Green-lipped mussel associations with macroalgae on a mussel reef in New Zealand

The goal of this study was to develop and test a tool that has the potential to aid restoration practitioners in the selection of suitable natural substrates for overcoming substrate limitation at mussel reef restoration sites in New Zealand and potentially elsewhere across the globe. This study aimed to; 1) create an index that scores the morphological features of macroalgae for their potential to support juvenile mussel recruitment based on prior published data, 2) compare the relative performance of various macroalgae species with differing morphology as attachment substrates for three size classes of juvenile green-lipped mussels or Perna canaliculus on two remnant mussel reefs, and 3) use these findings to evaluate the accuracy with which the index can predict the presence of attached juvenile mussels among macroalgal substrates based on the morphology of the macroalgae. This data details the presence of three different size classes of juvenile green-lipped mussels (i.e., <10 mm, 10 - <20 mm, and 20 - <30 mm in shell length) attached to different species of macroalgae sampled on two remnant intertidal mussel reefs in northeastern New Zealand (Waipu Cove and Pakiri Beach). The presence of juvenile mussels in each size class on macroalgal substrates is reported as total counts and as the proportion (%) of each species of macroalga observed during the reef sampling that exhibited juvenile mussel attachments. The macroalgae species that supported the highest juvenile mussel presence per mussel size class have the most promise for use in overcoming substrate limitation at mussel reef restoration sites.
Scores (1-3) for each of the eight morphological features of macroalgae (i.e., canopy cover, canopy height, holdfast complexity, clustering frequency, planes of branching, branch spacing, degree of branching, and branch width) were generated for each species of macroalga by the Macroalgal Morphology Index (MMI). These scores were tested against the presence of juvenile mussels attached to each species of macroalga using a linear discriminant analysis (LDA) to determine the accuracy of the MMI's scoring criteria. The LDA revealed that collectively, the index scores for each morphological feature were able to predict the likelihood of mussel attachments to macroalgae for juveniles <10 mm in shell length with 75% accuracy but with only 40-60% accuracy for juveniles 10 – <30 mm. Additionally, a 10-fold cross-validation was employed to assess the relative importance of each morphological feature of macroalgae as a predictor of juvenile mussel presence. This revealed that holdfast complexity and canopy cover were the strongest predictors of attached juvenile presence for all three mussel size classes, and planes of branching, branch width, degree of branching, and branch spacing were generally strong predictors of the presence of attached juvenile mussels <10 mm.

本研究旨在开发并测试一款工具，该工具可帮助修复从业者为新西兰贻贝礁修复场地乃至全球其他潜在区域，筛选可用于克服基质限制问题的天然基质。
本研究的具体目标包括：1）基于已发表的前期研究数据，构建一项评分指数，用于评估大型藻类（macroalgae）的形态特征对幼贻贝附着定植的潜在支持能力；2）在两处残存贻贝礁上，对比不同形态特征的多种大型藻类物种，作为三种规格幼绿唇贻贝（*Perna canaliculus*）附着基质的相对性能，三种规格分别对应壳长<10 mm、10~<20 mm以及20~<30 mm的幼贻贝；3）利用上述研究结果，评估该指数基于大型藻类形态特征，预测大型藻类基质上附着幼贻贝存在情况的准确性。
本数据集详细记录了在新西兰东北部两处残存潮间带贻贝礁（怀普湾和帕基里海滩）采集的不同大型藻类物种上的幼贻贝附着情况。研究同时报告了各规格幼贻贝在大型藻类基质上的附着总数量，以及采样期间观察到的各大型藻类物种中存在幼贻贝附着的比例（%）。
在各贻贝规格下，附着幼贻贝数量最多的大型藻类物种，最具备应用潜力，可用于解决贻贝礁修复场地的基质限制问题。
研究人员通过大型藻类形态指数（Macroalgal Morphology Index, MMI），为每种大型藻类生成了8项形态特征的1~3分评分，这8项特征分别为冠层覆盖度、冠层高度、固着器复杂度、簇生频率、分枝平面、分枝间距、分枝程度以及分枝宽度。
研究采用线性判别分析（Linear Discriminant Analysis, LDA），将上述评分与各大型藻类物种上附着的幼贻贝存在情况进行比对，以验证MMI评分标准的预测准确性。分析结果显示，综合各项形态特征的评分，能够以75%的准确率预测壳长<10 mm的幼贻贝在大型藻类上的附着概率，但对于10~<30 mm的幼贻贝，预测准确率仅为40%~60%。
此外，研究采用10折交叉验证（10-fold cross-validation）来评估各项大型藻类形态特征作为幼贻贝附着预测因子的相对重要性。结果显示，固着器复杂度和冠层覆盖度是三种规格幼贻贝附着情况的最强预测因子；而分枝平面、分枝宽度、分枝程度以及分枝间距通常是预测壳长<10 mm幼贻贝附着情况的强预测因子。