Environmental and vegetation control on active layer and soil temperature in an Arctic tundra ecosystem in Alaska

The sites of this research are near Utqiaġvik, Alaska (formerly known as Barrow), which is the largest town in the Alaskan North Slope Borough. Our team has and continues to maintain several micrometeorological and eddy covariance towers in Utqiaġvik over the last decades (Zona et al. 2016). This research was conducted near two of these eddy covariance towers, which were in operation from 2005-2009 (Fig. 1b, the US-Ben (North), and US-Bec (Central) sites, established during the Biocomplexity Experiment, see Zona et al. 2009; Zona et al. 2012). Environmental drivers (such as air temperature, and photosynthetic active radiation (PAR)) measured from an additional currently operational tower (US-Bes), in close proximity to the US-Ben and US-Bes sites, were also included in this study. The locations of the US-Bes, US-Bec, and US-Ben are: 71.2809N, -156.5965W; 71.28316N, -156.60342W; and 71.28628N, -156.60424W respectively (Zona et al. 2009). These sites are located in a drain lake basin ecosystem and a mixed polygon wet sedges ecosystem characterized by mosses, lichens, and graminoids with patches of water and partially to fully submerged patches of vegetation (Davidson et al. 2016). Given the proximity of these sites (US-Ben and US-Bec being within 662-meters and 356-meters of US-Bes respectively), we assume that the air temperature and PAR collected in US-Bes were representative of the US-Ben and US-Bec sites. Access to US-Bes, US-Bec, and US-Ben sites was facilitated by the establishment of boardwalks during the Biocomplexity experiment in summer 2005 (Zona et al. 2009). These boardwalks allowed sampling across the sites while limiting disturbance. Data was collected every two meters in both US-Ben and US-Bec across 124-meters transects that parallel historical water table data collection (Zona et al. 2012), for a total of 62 plots in US-Ben and 62 in US-Bec. The data collection was performed at 2-meter intervals along two 124-meter transects at US-Bec and US-Ben (Fig. 1b). In each of these plots we recorded the thickness of the total moss layer and the thickness of the green living moss layer. A previous in-depth vegetation analysis from our team showed that the general region where this data was collected was highly homogenous in its vegetative structure being composed primarily of graminoids as the dominant vascular plant group and mosses such as Sphagnum sp. and Drepanocladus sp. (Davidson et al. 2016). We also collected data on the dominant moss genus at every point and found Sphagnum sp. and Drepanocladus sp. to be the dominant genera at our sites with most of the plots being Sphagnum sp. (n = 55) or some combination of predominantly Sphagnum sp. and Drepanocladus sp. (n = 46) and a limited number of plots with only Drepanocladus sp. (n = 21). The moss layer thickness and genus identification were recorded in sections of approximately 25-square cm (5 cm x 5 cm) in each of the 2-meter plots only once (i.e., the first week of July in 2021 for all plots and a subset of these plots on the first week of July in 2022) during the peak season (i.e., the first week of July to the second week of August 2021) to reduce the disturbance to the moss layer. These sections were carefully removed using a serrated knife trying to limit damage to surrounding vegetation. Afterwards, the thickness of the entire moss mat was measured with a ruler, and the samples were reinserted into the hole created by sampling. In each of the 62 plots we also recorded moss and soil temperature every cm from 1-cm below the surface until 20-cm below ground on a weekly basis using type T thermocouples connected to a CR3000 datalogger (Campbell Scientific, Logan, UT, USA). These temperature profiles allowed us to determine how the presence and thickness of the moss layer affected the thermal difference across the moss and soil layers. These 21 thermocouples were attached to a fiberglass probe, which facilitated insertion in the moss layer and soil. Each point was measured for approximately 3 minutes as the temperature readings stabilized within the first couple of minutes. We also collected soil water content (percent water) weekly in the first 5-cm of the moss or soil layer using a FieldScout TDR300 (Spectrum Technologies, Aurora, IL, USA) and 5-cm rods (Beringer et al. 2005; Hayashi et al. 2007; Hrbáček et al. 2020). The FieldScout was calibrated using local water samples to account for nutrients which may influence conductivity. Thaw depth and water table levels (cm) were also collected weekly in each of the sampling plots using a metal and wooden probe respectively with markings indicating intervals of 1-cm depths. Water table measurements were collected inside PVC pipes (with holes every 1 cm) previously installed along the transects (Zona et al. 2009; Zona et al. 2012) in each of the sampling locations. This data collection was repeated for a second field season in a subset of the plots (n = 20 in Summer 2022 vs. n = 124 in Summer 2021) to reduce disturbance to the site, but both samples were compared for consistency in the thickness of the moss layers across different field seasons. This comparison showed good agreement between the measured moss thickness in both seasons (R2 = 91.82%, p-value less than 0.001, 2022 Moss Thickness (cm) = 1.01967 * 2021 Moss Thickness (cm) + 0.228272). Environmental variables collected by the eddy covariance US-Bes tower, included PAR, air temperature, local surface and subsurface soil temperature, relative humidity, wind speed, and net radiation. Elevation above sea level was collected in each of the sampling plots at US-Ben and US-Bec as reported in Zona et al. (2012). These measurements described the microtopography of each of the sampling plots and let us test its influence on the environmental conditions, vegetation, and active layer development.

本研究的观测站点位于阿拉斯加州乌特基亚维克（Utqiaġvik，旧称巴罗Barrow），该地是阿拉斯加北坡自治区规模最大的城镇。过往十余年间，我们团队已在乌特基亚维克布设并持续维护多座微气象塔（micrometeorological tower）与涡度协方差塔（eddy covariance tower）（Zona等，2016）。本研究的采样区域临近其中两座2005至2009年运行的涡度协方差塔，即图1b中的US-Ben（北）与US-Bec（中）站点，二者均为生物复杂性实验期间布设，详见Zona等，2009；Zona等，2012。本研究同时纳入了邻近US-Ben与US-Bec站点、当前仍在运行的US-Bes塔所采集的环境驱动因子数据，包括气温与光合有效辐射（Photosynthetic Active Radiation, PAR）等。US-Bes、US-Bec与US-Ben的坐标分别为71.2809°N，-156.5965°W；71.28316°N，-156.60342°W；以及71.28628°N，-156.60424°W（Zona等，2009）。上述站点地处排水湖盆生态系统与混合多边形湿草苔原生态系统，该生态系统以苔藓、地衣与禾本科植物为主要植被，伴生片状水体及部分至完全浸没的植被斑块（Davidson等，2016）。鉴于各站点间距较近（US-Ben与US-Bec分别距US-Bes仅662米与356米），本研究假定US-Bes塔采集的气温与光合有效辐射数据可代表US-Ben与US-Bec站点的环境条件。2005年夏季的生物复杂性实验期间搭建的栈道为US-Bes、US-Bec与US-Ben站点的野外作业提供了便利，同时最大限度降低了采样对原生环境的干扰。在US-Ben与US-Bec站点，我们沿平行于历史地下水位监测的两条124米样带，以每2米为间隔设置采样点，最终在US-Ben与US-Bec站点各布设62个样方（Zona等，2012；图1b）。在每个样方中，我们记录了总苔藓层厚度与绿色活苔藓层厚度。本团队此前的植被深度分析显示，本研究采样区域的植被结构均一性较强，优势维管植物类群为禾本科植物，苔藓则以泥炭藓属（Sphagnum sp.）和镰刀藓属（Drepanocladus sp.）为主（Davidson等，2016）。我们同时记录了每个采样点的优势苔藓属类，结果显示采样点的优势苔藓类群以泥炭藓属（共55个样方）、泥炭藓属与镰刀藓属混合（共46个样方）为主，仅少量样方以镰刀藓属为单一优势类群（共21个样方）。为降低对苔藓层的干扰，我们仅在植被盛季（2021年7月第一周至8月第二周）对每个2米间隔样方内约25平方厘米（5cm×5cm）的区域进行一次苔藓样本采集：于2021年7月第一周完成全部样方的采样，并于2022年7月第一周对其中部分样方进行复测。采样时使用锯齿小刀小心取下样本，尽量避免损伤周边植被，随后用直尺测量完整苔藓垫的厚度，再将样本重新放回采样孔中。在62个样方中，我们还使用连接至CR3000数据记录仪（Campbell Scientific, Logan, UT, USA）的T型热电偶，每周采集1厘米至20厘米深度的苔藓与土壤温度，每厘米记录一次。该温度剖面数据可用于分析苔藓层的存在与厚度对苔藓-土壤层间热温差的影响。21支热电偶被固定于玻璃纤维探针上，便于插入苔藓层与土壤中。每个采样点的温度测量持续约3分钟，直至前两分钟内读数趋于稳定。我们还使用FieldScout TDR300时域反射仪（Spectrum Technologies, Aurora, IL, USA）搭配5厘米探针，每周采集苔藓或土壤表层5厘米内的土壤含水量（以百分比表示），并使用当地水样对仪器进行校准，以消除可能影响电导率的营养盐干扰（Beringer等，2005；Hayashi等，2007；Hrbáček等，2020）。我们还分别使用带1厘米刻度标记的金属探针与木质探针，每周采集每个采样样方的冻融深度与地下水位（单位：厘米）数据。地下水位测量依托此前沿样带布设的带1厘米孔径PVC管进行（Zona等，2009；Zona等，2012）。为减少对站点的干扰，我们仅在2022年野外季对部分样方（2021年共124个样方，2022年复测20个）进行了重复数据采集，并对比了两次野外季的苔藓厚度数据以验证一致性。对比结果显示两次测量的苔藓厚度具有良好的一致性（决定系数R²=91.82%，p值<0.001，回归方程为：2022年苔藓厚度（厘米）=1.01967×2021年苔藓厚度（厘米）+0.228272）。涡度协方差塔US-Bes所采集的环境变量还包括光合有效辐射、气温、局地地表与地下土壤温度、相对湿度、风速与净辐射。我们参照Zona等（2012）的方法，在US-Ben与US-Bec站点的每个采样样方中采集了海拔高程数据，该数据可用于表征各采样样方的微地形特征，进而检验微地形对环境条件、植被状况与活动层发育的影响。