Uthicke Sven, Robson Barbara, Doyle Jason R, Logan Murray, Pratchett Morgan S, Lamare Miles
Australian Institute of Marine Science, PMB No. 3, Townsville MC, QLD 4810, Australia.
Australian Institute of Marine Science, PMB No. 3, Townsville MC, QLD 4810, Australia.
Sci Total Environ. 2022 Dec 10;851(Pt 1):158143. doi: 10.1016/j.scitotenv.2022.158143. Epub 2022 Aug 19.
Outbreaks of the corallivorous Crown-of-Thorns Seastar (CoTS) Acanthaster cf. solaris contribute significantly to coral reef loss. Control of outbreaks is hampered because standard monitoring techniques do not detect outbreaks at early (low density) stages, thus preventing early intervention. We previously demonstrated that eDNA monitoring can detect CoTS at intermediate densities. Here, we test whether detection probability can be improved by (i) targeted site selection or collection at specific times and (ii) moving from an average eDNA copy number approach (based on the limit of quantification) to a presence/absence approach (based on the limit of detection). Using a dataset collected over three years and multiple reef sites, we demonstrated that adding water residence age, sea surface level and temperature into generalized linear models explained low amounts of variance of eDNA copy numbers. Site specific CoTS density, by contrast, was a significant predictor for eDNA copy numbers. Bayesian multi-scale occupancy modelling of the presence/absence data demonstrated that the probability of sample capture (θ) on most reefs with intermediate or high CoTS densities was >0.8. Thus, confirming CoTS presence on these reefs would only require 2-3 samples. Sample capture decreased with decreasing CoTS density. Collecting ten filters was sufficient to reliably (based on the lower 95 % Credibility Interval) detect CoTS below nominal outbreak levels (3 Ind. ha). Copy number-based estimates may be more relevant to quantify CoTS at higher densities. Although water residence age did contribute little to our models, sites with higher residence times may serve as sentinel sites accumulating eDNA. The approach based on presence or absence of eDNA facilitates eDNA monitoring to detect CoTS densities below outbreak thresholds and we continue to further develop this method for quantification.
食珊瑚的棘冠海星（Acanthaster cf. solaris）爆发对珊瑚礁的损失有重大影响。由于标准监测技术无法在早期（低密度）阶段检测到海星爆发，从而无法进行早期干预，这阻碍了对爆发的控制。我们之前证明了环境DNA（eDNA）监测可以检测到中等密度的棘冠海星。在此，我们测试是否可以通过以下方式提高检测概率：（i）在特定时间进行有针对性的位点选择或样本采集，以及（ii）从基于定量限的平均eDNA拷贝数方法转变为基于检测限的存在/不存在方法。利用在三年时间里从多个珊瑚礁位点收集的数据集，我们证明在广义线性模型中加入水停留时间、海平面和温度只能解释eDNA拷贝数中少量的方差。相比之下，特定位点的棘冠海星密度是eDNA拷贝数的一个重要预测因子。对存在/不存在数据进行的贝叶斯多尺度占有率建模表明，在大多数中等或高密度棘冠海星的珊瑚礁上，样本捕获概率（θ）>0.8。因此，要确认这些珊瑚礁上存在棘冠海星，只需要2至3个样本。样本捕获概率随着棘冠海星密度的降低而下降。收集10个滤膜就足以可靠地（基于95%较低可信度区间）检测到低于名义爆发水平（3只/公顷）的棘冠海星。基于拷贝数的估计可能更适用于对较高密度棘冠海星进行量化。尽管水停留时间对我们的模型贡献不大，但停留时间较长的位点可能作为积累eDNA的哨兵位点。基于eDNA存在或不存在的方法有助于eDNA监测以检测低于爆发阈值的棘冠海星密度，我们将继续进一步开发这种量化方法。
