NH4+ and NO3) and soluble sugars spike within the first year following fire, leading to increased microbial biomass N and N leaching loss [52]. Thus, the flush of nutrients and organic matter into deeper mineral soil following post-harvest broadcast burning has the potential to impact soil C dynamics throughout the soil profile. The number of observations in deep and very deep layers is small, and consequently additional research is necessary to better differentiate between harvesting and fire effects in deep soil horizons.

5. Conclusions

We analyzed 945 studies from 112 publications to examine the effect of harvest on forest soil C around the globe. There is a significant overall reduction in forest soil C following harvest that occurs in both the O horizon and mineral soil. Significant variation in the response to harvesting was observed among different soil depths, among soil orders, between overstory forest types, and between different harvest intensities and pretreatment strategies. Broadcast burning, in particular, appears to exacerbate loss of soil C in both organic and mineral horizons following harvest. The recovery period of soil C following harvest depends upon soil type and takes at least 60 years in many production forests. One of the most important findings of this analysis is a significant loss (−17.7%) of soil C following harvest in very deep soil (60–100+ cm). Deep layers of the soil are greatly under-represented in the literature, and consequently, there is great uncertainty around this estimate. Examination of deep soil horizons in existing manipulative forest studies, in new studies, and in C inventory should be a clear objective for future research.

Supplementary Materials

The following are available online at www.mdpi.com/1999-4907/7/12/308/S1, Table S1: Harvest meta-analysis database (Excel file).

Acknowledgments

A huge thank you to Bryan James for assistance extracting data from publications and assembling the database. Thanks also to Lucas Nave and an anonymous reviewer for constructive recommendations on the manuscript. This manuscript was supported by funding from the University of Washington Stand Management Cooperative as well as a US Department of Agriculture McIntire-Stennis Grant.

Author Contributions

J.J. and R.H. conceived and designed the research questions for the meta-analysis; J.J. conducted the literature search; J.J. defined the inclusion criteria for the meta-analysis and extracted data from the publications; J.J. analyzed the data; and J.J. wrote the paper in consultation with R.H.

Conflicts of Interest

The authors declare no conflict of interest. The founding sponsors had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results.

Appendix A. Publications Providing Response Ratios for This Analysis

ReferenceYearkMax Depth (cm)Time Since Harvest a (years)Location
Alban and Perala [53]199275035MN, USA
Bauhus et al. [54]20046409Germany
Bisbing et al. [55]2010610040MT, USA
Black and Harden [56]1995152023CA, USA
Boerner et al. [57]20064102SC, USA
Borchers and Perry [58]199241514OR, USA
Bravo-Oviedo et al. [59]201583015Spain
Cade-Menun et al. [60]200012265BC, Canada
Carter et al. [61]20028152LA, TX, USA
Chatterjee et al. [62]2009195421WY, USA
Chen et al. [63]20132410029China
Chiti et al. [64]20162410024Ghana, Cameroon, Gabon
Christophel et al. [65]201363015Germany
Christophel et al. [66]2015183033Germany
Cromack et al. [67]1999110010OR, USA
Dai et al. [68]200137014NH, USA
DeByle et al. [69]19801053WY, USA
Deluca and Zouhar [52]2000685MT, USA
Diochon et al. [70]2009285035NS, Canada
Edmonds and McColl [71]19894203Australia
Edwards and Ross-Todd [72]19836451TN, USA
Elliott and Knoepp [73]20053153NC, USA
Ellis et al. [73]19824102Tasmania
Ellis and Graley [74]19832101Tasmania
Esquilin et al. [75]200811014CO, USA
Falsone et al. [76]201231305Russia
Fraterrigo et al. [77]200511530NC, USA
Frazer et al. [78]199041412CA, USA
Gartzia-Bengoetxea et al. [79]20092510Spain
Gillon et al. [80]1999201France
Goh and Phillips [81]19914602New Zealand
Goodale and Aber [82]200121085NH, USA
Gough et al. [83]2007158041MI, USA
Grady and Hart [84]200621512AZ, USA
Grand and Lavkulich [85]2012680 BC, Canada
Gresham [86]200263010SC, USA
Griffiths and Swanson [87]200131020OR, USA
Gundale et al. [88]20054103MT, USA
Gupta and DeLuca [89]201212505Wales
Hart et al. [90]20062151AZ, USA
Hendrickson and Chattarpaul [91]19896203ON, Canada
Herman et al. [92]2003298CA, USA
Holscher et al. [93]200122022Germany
Hwang and Son [94]20062302Korea
Jang and Page-Dumroese [95]201583038MT, USA
Johnson [96]19913203NH, USA
Johnson and Todd [97]199864515TN, USA
Johnson [98]199512 7NH, USA
Johnson et al. [99]199714536NH, USA
Johnson et al. [100]20144601CA, USA
Jones et al. [101]2011123015New Zealand
Kaye and Hart [102]19982151AZ, USA
Keenan et al. [103]19941204BC, Canada
Kelliher et al. [104]200445022OR, USA
Kishchuk et al. [105]2014476AB, Canada
Klockow et al. [106]20139201MN, USA
Klopatek [107]200262030WA, USA
Knoepp and Swank [108]199743033NC, USA
Korb et al. [109]20041101AZ, USA
Kraemer and Hermann [110]197921026WA, USA
Kurth et al. [111]201472308MI, MN, USA
Laiho et al. [112]20035225NC, LA, USA
Latty et al. [113]200421590NY, USA
Law et al. [114]2001310021OR, USA
Law et al. [115]2003910062OR, USA
Leduc and Rothstein [116]20071105MI, USA
Maassen and Wirth [117]200425 Germany
Mattson and Smith [118]1993301011WV, USA
Mattson and Swank [119]19898605NC, USA
May and Attiwill [120]20032105Australia
McLaughlin and Phillips [121]200625017ME, USA
McKee et al. [122]201386024AL, USA
McLaughlin [123]199610505MI, USA
Merino and Edeso [124]19996151Spain
Moreno-Fernandez et al. [125]2015545060Spain
Mu et al. [126]201318505China
Murphy et al. [127]200620601CA, USA
Neher et al. [128]20033202NC, USA
Norris et al. [129]20091510016SK, Canada
O’Brien et al. [130]200365018Australia
Powers et al. [131]2011203013MN, WI, USA
Prest et al. [132]201455035NS, Canada
Prietzel et al. [133]2004401WA, USA
Puhlick et al. [134]201610100 ME, USA
Rab [135]19968101Australia
Riley and Jones [136]20033101SC, USA
Roaldson et al. [137]201416205CA, USA
Rothstein and Spaulding [138]2010630 MI, USA
Sanchez et al. [139]200761052SC, USA
Sanscrainte et al. [140]2003470 WA, USA
Saynes et al. [141]20128511Mexico
Selig et al. [142]200833014VA, USA
Shelburne et al. [143]20044101SC, USA
Sheng et al. [144]201551008China
Skovsgaard et al. [145]200612300Denmark
Slesak et al. [146]201212605OR, WA, USA
Small and McCarthy [147]20053107OH, USA
Stone et al. [148]19991151AZ, USA
Stone and Elioff [149]19984305MN, USA
Strong [150]199784018MN, USA
Strukelj et al. [151]201512105QC, Canada
Tang et al. [152]2009126029MI, WI, USA
Trettin et al. [153]2011615011MI, USA
Ussiri and Johnson [154]200715608NH, USA
Vario et al. [155]201466049NH, USA
Vesterdal et al. [156]199590 Denmark
Waldrop et al. [157]2003301CA, USA
Wu et al. [158]201012010China
Xiang et al. [159]20098300China
Yanai et al. [160]200035029NH, USA
Zabowski et al. [161]200822025OR, WA, USA
Zhong and Makeshin [162]200321016Germany
Zummo and Friedland [163]201115603NH, USA
a For chronosequence studies, time since harvest in this table is averaged across all response ratios for that study.

References

  1. Dixon, R.K.; Brown, S.; Houghton, R.A.; Solomon, A.M.; Trexler, M.C.; Wisniewski, J. Carbon Pools and Flux of Global Forest Ecosystems. Science 1994, 263, 185–190. [Google Scholar] [CrossRef] [PubMed]
  2. Prentice, I.C.; Farquhar, G.; Fasham, M.J.R.; Goulden, M.L.; Heimann, M.; Jarmillo, V.J.; Kheshgi, H.S.; Le Quere, C.; Scholes, R.J.; Wallace, D.W.R. The Carbon Cycle and Atmospheric Carbon Dioxide; Cambridge University Press: Cambridge, UK, 2001; pp. 183–237. [Google Scholar]
  3. Batjes, N.H. Total carbon and nitrogen in the soils of the world. Eur. J. Soil Sci. 1996, 47, 151–163. [Google Scholar] [CrossRef]
  4. Jobbagy, E.G.; Jackson, R.B. The vertical distribution of soil organic carbon and its relation to climate and vegetation. Ecol. Appl. 2000, 10, 423–436. [Google Scholar] [CrossRef]
  5. Six, J.; Conant, R.T.; Paul, E.A.; Paustian, K. Stabilization mechanisms of soil organic matter: Implications for C-saturation of soils. Plant Soil 2002, 241, 155–176. [Google Scholar] [CrossRef]
  6. Lal, R. Forest soils and carbon sequestration. For. Ecol. Manag. 2005, 220, 242–258. [Google Scholar] [CrossRef]
  7. Schimel, D.S. Terrestrial ecosystems and the carbon cycle. Glob. Chang. Biol. 1995, 1, 77–91. [Google Scholar] [CrossRef]
  8. Jandl, R.; Lindner, M.; Vesterdal, L.; Bauwens, B.; Baritz, R.; Hagedorn, F.; Johnson, D.W.; Minkkinen, K.; Byrne, K.A. How strongly can forest management influence soil carbon sequestration? Geoderma 2007, 137, 253–268. [Google Scholar] [CrossRef]
  9. Nave, L.E.; Vance, E.D.; Swanston, C.W.; Curtis, P.S. Harvest impacts on soil carbon storage in temperate forests. For. Ecol. Manag. 2010, 259, 857–866. [Google Scholar] [CrossRef]
  10. Schlesinger, W.H.; Bernhardt, E.S. Biogeochemistry: An Analysis of Global Change; Academic Press: Oxford, UK, 2013. [Google Scholar]
  11. Brady, N.C.; Weil, R.R. The Nature and Properties of Soils; Macmillan Publ. Co.: New York, NY, USA, 2002; Volume 13, p. 960. [Google Scholar]
  12. Krull, E.S.; Skjemstad, J.O.; Baldock, J.A. Functions of Soil Organic Matter and the Effect on Soil Properties; Final Rep. CSO00029; Grains Research and Development Corporation: Glen Osmond, South Australia, 2009; p. 128. [Google Scholar]
  13. Six, J.; Paustian, K. Aggregate-associated soil organic matter as an ecosystem property and a measurement tool. Soil Biol. Biochem. 2014, 68, A4–A9. [Google Scholar] [CrossRef]
  14. Buchholz, T.; Friedland, A.J.; Hornig, C.E.; Keeton, W.S.; Zanchi, G.; Nunery, J. Mineral soil carbon fluxes in forests and implications for carbon balance assessments. Glob. Chang. Biol. Bioenergy 2014, 6, 305–311. [Google Scholar] [CrossRef]
  15. Ryan, M.G.; Harmon, M.E.; Birdsey, R.A.; Giardina, C.P.; Heath, L.S.; Houghton, R.A.; Jackson, R.B.; McKinley, D.C.; Morrison, J.F.; Murray, B.C.; et al. A synthesis of the science on forests and carbon for U.S. Forests. Issues Ecol. 2010, 13, 1–16. [Google Scholar]
  16. McKinley, D.C.; Ryan, M.G.; Birdsey, R.A.; Giardina, C.P.; Harmon, M.E.; Heath, L.S.; Houghton, R.A.; Jackson, R.B.; Morrison, J.F.; Murray, B.C.; et al. A synthesis of current knowledge on forests and carbon storage in the United States. Ecol. Appl. 2011, 21, 1902–1924. [Google Scholar] [CrossRef] [PubMed]
  17. Fahey, T.J.; Woodbury, P.B.; Battles, J.J.; Goodale, C.L.; Hamburg, S.P.; Ollinger, S.V.; Woodall, C.W. Forest carbon storage: Ecology, management, and policy. Front. Ecol. Environ. 2010, 8, 245–252. [Google Scholar] [CrossRef]
  18. American Carbon Registry. The ACR Forest Carbon Project Standard v 2.1; Winrock International: Arlington, VA, USA, 2011. [Google Scholar]
  19. Verified Carbon Standard. Agriculture, Forestry and other Land Use (AFOLU) Requirements; VCS Version 3 Requirements Doc., v3.4 2013; Verified Carbon Standard: Washington, DC, USA, 2013. [Google Scholar]
  20. IPCC. 2006 IPCC Guidelines for national greenhouse gas inventories. In Agriculture, Forestry and Other Land Use; Institute for Global Environmental Strategies: Hayama, Japan, 2006; Volume 4, pp. 1–29. [Google Scholar]
  21. Bechtold, W.A.; Patterson, P.L. The Enhanced Forest Inventory and Analysis Program—National Sampling Design and Estimation Procedures. In USDA General Technical Report SRS-80; U.S. Department of Agriculture, Forest Service, Southern Research Station: Asheville, NC, USA, 2005; p. 85. [Google Scholar]
  22. Zanchi, G.; Pena, N.; Bird, N. Is woody bioenergy carbon neutral? A comparative assessment of emissions from consumption of woody bioenergy and fossil fuel. GCB Bioenergy 2012, 4, 761–772. [Google Scholar] [CrossRef]
  23. Homann, P.S.; Bormann, B.T.; Boyle, J.R. Detecting treatment differences in soil carbon and nitrogen resulting from forest manipulations. Soil Sci. Soc. Am. J. 2001, 65, 463–469. [Google Scholar] [CrossRef]
  24. Hedges, L.V.; Gurevitch, J.; Curtis, P.S. The meta-analysis of response ratios in experimental ecology. Ecology 1999, 80, 1150–1156. [Google Scholar] [CrossRef]
  25. Johnson, D.W.; Curtis, P.S. Effects of forest management on soil C and N storage: Meta analysis. For. Ecol. Manag. 2001, 140, 227–238. [Google Scholar] [CrossRef]
  26. Harrison, R.B.; Footen, P.W.; Strahm, B.D. Deep soil horizons: Contribution and importance to soil carbon pools and in assessing whole-ecosystem response to management and global change. For. Sci. 2011, 57, 67–76. [Google Scholar]
  27. Schmidt, M.W.I.; Torn, M.S.; Abiven, S.; Dittmar, T.; Guggenberger, G.; Janssens, I.A.; Kleber, M.; Koegel-Knabner, I.; Lehmann, J.; Manning, D.A.C.; et al. Persistence of soil organic matter as an ecosystem property. Nature 2011, 478, 49–56. [Google Scholar] [CrossRef] [PubMed]
  28. Deb Richter, D.; Yaalon, D.H. “The Changing Model of Soil” Revisited. Soil Sci. Soc. Am. J. 2012, 76, 76–778. [Google Scholar] [CrossRef]
  29. Marin-Spiotta, E.; Gruley, K.E.; Crawford, J.; Atkinson, E.E.; Miesel, J.R.; Greene, S.; Cardona-Correa, C.; Spencer, R.G.M. Paradigm shifts in soil organic matter research affect interpretations of aquatic carbon cycling: Transcending disciplinary and ecosystem boundaries. Biogeochemistry 2014, 117, 279–297. [Google Scholar] [CrossRef]
  30. Lehmann, J.; Kleber, M. The contentious nature of soil organic matter. Nature 2015, 528, 60–68. [Google Scholar] [CrossRef] [PubMed]
  31. Fontaine, S.; Barot, S.; Barre, P.; Bdioui, N.; Mary, B.; Rumpel, C. Stability of organic carbon in deep soil layers controlled by fresh carbon supply. Nature 2007, 450, 277–280. [Google Scholar] [CrossRef] [PubMed]
  32. Adams, D.C.; Gurevitch, J.; Rosenberg, M.S. Resampling tests for meta-analysis of ecological data. Ecology 1997, 78, 1277–1283. [Google Scholar] [CrossRef]
  33. Gerlanc, D.; Kirby, K. bootES: Bootstrap Effect Sizes, R Package, Version 1.2; R Foundation for Statistical Computing: Vienna, Austria, 2015.
  34. R Development Core Team. A Language and Environment for Statistical Computing, R Foundation for Statistical Computing: Vienna, Austria, 2016.
  35. Currie, W.S. Responsive C and N biogeochemistryof the temperate forest floor. Trends Ecol. Evol. 1999, 14, 316–320. [Google Scholar] [CrossRef]
  36. Silver, W.; Miya, R. Global patterns in root decomposition: Comparisons of climate and litter quality effects. Oecologia 2001, 129, 407–419. [Google Scholar] [CrossRef]
  37. James, J.; Devine, W.; Harrison, R.; Terry, T. Deep soil carbon: quantification and modeling in subsurface layers. Soil Sci. Soc. Am. J. 2014, 78, S1–S9. [Google Scholar] [CrossRef]
  38. Richter, D.D.; Markewitz, D. How Deep Is Soil? Bioscience 1995, 45, 600–609. [Google Scholar] [CrossRef]
  39. Canadell, J.; Jackson, R.B.; Ehleringer, J.R.; Mooney, H.A.; Sala, O.E.; Schulze, E.D. Maximum rooting depth of vegetation types at the global scale. Oecologia 1996, 108, 583–595. [Google Scholar] [CrossRef]
  40. Strahm, B.D.; Harrison, R.B.; Terry, T.A.; Flaming, B.L.; Licata, C.W.; Petersen, K.S. Soil solution nitrogen concentrations and leaching rates as influenced by organic matter retention on a highly productive Douglas-fir site. For. Ecol. Manag. 2005, 218, 74–88. [Google Scholar] [CrossRef]
  41. Likens, G.E.; Bormann, F.H.; Johnson, N.M. Nitrification—Importance to nutrient losses from a cutover forested ecosystem. Science 1969, 163, 1205–1206. [Google Scholar] [CrossRef] [PubMed]
  42. Kuzyakov, Y. Priming effects: Interactions between living and dead organic matter. Soil Biol. Biochem. 2010, 42, 1363–1371. [Google Scholar] [CrossRef]
  43. Blagodatskaya, E.; Kuzyakov, Y. Mechanisms of real and apparent priming effects and their dependence on soil microbial biomass and community structure: Critical review. Biol. Fertil. Soils 2008, 45, 115–131. [Google Scholar] [CrossRef]
  44. Mobley, M.L.; Lajtha, K.; Kramer, M.G.; Bacon, A.R.; Heine, P.R.; Richter, D.D. Surficial gains and subsoil losses of soil carbon and nitrogen during secondary forest development. Glob. Chang. Biol. 2015, 21, 986–996. [Google Scholar] [CrossRef] [PubMed]
  45. Soil Survey Staff. Keys to Soil Taxonomy; Department of Agriculture, Natural Resources Conservation Service: Washington, DC, USA, 2014.
  46. Wada, K. Minerals formed and mineral formation from volcanic ash by weathering. Chem. Geol. 1987, 60, 17–28. [Google Scholar] [CrossRef]
  47. Nanzyo, M.; Shoji, S.; Dahlgren, R. Volcanic Ash Soils—Genesis, Properties and Utilization; Elsevier: Amsterdam, The Netherlands, 1993; Volume 21. [Google Scholar]
  48. Blagodatskaya, E.V.; Blagodatsky, S.A.; Anderson, T.H.; Kuzyakov, Y. Priming effects in Chernozem induced by glucose and N in relation to microbial growth strategies. Appl. Soil Ecol. 2007, 37, 95–105. [Google Scholar] [CrossRef]
  49. Kuzyakov, Y.; Friedel, J.K.; Stahr, K. Review of mechanisms and quantification of priming effects. Soil Biol. Biochem. 2000, 32, 1485–1498. [Google Scholar] [CrossRef]
  50. Valette, J.; Gomendy, V.; Marechal, J.; Houssard, C.; Gillon, D. Heat Transfer in the Soil During Very Low-intensity Experimental Fires: The Role of Duff and Soil Moisture Content. Int. J. Wildland Fire 1994, 4, 225–237. [Google Scholar] [CrossRef]
  51. Certini, G. Effects of fire on properties of forest soils: A review. Oecologia 2005, 143, 1–10. [Google Scholar] [CrossRef] [PubMed]
  52. DeLuca, T.H.; Zouhar, K.L. Effects of selection harvest and prescribed fire on the soil nitrogen status of ponderosa pine forests. For. Ecol. Manag. 2000, 138, 263–271. [Google Scholar] [CrossRef]
  53. Alban, D.H.; Perala, D.A. Carbon storage in lake states aspen ecosystems. Can. J. For. Res. 1992, 22, 1107–1110. [Google Scholar] [CrossRef]
  54. Bauhus, J.; Vor, T.; Bartsch, N.; Cowling, A. The effects of gaps and liming on forest floor decomposition and soil C and N dynamics in a Fagus sylvatica forest. Can. J. For. Res. 2004, 34, 509–518. [Google Scholar] [CrossRef]
  55. Bisbing, S.M.; Alaback, P.B.; DeLuca, T.H. Carbon storage in old-growth and second growth fire-dependent western larch (Larix occidentalis Nutt.) forests of the Inland Northwest, USA. For. Ecol. Manag. 2010, 259, 1041–1049. [Google Scholar] [CrossRef]
  56. Black, T.A.; Harden, J.W. Effect of timber harvest on soil carbon storage at Blodgett Experimental Forest, California. Can. J. For. Res. 1995, 25, 1385–1396. [Google Scholar] [CrossRef]
  57. Boerner, R.E.J.; Waldrop, T.A.; Shelburne, V.B. Wildfire mitigation strategies affect soil enzyme activity and soil organic carbon in loblolly pine (Pinus taeda) forests. Can. J. For. Res. 2006, 36, 3148–3154. [Google Scholar] [CrossRef]
  58. Borchers, J.G.; Perry, D.A. The influence of soil texture and aggregation on carbon and nitrogen dynamics in southwest Oregon forest and clearcuts. Can. J. For. Res. 1992, 22, 298–305. [Google Scholar] [CrossRef]
  59. Bravo-Oviedo, A.; Ruiz-Peinado, R.; Modrego, P.; Alonso, R.; Montero, G. Forest thinning impact on carbon stock and soil condition in Southern European populations of P. sylvestris L. For. Ecol. Manag. 2015, 357, 259–267. [Google Scholar] [CrossRef]
  60. Cade-Menun, B.J.; Berch, S.M.; Preston, C.M.; Lavkulich, L.M. Phosphorus forms and related soil chemistry of Podzolic soils on northern Vancouver Island. II. The effects of clear-cutting and burning. Can. J. For. Res. 2000, 30, 1726–1741. [Google Scholar] [CrossRef]
  61. Carter, M.C.; Dean, T.J.; Zhou, M.; Messina, M.G.; Wang, Z. Short-term changes in soil C, N, and biota following harvesting and regeneration of loblolly pine (Pinus taeda L.). For. Ecol. Manag. 2002, 164, 67–88. [Google Scholar] [CrossRef]
  62. Chatterjee, A.; Vance, G.F.; Tinker, D.B. Carbon pools of managed and unmanaged stands of ponderosa and lodgepole pine forests in Wyoming. Can. J. For. Res. 2009, 39, 1893–1900. [Google Scholar] [CrossRef]
  63. Chen, L.-C.; Liang, M.-J.; Wang, S.-L. Carbon stock density in planted versus natural Pinus massoniana forests in sub-tropical China. Ann. For. Sci. 2016, 73, 461–472. [Google Scholar] [CrossRef]
  64. Chiti, T.; Perugini, L.; Vespertino, D.; Valentini, R. Effect of selective logging on soil organic carbon dynamics in tropical forests in central and western Africa. Plant Soil 2015, 399, 283–294. [Google Scholar] [CrossRef]
  65. Christophel, D.; Spengler, S.; Schmidt, B.; Ewald, J.; Prietzel, J. Customary selective harvesting has considerably decreased organic carbon and nitrogen stocks in forest soils of the Bavarian Limestone Alps. For. Ecol. Manag. 2013, 305, 167–176. [Google Scholar] [CrossRef]
  66. Christophel, D.; Höllerl, S.; Prietzel, J.; Steffens, M. Long-term development of soil organic carbon and nitrogen stocks after shelterwood- and clear-cutting in a mountain forest in the Bavarian Limestone Alps. Eur. J. For. Res. 2015, 134, 623–640. [Google Scholar] [CrossRef]
  67. Cromack, K.; Miller, R.E.; Anderson, H.W.; Helgerson, O.T.; Smith, R.B. Soil Carbon and Nutrients in a Coastal Oregon Douglas-Fir Plantation with Red Alder. Soil Sci. Soc. Am. J. 1999, 63, 232–239. [Google Scholar] [CrossRef]
  68. Dai, K.H.; Johnson, C.E.; Driscoll, C.T. Organic matter chemistry and dynamics in clear-cut and unmanaged hardwood forest ecosystems. Biogeochemistry 2001, 54, 51–83. [Google Scholar] [CrossRef]
  69. DeByle, N.V. Harvesting and Site Treatment Influences on the Nutrient Status of Lodgepole Pine Forests in Western Wyoming. In General Technical Report INT-90; USDA Forest Service Intermountain Forest And Range Experimental Station: Ogden, UT, USA, 1979; pp. 137–156. [Google Scholar]
  70. Diochon, A.; Kellman, L.; Beltrami, H. Looking deeper: An investigation of soil carbon losses following harvesting from a managed northeastern red spruce (Picea rubens Sarg.) forest chronosequence. For. Ecol. Manag. 2009, 257, 413–420. [Google Scholar] [CrossRef]
  71. Edmonds, R.L.; McColl, G.J. Effects of forest management on soil nitrogen in Pinus radiata stands in the Australian Capital Territory. For. Ecol. Manag. 1989, 29, 199–212. [Google Scholar] [CrossRef]
  72. Edwards, N.T.; Ross-Todd, B.M. Soil carbon dynamics in a mixed deciduous forest following clear-cutting with and without residue removal. Soil Sci. Soc. Am. J. 1983, 47, 1014–1021. [Google Scholar] [CrossRef]
  73. Elliott, K.J.; Knoepp, J.D. The effects of three regeneration harvest methods on plant diversity and soil characteristics in the southern Appalachians. For. Ecol. Manag. 2005, 211, 296–317. [Google Scholar] [CrossRef]
  74. Ellis, R.; Graley, A.M. Gains and losses in soil nutrients associated with harvesting and burning eucalypt rainforest. Plant Soil 1983, 74, 437–450. [Google Scholar] [CrossRef]
  75. Esquilín, J.; Aida, E.; Stromberger, M.E.; Shepperd, W.D. Soil scarification and wildfire interactions and effects on microbial communities and carbon. Soil Sci. Soc. Am. J. 2008, 72, 111–118. [Google Scholar] [CrossRef]
  76. Falsone, G.; Celi, L.; Caimi, A.; Simonov, G.; Bonifacio, E. The effect of clear cutting on podzolisation and soil carbon dynamics in boreal forests (Middle Taiga zone, Russia). Geoderma 2012, 177–178, 27–38. [Google Scholar] [CrossRef]
  77. Fraterrigo, J.; Turner, M.; Pearson, S.M.; Dixon, P. Effects of past land use on spatial heterogeneity of soil nutrients in southern Appalachian forests. Ecol. Monogr. 2005, 75, 215–230. [Google Scholar] [CrossRef]
  78. Frazer, D.W.; McColl, J.G.; Powers, R.F. Soil Nitrogen Mineralization in a Clearcutting Chronosequence in a Northern California Conifer Forest. Soil Sci. Soc. Am. J. 1990, 54, 1145. [Google Scholar] [CrossRef]
  79. Gartzia-Bengoetxea, N.; González-Arias, A.; Merino, A.; De Arano, I.M. Soil organic matter in soil physical fractions in adjacent semi-natural and cultivated stands in temperate Atlantic forests. Soil Biol. Biochem. 2009, 41, 1674–1683. [Google Scholar] [CrossRef]
  80. Gillon, D.; Houssard, C.; Valette, J.C.; Rigolot, E. Nitrogen and phosphorus cycling following prescribed burning in natural and managed Aleppo pine forests. Can. J. For. Res. 1999, 29, 1237–1247. [Google Scholar] [CrossRef]
  81. Goh, K.M.; Phillips, M.J. Effects of clearfell logging and clearfell logging and burning of a Nothofagus forest on soil nutrient dynamics in South Island, New Zealand—Changes in forest floor organic matter and nutrient status. N. Z. J. Bot. 1991, 29, 367–384. [Google Scholar] [CrossRef]
  82. Goodale, C.L.; Aber, J.D. The Long-term effects of land-use history on nitrogen cycling in northern hardwood forests. Ecol. Appl. 2001, 11, 253–267. [Google Scholar] [CrossRef]
  83. Gough, C.M.; Vogel, C.S.; Harrold, K.H.; George, K.; Curtis, P.S. The legacy of harvest and fire on ecosystem carbon storage in a north temperate forest. Glob. Chang. Biol. 2007, 13, 1935–1949. [Google Scholar] [CrossRef]
  84. Grady, K.C.; Hart, S.C. Influences of thinning, prescribed burning, and wildfire on soil processes and properties in southwestern ponderosa pine forests: A retrospective study. For. Ecol. Manag. 2006, 234, 123–135. [Google Scholar] [CrossRef]
  85. Grand, S.; Lavkulich, L.M. Effects of forest harvest on soil carbon and related variables in Canadian spodosols. Soil Sci. Soc. Am. J. 2012, 76, 1816–1827. [Google Scholar] [CrossRef]
  86. Gresham, C.A. Sustainability of intensive loblolly pine plantation management in the South Carolina Coastal Plain, USA. For. Ecol. Manag. 2002, 155, 69–80. [Google Scholar] [CrossRef]
  87. Griffiths, R.P.; Swanson, A.K. Forest soil characteristics in a chronosequence of harvested Douglas-fir forests. Can. J. For. Res. 2001, 31, 1871–1879. [Google Scholar] [CrossRef]
  88. Gundale, M.J.; DeLuca, T.H.; Fiedler, C.E.; Ramsey, P.W.; Harrington, M.G.; Gannon, J.E. Restoration treatments in a Montana ponderosa pine forest: Effects on soil physical, chemical and biological properties. For. Ecol. Manag. 2005, 213, 25–38. [Google Scholar] [CrossRef]
  89. Das Gupta, S.; DeLuca, T.H. Short-term changes in belowground C, N stocks in recently clear felled Sitka spruce plantations on podzolic soils of North Wales. For. Ecol. Manag. 2012, 281, 48–58. [Google Scholar] [CrossRef]
  90. Hart, S.C.; Selmants, P.C.; Boyle, S.I.; Overby, S.T. Carbon and nitrogen cycling in southwestern ponderosa pine forests. For. Sci. 2006, 52, 683–693. [Google Scholar]
  91. Hendrickson, O.; Chatarpaul, L. Nutrient cycling following whole-tree and conventional harvest in a northern mixed forest. Can. J. For. Res. 2016, 19, 725–735. [Google Scholar] [CrossRef]
  92. Herman, D.J.; Halverson, L.J.; Firestone, M.K. Nitrogen dynamics in an annual grassland: Oak canopy, climate, and microbial population effects. Ecol. Appl. 2003, 13, 593–604. [Google Scholar] [CrossRef]
  93. Hölscher, D.; Schade, E.; Leuschner, C. Effects of coppicing in temperate deciduous forests on ecosystem nutrient pools and soil fertility. Basic Appl. Ecol. 2001, 2, 155–164. [Google Scholar] [CrossRef]
  94. Hwang, J.; Son, Y. Short-term effects of thinning and liming on forest soils of pitch pine and Japanese larch plantations in central Korea. Ecol. Res. 2006, 21, 671–680. [Google Scholar] [CrossRef]
  95. Jang, W.; Page-Dumroese, D.S.; Keyes, C.R. Long-term soil changes from forest harvesting and residue management in the Northern Rocky Mountains. Soil Sci. Soc. Am. J. 2016. [Google Scholar] [CrossRef]
  96. Johnson, C.E. Whole-tree clear-cutting effects on soil horizons and organic matter pools. Soil Sci. Soc. Am. J. 1991, 55, 497–502. [Google Scholar] [CrossRef]
  97. Johnson, D.; Todd, D.E. Harvesting effects on long-term changes in nutrient pools of mixed oak forest. Soil Sci. Soc. Am. J. 1998, 62, 1725–1735. [Google Scholar] [CrossRef]
  98. Johnson, C.E. Soil nitrogen status 8 years after whole-tree clear-cutting. Can. J. For. Res. 1995, 1346–1355. [Google Scholar] [CrossRef]
  99. Johnson, C.E.; Romanowicz, R.B.; Siccama, T.G. Conservation of exchangeable cations after clear-cutting of a northern hardwood forest. Can. J. For. Res. 1997, 27, 859–868. [Google Scholar] [CrossRef]
  100. Johnson, D.W.; Walker, R.F.; Glass, D.W.; Stein, C.M.; Murphy, J.B.; Blank, R.R.; Miller, W.W. Effects of thinning, residue mastication, and prescribed fire on soil and nutrient budgets in a Sierra Nevada mixed-conifer forest. For. Sci. 2014, 60, 170–179. [Google Scholar] [CrossRef]
  101. Jones, H.S.; Beets, P.N.; Kimberley, M.O.; Garrett, L.G. Harvest residue management and fertilisation effects on soil carbon and nitrogen in a 15-year-old Pinus radiata plantation forest. For. Ecol. Manag. 2011, 262, 339–347. [Google Scholar] [CrossRef]
  102. Kaye, J.P.; Hart, S.C. Restoration and canopy-type effects on soil respiration in a ponderosa pine-bunchgrass ecosystem. Soil Sci. Soc. Am. J. 1998, 62, 1062–1072. [Google Scholar] [CrossRef]
  103. Keenan, R.J.; Messier, C.; Kimmins, J.P. (Hamish) Effects of clearcutting and soil mixing on soil properties and understorey biomass in western red cedar and western hemlock forests on northern Vancouver Island, Canada. For. Ecol. Manag. 1994, 68, 251–261. [Google Scholar] [CrossRef]
  104. Kelliher, F.M.; Ross, D.J.; Law, B.E.; Baldocchi, D.D.; Rodda, N.J. Limitations to carbon mineralization in litter and mineral soil of young and old ponderosa pine forests. For. Ecol. Manag. 2004, 191, 201–213. [Google Scholar] [CrossRef]
  105. Kishchuk, B.E.; Thiffault, E.; Lorente, M.; Quideau, S.; Keddy, T.; Sidders, D. Decadal soil and stand response to fire, harvest, and salvage-logging disturbances in the western boreal mixedwood forest of Alberta, Canada. Can. J. For. Res. 2015, 45, 141–152. [Google Scholar] [CrossRef]
  106. Klockow, P.A.; D’Amato, A.W.; Bradford, J.B. Impacts of post-harvest slash and live-tree retention on biomass and nutrient stocks in Populus tremuloides Michx.-dominated forests, northern Minnesota, USA. For. Ecol. Manag. 2013, 291, 278–288. [Google Scholar] [CrossRef]
  107. Klopatek, J.M. Belowground carbon pools and processes in different age stands of Douglas-fir. Tree Physiol. 2002, 22, 197–204. [Google Scholar] [CrossRef] [PubMed]
  108. Knoepp, J.D.; Swank, W.T. Forest management effects on surface soil carbon and nitrogen. Soil Sci. Soc. Am. J. 1997, 61, 928–935. [Google Scholar] [CrossRef]
  109. Korb, J.E.; Johnson, N.C.; Covington, W.W. Slash pile burning effects on soil biotic and chemical properties and plant establishment: Recommendations for amelioration. Restor. Ecol. 2004, 12, 52–62. [Google Scholar] [CrossRef]
  110. Kraemer, J.F.; Hermann, R.K. Broadcast burning: 25-year effects on forest soils in the western flanks of the Cascade Mountains. For. Sci. 1979, 25, 427–439. [Google Scholar]
  111. Kurth, V.J.; D’Amato, A.W.; Palik, B.J.; Bradford, J.B. Fifteen-year patterns of soil carbon and nitrogen following biomass harvesting. Soil Sci. Soc. Am. J. 2014, 78, 624–633. [Google Scholar] [CrossRef]
  112. Laiho, R.; Sanchez, F.; Tiarks, A.; Dougherty, P.M.; Trettin, C.C. Impacts of intensive forestry on early rotation trends in site carbon pools in the southeastern US. For. Ecol. Manag. 2003, 174, 177–189. [Google Scholar] [CrossRef]
  113. Latty, E.F.; Canham, C.D.; Marks, P.L. The effects of land-use history on soil properties and nutrient dynamics in northern hardwood forests of the Adirondack Mountains. Ecosystems 2004, 7, 193–207. [Google Scholar] [CrossRef]
  114. Law, B.E.; Thornton, P.E.; Irvine, J.; Anthoni, P.M.; Van Tuyl, S. Carbon storage and fluxes in Ponderosa pine forests at different developmental stages. Glob. Chang. Biol. 2001, 7, 755–777. [Google Scholar] [CrossRef]
  115. Law, B.E.; Sun, O.J.; Campbell, J.; Van Tuyl, S.; Thornton, P.E. Changes in carbon storage and fluxes in a chronoseuence of Ponderosa pine. Glob. Chang. Biol. 2003, 4, 510–524. [Google Scholar] [CrossRef]
  116. LeDuc, S.D.; Rothstein, D.E. Initial recovery of soil carbon and nitrogen pools and dynamics following disturbance in jack pine forests: A comparison of wildfire and clearcut harvesting. Soil Biol. Biochem. 2007, 39, 2865–2876. [Google Scholar] [CrossRef]
  117. Maassen, S.; Wirth, S. Soil microbiological monitoring of a pine forest after partial thinning for stand regeneration with beech seedlings. Soil Sci. Plant Nutr. 2004, 50, 815–819. [Google Scholar] [CrossRef]
  118. Mattson, K.G.; Smith, H.C. Detrital organic matter and soil CO2 efflux in forests regenerating from cutting in West Virginia. Soil Biol. Biochem. 1993, 25, 1241–1248. [Google Scholar] [CrossRef]
  119. Mattson, K.G.; Swank, W.T. Soil and detrital carbon dynamics following forest cutting in the Southern Appalachians. Biol. Fertil. Soils 1989, 7, 247–253. [Google Scholar] [CrossRef]
  120. May, B.M.; Attiwill, P.M. Nitrogen-fixation by Acacia dealbata and changes in soil properties 5 years after mechanical disturbance or slash-burning following timber harvest. For. Ecol. Manag. 2003, 181, 339–355. [Google Scholar] [CrossRef]
  121. McLaughlin, J.W.; Phillips, S.A. Soil carbon, nitrogen, and base cation cycling 17 years after whole-tree harvesting in a low-elevation red spruce (Picea rubens)-balsam fir (Abies balsamea) forested watershed in central Maine, USA. For. Ecol. Manag. 2006, 222, 234–253. [Google Scholar] [CrossRef]
  122. McKee, S.E.; Seiler, J.R.; Aust, W.M.; Strahm, B.D.; Schilling, E.B.; Brooks, S. Carbon pools and fluxes in a tupelo (Nyssa aquatica)-baldcypress (Taxodium distichum) swamp24-years after harvest disturbances. Biomass Bioenergy 2013, 55, 130–140. [Google Scholar] [CrossRef]
  123. McLaughlin, J.W.; Liu, G.; Jurgensen, M.F.; Gale, M.R. Organic carbon characteristics in a spruce swamp five years after harvesting. Soil Sci. Soc. Am. J. 1996, 60, 1228–1236. [Google Scholar] [CrossRef]
  124. Merino, A.; Edeso, J.M. Soil fertility rehabilitation in young Pinus radiata D. Don. plantations from northern Spain after intensive site preparation. For. Ecol. Manag. 1999, 116, 83–91. [Google Scholar] [CrossRef]
  125. Moreno-Fernandez, D.; Diaz-Pines, E.; Barbeito, I.; Sanchez-Gonzalez, M.; Montes, F.; Rubio, A.; Canellas, I. Temporal carbon dynamics over the rotation period of two alternative management systems in Mediterranean mountain Scots pine forests. For. Ecol. Manag. 2015, 348, 186–195. [Google Scholar] [CrossRef]
  126. Mu, C.C.; Lu, H.C.; Bao, X.; Wang, B.; Cui, W. Effects of selective cutting on vegetation carbon storage of boreal Larix gmelinii-Carex schmidtii forested wetlands in Daxing’anling, China. Shengtai Xuebao/Acta Ecol. Sin. 2013, 33, 5286–5298. [Google Scholar]
  127. Murphy, J.D.; Johnson, D.W.; Miller, W.W.; Walker, R.F.; Blank, R.R. Prescribed fire effects on forest floor and soil nutrients in a Sierra Nevada Forest. Soil Sci. 2006, 171, 181–199. [Google Scholar] [CrossRef]
  128. Neher, D.A.; Barbercheck, M.E.; El-Allaf, S.M.; Anas, O. Effects of disturbance and ecosystem on decomposition. Appl. Soil Ecol. 2003, 23, 165–179. [Google Scholar] [CrossRef]
  129. Norris, C.E.; Quideau, S.A.; Bhatti, J.S.; Wasylishen, R.E.; MacKenzie, M.D. Influence of fire and harvest on soil organic carbon in jack pine sites. Can. J. For. Res. 2009, 39, 642–654. [Google Scholar] [CrossRef]
  130. O’Brien, N.D.; Attiwill, P.M.; Weston, C.J. Stability of soil organic matter in Eucalyptus regnans forests and Pinus radiata plantations in south eastern Australia. For. Ecol. Manag. 2003, 185, 249–261. [Google Scholar] [CrossRef]
  131. Powers, M.; Kolka, R.; Palik, B.; McDonald, R.; Jurgensen, M. Long-term management impacts on carbon storage in Lake States forests. For. Ecol. Manag. 2011, 262, 424–431. [Google Scholar] [CrossRef]
  132. Prest, D.; Kellman, L.; Lavigne, M.B. Mineral soil carbon and nitrogen still low three decades following clearcut harvesting in a typical Acadian Forest stand. Geoderma 2014, 214–215, 62–69. [Google Scholar] [CrossRef]
  133. Prietzel, J.; Wagoner, G.L.; Harrison, R.B. Long-term effects of repeated urea fertilization in Douglas-fir stands on forest floor nitrogen pools and nitrogen mineralization. For. Ecol. Manag. 2004, 193, 413–426. [Google Scholar] [CrossRef]
  134. Puhlick, J.J.; Fernandez, I.J.; Weiskittel, A.R. Evaluation of forest management effects on the mineral soil carbon pool of a lowland, mixed-species forest in Maine, USA. Can. J. Soil Sci. 2016, 96, 207–218. [Google Scholar] [CrossRef]
  135. Rab, M.A. Soil physical and hydrological properties following logging and slash burning in the Eucalyptus regnans forest of southeastern Australia. For. Ecol. Manag. 1996, 84, 159–176. [Google Scholar] [CrossRef]
  136. Riley, J.M.; Jones, R.H. Factors limiting regeneration of Quercus alba and Cornus florida in formerly cultivated coastal plain sites, South Carolina. For. Ecol. Manag. 2003, 177, 571–586. [Google Scholar] [CrossRef]
  137. Roaldson, L.M.; Johnson, D.W.; Miller, W.W.; Murphy, J.D.; Walker, R.F.; Stein, C.M.; Glass, D.W.; Blank, R.R. Prescribed Fire and Timber Harvesting Effects on Soil Carbon and Nitrogen in a Pine Forest. Soil Sci. Soc. Am. J. 2014, 78, S48. [Google Scholar] [CrossRef]
  138. Rothstein, D.E.; Spaulding, S.E. Replacement of wildfire by whole-tree harvesting in jack pine forests: Effects on soil fertility and tree nutrition. For. Ecol. Manag. 2010, 260, 1164–1174. [Google Scholar] [CrossRef]
  139. Sanchez, F.G.; Scott, D.A.; Ludovici, K.H. Negligible effects of severe organic matter removal and soil compaction on loblolly pine growth over 10 years. For. Ecol. Manag. 2006, 227, 145–154. [Google Scholar] [CrossRef]
  140. Sanscrainte, C.L.; Peterson, D.L.; McKay, S. Carbon storage and soil properties in late-successional and second-growth subalpine forests in the North Cascade Range, Washington. Northwest Sci. 2003, 77, 297–307. [Google Scholar]
  141. Saynes, V.; Etchevers, J.D.; Galicia, L.; Hidalgo, C.; Campo, J. Soil carbon dynamics in high-elevation temperate forests of Oaxaca (Mexico): Thinning and rainfall effects. Bosque (Valdivia) 2012, 33, 3–11. [Google Scholar] [CrossRef]
  142. Selig, M.F.; Seiler, J.R.; Tyree, M.C. Soil carbon and CO2 efflux as influenced by the thinning of loblolly pine (Pinus taeda L.) plantations on the Piedmont of Virginia. For. Sci. 2008, 54, 58–66. [Google Scholar]
  143. Shelburne, V.B.; Boyle, M.F.; Lione, D.J.; Waldrop, T.A. Preliminary effects of prescribed burning and thinning as fuel reduction treatments on the piedmont soils of the clemson experimental forest. In General Technical Report SRS-71; U.S. Department of Agriculture, Forest Service, Southern Research Station: Asheville, NC, USA, 2004; pp. 35–38. [Google Scholar]
  144. Sheng, H.; Zhou, P.; Zhang, Y.; Kuzyakov, Y.; Zhou, Q.; Ge, T.; Wang, C. Loss of labile organic carbon from subsoil due to land-use changes insubtropical China. Soil Biol. Biochem. 2015, 88, 148–157. [Google Scholar] [CrossRef]
  145. Skovsgaard, J.P.; Stupak, I.; Vesterdal, L. Distribution of biomass and carbon in even-aged stands of Norway spruce (Picea abies (L.) Karst.): A case study on spacing and thinning effects in northern Denmark. Scand. J. For. Res. 2006, 21, 470–488. [Google Scholar] [CrossRef]
  146. Slesak, R.A.; Schoenholtz, S.H.; Harrington, T.B. Soil carbon and nutrient pools in Douglas-fir plantations 5 years after manipulating biomass and competing vegetation in the Pacific Northwest. For. Ecol. Manag. 2011, 262, 1722–1728. [Google Scholar] [CrossRef]
  147. Small, C.J.; McCarthy, B.C. Relationship of understory diversity to soil nitrogen, topographic variation, and stand age in an eastern oak forest, USA. For. Ecol. Manag. 2005, 217, 229–243. [Google Scholar] [CrossRef]
  148. Stone, J.E.; Kolb, T.E.; Covington, W.W. Effects of restoration thinning on presettlement Pinus ponderosa in northern Arizona. Restor. Ecol. 1999, 7, 172–182. [Google Scholar] [CrossRef]
  149. Stone, D.; Elioff, J. Soil properties and aspen development five years after compaction and forest floor removal. Can. J. Soil Sci. 1998, 78, 51–58. [Google Scholar] [CrossRef]
  150. Strong, T.F. Harvesting intensity influences the carbon distribution in a northern hardwood ecosystem. In Research Paper NC-329; Department of Agriculture, Forest Service, North Central Research Station: St. Paul, MN, USA, 1997. [Google Scholar]
  151. Strukelj, M.; Brais, S.; Paré, D. Nine-year changes in carbon dynamics following different intensities of harvesting in boreal aspen stands. Eur. J. For. Res. 2015, 134, 737–754. [Google Scholar] [CrossRef]
  152. Tang, J.; Bolstad, P.V.; Martin, J.G. Soil carbon fluxes and stocks in a Great Lakes forest chronosequence. Glob. Chang. Biol. 2009, 15, 145–155. [Google Scholar] [CrossRef]
  153. Trettin, C.C.; Jurgensen, M.F.; Gale, M.R.; McLaughlin, J.W. Recovery of carbon and nutrient pools in a northern forested wetland 11 years after harvesting and site preparation. For. Ecol. Manag. 2011, 262, 1826–1833. [Google Scholar] [CrossRef]
  154. Ussiri, D.A.N.; Johnson, C.E. Organic matter composition and dynamics in a northern hardwood forest ecosystem 15 years after clear-cutting. For. Ecol. Manag. 2007, 240, 131–142. [Google Scholar] [CrossRef]
  155. Vario, C.L.; Neurath, R.A.; Friedland, A.J. Response of mineral soil carbon to clear-cutting in a northern hardwood forest. Soil Sci. Soc. Am. J. 2014, 78, 309–318. [Google Scholar] [CrossRef]
  156. Vesterdal, L.; Dalsgaard, M.; Felby, C.; Raulund-Rasmussen, K.; Jørgensen, B.B. Effects of thinning and soil properties on accumulation of carbon, nitrogen and phosphorus in the forest floor of Norway spruce stands. For. Ecol. Manag. 1995, 77, 1–10. [Google Scholar] [CrossRef]
  157. Waldrop, M.P.; McColl, J.G.; Powers, R.F. Effects of forest postharvest management practices on enzyme activities in decomposing litter. Soil Sci. Soc. Am. J. 2003, 67, 1250–1256. [Google Scholar] [CrossRef]
  158. Wu, J.-S.; Jiang, P.-K.; Chang, S.X.; Xu, Q.-F.; Lin, Y. Dissolved soil organic carbon and nitrogen were affected by conversion of native forests to plantations in subtropical China. Can. J. Soil Sci. 2010, 90, 27–36. [Google Scholar] [CrossRef]
  159. Xiang, W.; Chai, H.; Tian, D.; Peng, C. Marginal effects of silvicultural treatments on soil nutrients following harvest in a Chinese fir plantation. Soil Sci. Plant Nutr. 2009, 55, 523–531. [Google Scholar] [CrossRef]
  160. Yanai, R.D.; Arthur, M.A.; Siccama, T.G.; Federer, C.A. Challenges of measuring forest floor organic matter dynamics: Repeated measures from a chronosequence. For. Ecol. Manag. 2000, 138, 273–283. [Google Scholar] [CrossRef]
  161. Zabowski, D.; Chambreau, D.; Rotramel, N.; Thies, W.G. Long-term effects of stump removal to control root rot on forest soil bulk density, soil carbon and nitrogen content. For. Ecol. Manag. 2008, 255, 720–727. [Google Scholar] [CrossRef]
  162. Zhong, Z.; Makeschin, F. Soil biochemical and chemical changes in relation to mature spruce (Picea abies) forest conversion and regeneration. J. Plant Nutr. Soil Sci. 2003, 166, 291–299. [Google Scholar] [CrossRef]
  163. Zummo, L.M.; Friedland, A.J. Soil carbon release along a gradient of physical disturbance in a harvested northern hardwood forest. For. Ecol. Manag. 2011, 261, 1016–1026. [Google Scholar] [CrossRef]
Figure 1. Response of soil C to forest harvesting, overall and faceted by soil depth. All points are back-transformed mean effect sizes ± 95% confidence intervals calculated by nonparametric bootstrap. The number of response ratios (k) that make up each mean effect is listed on the right. Mean effects with confidence intervals overlapping the dashed line (0%) show no significant change in soil C due to harvesting. Within each facet, mean effect sizes are shown for the overall effect as well as separately for studies reporting C pool units or concentration units.
Figure 1. Response of soil C to forest harvesting, overall and faceted by soil depth. All points are back-transformed mean effect sizes ± 95% confidence intervals calculated by nonparametric bootstrap. The number of response ratios (k) that make up each mean effect is listed on the right. Mean effects with confidence intervals overlapping the dashed line (0%) show no significant change in soil C due to harvesting. Within each facet, mean effect sizes are shown for the overall effect as well as separately for studies reporting C pool units or concentration units.
Forests 07 00308 g001
Figure 2. Response of soil C to harvesting in different soil orders. Mean effect sizes ± 95% confidence intervals calculated by nonparametric bootstrap are shown for all response ratios in each soil order (Overall) and broken out into mineral soil or O horizon. The number of response ratios (k) comprising each mean effect are listed on the right. Effect sizes were calculated only on response ratios reported in pool units (k = 746).
Figure 2. Response of soil C to harvesting in different soil orders. Mean effect sizes ± 95% confidence intervals calculated by nonparametric bootstrap are shown for all response ratios in each soil order (Overall) and broken out into mineral soil or O horizon. The number of response ratios (k) comprising each mean effect are listed on the right. Effect sizes were calculated only on response ratios reported in pool units (k = 746).
Forests 07 00308 g002
Figure 3. Response of soil C harvesting at different depths in soil, broken down by hardwood or conifer/mixed forest types. Mean effect sizes ± 95% confidence intervals calculated by nonparametric bootstrap are shown for hardwood and conifer/mixed forests. Blue circles show the mean difference between these forest types (Hardwood–Conifer/Mix) ± 95% confidence interval for the difference. Differences are calculated on the logarithmic effect size scale, and then back-transformed to % change, and thus do not necessarily add up on the % change scale. The number of response ratios (k) in each forest type at each depth is listed on the right. Data for very deep soil is not shown because there were no observations for this soil layer in hardwood forests.
Figure 3. Response of soil C harvesting at different depths in soil, broken down by hardwood or conifer/mixed forest types. Mean effect sizes ± 95% confidence intervals calculated by nonparametric bootstrap are shown for hardwood and conifer/mixed forests. Blue circles show the mean difference between these forest types (Hardwood–Conifer/Mix) ± 95% confidence interval for the difference. Differences are calculated on the logarithmic effect size scale, and then back-transformed to % change, and thus do not necessarily add up on the % change scale. The number of response ratios (k) in each forest type at each depth is listed on the right. Data for very deep soil is not shown because there were no observations for this soil layer in hardwood forests.
Forests 07 00308 g003
Figure 4. Differences in response of soil C to harvesting between treatment strategies. Differences are calculated by subtracting [more intensive treatment] − [less intensive treatment], such that positive differences represent reduced loss of C due to more intensive treatment, and negative differences represent increased losses of C due to more intensive treatment. Point estimates are back-transformed differences between mean effect sizes ± 95% confidence intervals calculated by nonparametric bootstrap. Mean effect differences with confidence intervals overlapping the dashed line (0%) show no significant difference between the harvesting, residual management, or pretreatment strategies. The number of response ratios (k) for the intensive treatment in each comparison appear on the left and for the less intensive treatment on the right.
Figure 4. Differences in response of soil C to harvesting between treatment strategies. Differences are calculated by subtracting [more intensive treatment] − [less intensive treatment], such that positive differences represent reduced loss of C due to more intensive treatment, and negative differences represent increased losses of C due to more intensive treatment. Point estimates are back-transformed differences between mean effect sizes ± 95% confidence intervals calculated by nonparametric bootstrap. Mean effect differences with confidence intervals overlapping the dashed line (0%) show no significant difference between the harvesting, residual management, or pretreatment strategies. The number of response ratios (k) for the intensive treatment in each comparison appear on the left and for the less intensive treatment on the right.
Forests 07 00308 g004
Figure 5. Temporal patterns in both O horizon (yellow triangle) and mineral soil (blue circle) C pools for Alfisol, Inceptisol, Spodosol, and Ultisol orders. Other soil orders are not shown due to an inadequate number of response ratios over time. Regression lines show trends with time using a second order polynomial. For the overall model, F = 9.205 on 7 and 532 degrees of freedom, Adj. R2 = 0.096, and p < 0.0001 (Table 2).
Figure 5. Temporal patterns in both O horizon (yellow triangle) and mineral soil (blue circle) C pools for Alfisol, Inceptisol, Spodosol, and Ultisol orders. Other soil orders are not shown due to an inadequate number of response ratios over time. Regression lines show trends with time using a second order polynomial. For the overall model, F = 9.205 on 7 and 532 degrees of freedom, Adj. R2 = 0.096, and p < 0.0001 (Table 2).
Forests 07 00308 g005
Figure 6. Number of response ratios plotted by the maximum depth of sampling for each observation. Response ratios calculated from concentration are in blue, and from pools in orange. The average depth of all response ratios is denoted by the solid green line (21.3 cm, n = 945). The average maximum sampling depth for all 112 publications in the meta-analysis is denoted by the dashed red line (35.9 cm).
Figure 6. Number of response ratios plotted by the maximum depth of sampling for each observation. Response ratios calculated from concentration are in blue, and from pools in orange. The average depth of all response ratios is denoted by the solid green line (21.3 cm, n = 945). The average maximum sampling depth for all 112 publications in the meta-analysis is denoted by the dashed red line (35.9 cm).
Forests 07 00308 g006
Figure 7. Absolute change in soil C due to harvest for each soil depth in this analysis (O horizon, top, mid, deep, very deep, and whole mineral soil). Different points show burned (yellow triangle) and unburned (blue circle) pretreatment strategies. Dashed 1:1 lines in each facet represent no response due to harvest. The total number of responses shown is k = 746.
Figure 7. Absolute change in soil C due to harvest for each soil depth in this analysis (O horizon, top, mid, deep, very deep, and whole mineral soil). Different points show burned (yellow triangle) and unburned (blue circle) pretreatment strategies. Dashed 1:1 lines in each facet represent no response due to harvest. The total number of responses shown is k = 746.
Forests 07 00308 g007
Table 1. Factors gathered as potential predictor variables in this meta-analysis.
Table 1. Factors gathered as potential predictor variables in this meta-analysis.
FactorLevels
Reporting unitsPool (Mg·ha−1), concentration (% or mg·g−1)
Soil DepthO horizonForest Floor
Top Soil0–15 cm
Mid Soil15–30 cm
Deep Soil30–60 cm
Very Deep60–100+ cm
Overstory speciesHardwood, conifer/mixed
Soil orderAlfisol, Andisol, Entisol, Inceptisol, Mollisol, Spodosol, Oxisol, Ultisol
Geographic groupNE North America, NW North America, SE North America, SW North America, Europe, Asia, Pacific (Australia, New Zealand)
Harvest typeClearcut, thin
Harvest intensityStem only, whole tree
Residue managementRemoved, spread
Site preparationBroadcast burn, tillage/scarification
Soil textureFine (mostly silt or clay), coarse (mostly sand), organic
Time since harvestContinuous
Mean Annual Temperature0–5, 5–7.5, 7.6–10, 10.1–15, 15.1–20, >20 (°C)
Mean Annual Precipitation<500, 500−750, 751−1000, 1001−1400, 1401−1800, >1800 (mm)
Table 2. Linear regression coefficients and significance for second degree polynomial model of response of soil C to harvesting over time.
Table 2. Linear regression coefficients and significance for second degree polynomial model of response of soil C to harvesting over time.
CoefficientEstimateSEt-Valuep-Value
Intercept (Alfisol, mineral soil)12.7023.5873.5410.0004
O horizon−21.4753.766−5.703<0.0001
Inceptisol−10.8765.717−1.9020.0577
Spodosol−14.7174.320−3.4070.0007
Ultisol−24.7765.391−4.596<0.0001
Time−67.83441.56−1.6320.10325
Time2120.41240.3612.9830.0030
Residual SE: 40.24 on 533 DF
F-Statistic: 10.74 on 6 and 533 df, p < 0.0001R2 = 0.108Adj. R2 = 0.098
Forests EISSN 1999-4907 Published by MDPI AG, Basel, Switzerland RSS E-Mail Table of Contents Alert
Back to Top