Prospect Hill Soil Warming Experiment at Harvard Forest since 1991

The soil warming experiment was installed on the Prospect Hill tract in 1991 to allow us to investigate the effects of a 5 deg C temperature increase on soil processes fundamental to the global cycling of carbon and nitrogen. The experiment is located in an even-aged mixed hardwood forest. Six replicates of three treatments, Heated (resistance heating cables buried at 10cm and maintained at a 5 deg C differential from the control plots), Disturbance Control (cables installed but not powered) and Control treatments make up the randomized block design. The temperature differential is maintained with monitoring at five minute increments by an automated thermistor network in the plots, wired to a multiplexer and a datalogger in the control shed. In the plots, measurements of trace gasses (CO2, N2O and CH4), nitrogen mineralization, soil moisture and soil water chemistry have allowed us to quantify changes in the soil system.
Ten years of elevated soil temperatures at the Harvard Forest soil warming experiment suggest that there are limits to a positive feedback to the global warming cycle. After many early years of increased CO2 fluxes from the warmed plots, years nine and ten have revealed no significant differences in releases of CO2 between the heated and control plots. Nitrogen mineralization has shown a large response to warming as well, with twice the rate of N mineralized in years 1-4, followed by a gradual decrease in rates to about the 40% level in 1998. Resumption of mineralization measurements in 2001 reveals a continued decrease in mineralization rates.
Field results from the soil warming experiment indicate that only a small fraction of the soil carbon in this mid-latitude forest ecosystem will be lost to the atmosphere in response to warming. We find that a 5 deg C warming of the soil for a decade results in a loss of about 11% of the carbon stored in the top 60 cm of soil, with most of this loss occurring in the first four to five years. By the end of the decade, warming no longer stimulates soil carbon loss. In addition, we find that warming accelerates the soil nitrogen cycle, particularly in the early years of warming. This acceleration has the potential to stimulate plant carbon storage.
Calculations based on nitrogen uptake in vegetation at the Chronic N Experiment (Magill, Aber et al. 2000) allow us to estimate how much additional carbon might be stored in plants as a result increased N availability at soil warming. Our calculations show that the potential increase in carbon stored is large enough to more than compensate for the soil carbon loss, so that warming may actually stimulate carbon storage at the ecosystem level in this mid-latitude forest.
We are currently expanding our research from the original soil warming experiment to investigate this important feedback at the Harvard Forest. The existing 6x6m plots on the Prospect Hill tract are not large enough to allow for the study of treatment effects on trees, so we are installing a larger warming experiment. The new experiment will have two 30x30m plots, a heated and a control plot, in the Slab City Tract of the Harvard Forest. Many aspects of this experiment will be the same as the original. Using resistance heating cables and thermistors, we will maintain a 5 deg C differential between the heated and control plots, year-round. In addition to the trace gas, nitrogen mineralization and soil water parameters measured to date, we will also be measuring woody increment and other vegetation metrics to allow us to investigate the plant response to perturbations in the carbon and nitrogen cycles.

1991年，土壤增温实验（soil warming experiment）于展望山样地（Prospect Hill tract）启动，旨在探究5℃升温对全球碳、氮循环核心土壤过程的影响。本实验设置于同龄混交硬木林内，采用随机区组设计（randomized block design），包含3种处理、共6个重复：加热组（电阻加热电缆埋深10cm，与对照样地维持5℃温差）、干扰对照（安装电缆但不通电）与对照组。
样地内布设自动热敏电阻网络（automated thermistor network），以5分钟为间隔开展监测，并通过线缆连接至控制棚内的多路复用器（multiplexer）与数据采集器（datalogger），以维持设定温差。研究团队针对样地内的痕量气体（trace gasses，CO₂、N₂O与CH₄）、氮矿化（nitrogen mineralization）、土壤湿度及土壤水化学特征开展测定，由此量化土壤系统的动态变化。

哈佛森林（Harvard Forest）土壤增温实验的十年原位升温结果显示，全球变暖循环的正反馈效应存在阈值。在升温初期数年，升温样地的CO₂通量显著升高，但第9、10年时，加热组与对照组的CO₂释放量已无显著差异。氮矿化对升温也表现出强烈响应：第1-4年的氮矿化速率为对照的2倍，随后速率逐渐下降，至1998年降至约对照的40%；2001年恢复监测后发现，矿化速率仍在持续降低。

本土壤增温实验的野外结果表明，该中纬度森林生态系统中仅极小部分土壤碳会因升温向大气释放。研究发现，土壤持续10年的5℃升温会导致0-60cm土层储存的碳损失约11%，其中绝大部分损失发生在最初4-5年；至实验十年末，升温不再促进土壤碳释放。此外，升温会加速土壤氮循环，尤其在升温初期，这种加速效应或可促进植物碳储存。

基于长期氮输入实验（Chronic N Experiment，Magill、Aber等，2000）中的植被氮吸收数据，研究团队可估算土壤升温后氮有效性提升所带来的植物碳储存增量。计算结果显示，潜在的植物碳储存增量足以抵消土壤碳损失，因此在该中纬度森林生态系统中，升温实际上可能促进整体碳储存。

目前研究团队正基于原始土壤增温实验拓展研究，以在哈佛森林中探究这一关键反馈机制。展望山样地原有的6×6m样地尺寸过小，无法开展处理措施对树木影响的相关研究，因此正布设更大规模的增温实验。新实验设置于哈佛森林的板城样地（Slab City Tract），包含两块30×30m样地，分别为加热组与对照组。
新实验的诸多设置与原始实验一致：将采用电阻加热电缆与热敏电阻，全年维持加热组与对照组之间的5℃温差。除目前已测定的痕量气体、氮矿化与土壤水参数外，研究团队还将测定木质增量与其他植被指标，以探究植物对碳、氮循环扰动的响应。