Size-dependent predation and intraspecific inhibition of an estuarine snail feeding on oysters

This experiment used drills and oysters that were collected from subtidal habitats in Apalachicola Bay, FL USA (both collections occurred around 29º 40’ 32.56” N, 84º 51’ 36.76” W and 29º 42’ 7.2” N, 84º 49’ 31.76” W), and was carried out at the University of South Florida, College of Marine Science in St. Petersburg, FL (USFCMS). At USFCMS three separate, closed-seawater systems, each comprising ten experimental tanks (47.3 l; 68 ´ 40 ´ 27.5 cm), were constructed for a total of 30 tanks. The base area of these tanks was similar to the first experiment, 0.27 m2. Within each system, approximately 950 l of artificial seawater was recirculated through the experimental tanks and a large sump (633 l; 76 ´ 45.5 ´ 183 cm). In Experiment II at UFSCMS, salinity was maintained at 25 by mixing deionized water with Instant Ocean sea salt (Instant Ocean Spectrum Brands, Blacksburg, VA, USA), and checked daily with a YSI 85 (Xylem Inc, Yellow Springs, OH, USA). Water temperature was maintained between 20-22°C using Finnex TH-800 Plus, 800-watt titanium heating rods (Finnex, Chicago IL, USA) with a Reef Octopus Heater Controller (Honya Co. Ltd, Shenzhen, China). The temperature and salinity were chosen based on conditions that occurred during an outbreak in the northern GOM (Menzel et al.,1966; FFWCC, 2013). During both experiments and the one week holding prior to Experiment II, oysters were fed Instant Algae Shellfish Diet 1800 (Reed Mariculture Inc, San Jose, CA) daily, following the manufacturer’s instructions of 3.6 ml per 100 g of oyster wet weight. In the seawater system at USFCMS, six drill abundances (2, 3, 4, 6, 8, and 12 drills per tank) were orthogonally crossed with five oyster abundances (3, 4, 6, 8, and 12 per tank), yielding two replicates for each orthogonal cross. The wide range of abundances reflects a response-surface approach to the regression analysis, improving the chances of detecting interactions between experimental factors despite low replication at any one combination of treatments. The treatment abundances were equivalent to densities ranging from 7-44 m-2 (drills) and 16-65 m-2 (oysters), both within the natural range observed for each species during and after outbreak conditions (See Supplemental Fig. 1). In this second experiment, drills ranged in size from 30-81 mm and were separated into small (30-55 mm) and large (56-81 mm) size classes. Oysters ranged in size from 25-99 mm and were separated into small (25-49 mm), medium (50-74 mm), and large (75-99 mm) size classes, as in Experiment I. Abundances of drills and oysters were split equally among each size class. To quantify oyster mortality in the absence of drills, each level of oyster abundance was paired with a corresponding control treatment (n=2) that lacked an oyster drill; no oysters died in these controls. After collection from the field, drills were held for one week and allowed to feed ad libitum on oysters before the start of the experiment to reduce any transportation effect to USFCMS from Apalachicola Bay, FL. The experiment began with a 5-day starvation period, based on the feeding rate of individual-housed drills from Experiment I, to standardize predator hunger. During feeding trials, tanks were checked twice daily for dead oysters, and the number of drills on each oyster was counted. Any dead oysters were replaced with a live oyster from the same size class to preserve a constant prey density, as assumed in predator functional response models that quantify the instantaneous feeding rate as opposed to the integrated feeding rate. After the 15-day experimental period, both the per capita feeding rate and the aggregate feeding rate (oysters consumed per replicate) were calculated.

本实验所用的牡蛎捕食螺（oyster drill）与牡蛎采集自美国佛罗里达州阿巴拉契科拉湾的潮下栖息地（subtidal habitats），两处采集点的坐标分别为29°40′32.56″N、84°51′36.76″W与29°42′7.2″N、84°49′31.76″W，实验于美国佛罗里达州圣彼得斯堡的南佛罗里达大学海洋科学学院（University of South Florida College of Marine Science, USFCMS）开展。
USFCMS内搭建了3套独立的封闭海水系统，每套系统包含10个实验水槽（容积47.3 L，尺寸68×40×27.5 cm），总计30个水槽。该类水槽的底面积与首次实验一致，为0.27 m²。每套系统内约950 L人工海水经实验水槽与大型集水槽（容积633 L，尺寸76×45.5×183 cm）循环流动。
在USFCMS开展的实验II中，研究人员通过将去离子水与Instant Ocean海盐（Instant Ocean Spectrum Brands，美国弗吉尼亚州布莱克斯堡）混合，将盐度维持在25，并每日使用YSI 85型水质分析仪（Xylem Inc，美国俄亥俄州耶洛斯普林斯）检测盐度。水温通过Finnex TH-800 Plus型800 W钛加热棒（Finnex，美国伊利诺伊州芝加哥）搭配Reef Octopus温控器（Honya Co. Ltd，中国深圳）维持在20~22℃。该水温与盐度参照了墨西哥湾北部灾害暴发期间的环境条件（Menzel等，1966；佛罗里达鱼类与野生动物保护委员会（FFWCC），2013）。
在两次实验及实验II开展前的一周暂养阶段，研究人员每日按照制造商推荐剂量——每100 g牡蛎湿重投喂3.6 ml——使用Instant Algae Shellfish Diet 1800（Reed Mariculture Inc，美国加利福尼亚州圣何塞）饲喂牡蛎。
USFCMS的海水系统中，设置了6个螺类密度梯度（每个水槽2、3、4、6、8、12只螺），并与5个牡蛎密度梯度（每个水槽3、4、6、8、12只牡蛎）进行正交组合，每个正交组合设置2个重复。该宽范围的密度梯度采用响应面法开展回归分析，可在单一处理组合重复数较低的情况下，提升检测实验因子间交互作用的概率。本次实验的处理密度对应种群密度范围为：螺类7~44 只·m⁻²，牡蛎16~65 只·m⁻²，二者均处于灾害暴发期间及之后野外观测到的各物种种群密度自然范围内（详见补充图1）。
本次实验II中，螺类的体长范围为30~81 mm，分为小型（30~55 mm）与大型（56~81 mm）两个规格等级；牡蛎体长范围为25~99 mm，与首次实验一致，分为小型（25~49 mm）、中型（50~74 mm）与大型（75~99 mm）三个规格等级。螺类与牡蛎的种群密度在各规格等级中均匀分配。
为量化无牡蛎捕食螺时的牡蛎死亡率，每个牡蛎密度梯度均设置了对应的空白对照处理（n=2），对照组水槽未放置牡蛎捕食螺；所有对照组均未出现牡蛎死亡情况。野外采集的螺类先暂养1周，使其自由取食牡蛎，以消除从阿巴拉契科拉湾转运至USFCMS带来的运输应激影响。实验开始前设置了5天的饥饿期，该设置参照首次实验中单只饲养螺类的摄食速率，以统一捕食者的饥饿状态。
摄食试验期间，每日对水槽进行两次检查，记录死亡牡蛎的数量以及附着在每只牡蛎上的螺类数量。为维持猎物密度恒定，死亡牡蛎会被替换为同规格等级的活牡蛎，该操作符合捕食者功能响应模型的假设——此类模型量化的是瞬时摄食速率而非累计摄食速率。15天实验周期结束后，研究人员计算得到个体摄食率与总摄食率（每个重复的牡蛎消耗量）。