Eastern Redback Salamander Abundance at the Arnold Arboretum 2004-2005

Terrestrial salamanders are significant contributors to the overall vertebrate biomass in a forest (Burton and Likens 1975a). Due to this abundance, they are ecologically important as both predators of soil fauna, and as prey for larger animals such as birds, reptiles, and small mammals (Welsh and Droege 2001). In 1997, the invasive insect pest, hemlock woolly adelgid (Adelges tsugae) or HWA, was discovered on Hemlock Hill at Harvard University’s Arnold Arboretum in Jamaica Plain, Massachusetts. Unfortunately, this infestation has led to the death, decline, and removal of a larger percentage of the eastern hemlock (Tsuga canadensis) trees in this section of the Arboretum. In April 2004, researchers from the Harvard Forest and the Arnold Arboretum began studying the environmental impacts of this disturbance event (HF061). To complement this research twenty-four artificial cover objects (ACOs) were installed in three plots on Hemlock Hill in the summer of 2004 to monitor the relative abundance of eastern red-backed salamanders (Plethodon cinereus) and any other herpetofauna encountered. The timing of the installation permitted an analysis of the effects of a logging treatment conducted in the winter of 2004-2005 on the relative abundance of terrestrial salamanders. Logging had an immediate impact on the relative abundance of eastern red-backed salamanders in both logged plots as seen by an 83 percent decline in the relative abundance of red-backs in plot 1 and a 63 percent decline in plot 2 in the spring following logging versus the fall prior to logging. In the unlogged control plot (plot 3) the relative abundance declined 9 percent in the spring versus the fall. The lower relative abundances in the logged plots were likely due to higher temperatures on the surface of the soil (5.7 deg C) and five centimeters below the soil surface (1.2 deg C), and lower relative humidity on the surface of the soil (3.4%) than in the unlogged plot. By Fall 2005, the relative abundance of red-backs had nearly recovered in plot 2 with only a 7% decline from Fall 2004, compared to a 15% increase in the control plot. However, in plot 1 the relative abundance of red-backs declined further in Fall 2005 with 94% fewer salamanders than in Fall 2004. These differences are likely because plot 1 is on a steeper slope leading to less retention of moisture and higher levels of sunlight reaching the forest floor than plot 2. This physical difference led to plot 1 having higher temperatures on the surface of the soil, higher soil temperatures, and lower relative humidity on the surface of the soil than plot 2. In addition to the 147 Eastern Red-back Salamander observations, twelve American Toad (Bufo americanus), three Northern Dusky Salamander (Desmognathus fuscus), and one Garter Snake (Thamnophis sirtalis) observation were made under ACOs. It is important to point out that many repeat observations were likely made, and that 147 observations does not suggest 147 unique red-backs observed. Of the Eastern Red-backed Salamanders observed 40% were leadback morphs and 60% were red-back morphs. This is a higher percentage of leadback morphs than in any of the 50 Eastern Red-backed Salamander populations in New England studied by Lotter and Scott (1977), and more similar to the percentages seen in populations in Pennsylvania, Maryland, and Ohio. Interestingly, Lotter and Scott found that the frequency of leadback morphs is positively correlated with warmer climates. All three Northern Dusky Salamander observations were made under ACO 6 in plot 1 in 2004, and this species was not observed again in 2005. The Garter Snake observation was made in 2005 under ACO 10 in plot 2. The relative abundance of red-backs in the control plot was higher than in any of the thirty second-growth eastern hemlock-dominated and mixed deciduous forests studied by the author during the same period in 2005. However, this may be more a reflection of a lower abundance of coarse woody debris on the forest floor of Hemlock Hill than in the unmanaged forests studied in north central Massachusetts than an actual higher relative abundance of salamanders. That is, fewer pieces of coarse woody debris would mean fewer natural cover objects for salamanders, thereby making ACOs more attractive. Evidence that natural cover objects were less abundant at the Arboretum than in north central Massachusetts was the much higher frequency of multiple red-backs observed under the same ACO. Two red-backs were observed under an ACO in 5% of observations at the Arboretum versus 0.6% of observations in north central Massachusetts, three observations were made under the same ACO in 2% of observations versus 0.08%, and four or more observations were made in 1% of observations versus 0%. The Arnold Arboretum has a long history of breeding bird studies spanning from 1895 to 2005 which have allowed researchers to track changes in species composition over more than a century (Mayer 2005). Similarly, the baseline data provided here will hopefully be used by future studies to assess potential changes in the relative abundance of terrestrial herpetofauna on Hemlock Hill at the Arnold Arboretum as it undergoes significant vegetative changes due to HWA infestation. Another interesting question that future research will be able to assess is whether the percentage of leadback morphs in the population changes over time, given the potential changes in climate. In the future hopefully other areas of the Arboretum will be studied as well including the woods to the west of the Bradley Collection of Rosaceous Plants where spotted salamanders (Ambystoma maculatum) could be breeding in the adjacent ponds, and Bussey Brook and the stream running through the Meadow which may provide habitat for northern two-lined salamanders (Eurycea bislineata).

陆地蝾螈是森林生态系统脊椎动物总生物量的重要组成部分（Burton与Likens，1975a）。鉴于其种群丰度较高，该类群在生态系统中兼具双重生态功能：既是土壤动物区系的捕食者，同时也是鸟类、爬行动物与小型哺乳类等大型动物的猎物（Welsh与Droege，2001）。

1997年，入侵性昆虫害虫铁杉球蚜（hemlock woolly adelgid, Adelges tsugae，简称HWA）在马萨诸塞州牙买加平原哈佛大学阿诺德植物园的铁杉山被发现。此次侵染导致该区域内大量东部铁杉（Tsuga canadensis）死亡、长势衰退乃至被移除。

2004年4月，哈佛森林与阿诺德植物园的研究人员启动了针对此次干扰事件的环境影响研究（项目编号HF061）。为补充该研究内容，研究团队于2004年夏季在铁杉山的3个样地内布设了24个人工覆盖物（artificial cover objects, ACOs），用以监测东部红背蝾螈（Plethodon cinereus）及其他两栖爬行动物类群（herpetofauna）的相对丰度。本次布设的时间窗口恰好可用于分析2004-2005年冬季实施的采伐处理对陆地蝾螈相对丰度的影响。

采伐对两个采伐样地内东部红背蝾螈的相对丰度产生了即时影响：相较于采伐前秋季的监测数据，采伐后次年春季，样地1内红背蝾螈的相对丰度下降了83%，样地2下降了63%。未采伐的对照样地（样地3）的相对丰度仅下降了9%。采伐样地内相对丰度较低的原因，可能是土壤表层（5.7℃）与地下5厘米处（1.2℃）的温度更高，且土壤表层相对湿度（3.4%）低于对照样地。

至2005年秋季，样地2内红背蝾螈的相对丰度已基本恢复，较2004年秋季仅下降7%；而对照样地的相对丰度则上升了15%。但样地1内红背蝾螈的相对丰度在2005年秋季进一步下降，较2004年秋季减少了94%。这一差异可能源于样地1的坡度更陡，导致水分留存更少、林地表层接收到的光照强度高于样地2，进而使得样地1的土壤表层温度、土壤温度更高，土壤表层相对湿度更低。

除147次东部红背蝾螈观测记录外，研究团队还在人工覆盖物下记录到12次美洲蟾蜍（Bufo americanus）、3次北暗蝾螈（Desmognathus fuscus）以及1次束带蛇（Thamnophis sirtalis）的观测。需要说明的是，本次观测中存在大量重复记录，因此147次观测并不代表共观测到147只独立的红背蝾螈。在观测到的东部红背蝾螈中，40%为铅背色型（leadback morphs），60%为红背色型（red-back morphs）。这一铅背色型的占比高于Lotter与Scott（1977）研究的新英格兰地区50个东部红背蝾螈种群，更接近宾夕法尼亚州、马里兰州与俄亥俄州种群的占比水平。值得注意的是，Lotter与Scott发现铅背色型的出现频率与温暖气候呈正相关。

所有3次北暗蝾螈的观测均于2004年在样地1的ACO 6下完成，该物种在2005年未被再次观测到。束带蛇的观测记录于2005年在样地2的ACO 10下获取。

2005年同期，对照样地内红背蝾螈的相对丰度高于作者研究过的30片以次生东部铁杉为优势种的落叶混交林。但这一现象可能更多源于铁杉山林地表层的粗木质残体（coarse woody debris）数量低于马萨诸塞州中北部未受管理的森林，而非蝾螈实际相对丰度更高。换言之，粗木质残体越少，蝾螈可利用的天然覆盖物就越少，因此人工覆盖物的吸引力也就越强。相较于马萨诸塞州中北部的森林，阿诺德植物园内天然覆盖物更为匮乏的证据之一，是同一块人工覆盖物下多次观测到红背蝾螈的频率更高：在阿诺德植物园的观测中，5%的记录为同一块ACO下观测到2只红背蝾螈，而马萨诸塞州中北部的这一比例仅为0.6%；2%的记录为同一块ACO下观测到3只，后者仅为0.08%；1%的记录为同一块ACO下观测到4只及以上，后者则为0%。

阿诺德植物园自1895年至2005年开展了长期的繁殖鸟类研究，跨度超过一个世纪，为研究人员追踪物种组成的变化提供了数据支撑（Mayer，2005）。同理，本研究提供的基线数据有望为未来研究提供支持，用以评估阿诺德植物园铁杉山地区因HWA侵染而发生的显著植被变化下，陆地两栖爬行动物类群相对丰度的潜在变化。未来研究还可探讨另一个有趣的问题：鉴于气候的潜在变化，种群内铅背色型的占比是否会随时间发生改变。

未来还可对植物园的其他区域开展研究，例如布拉迪玫瑰植物收藏区西侧的林地——其邻近的池塘可能是斑点蝾螈（Ambystoma maculatum）的繁殖场所，以及巴斯溪与穿过草甸的溪流，这些生境可能为北两线蝾螈（Eurycea bislineata）提供栖息环境。