Abstract Despite the widespread need to map the abundance of invasive quagga and zebr mussels ( Dreissena spp.) in lakes and reservoirs worldwide, current methods remain limited by data collection and analysis bottlenecks. Autonomous underwater vehicles (AUVs) combined with machine learning (ML) offer a scalable remote sensing approach for estimating mussel density (individuals m −2 ) across large spatial domains and a range of habitats that are difficult to sample with current techniques. We present MusselFinder: a deep neural network (DNN)‐based pipeline that detects and enumerates live mussels in AUV‐collected benthic images. MusselFinder estimates live mussel percent cover and density in nadir monocular images captured at known altitudes above the lakebed. In images of suitable clarity, MusselFinder achieved strong pixel‐level localization (mussel‐class Intersection over Union [IoU] > 0.76) and density prediction (log‐density R 2 = 0.79–0.95) over a variety of bottom types. AUV‐based image acquisition is more than three orders of magnitude faster than physical sampling by divers or ponar grabs, and DNN‐based analysis can proceed at rates comparable to image acquisition. We used MusselFinder to analyze over 50,000 spatially referenced images at a spawning reef in Lake Michigan, representing a shift from small‐scale, diver‐collected datasets to large‐scale monitoring in unconstrained environments. The method can be adapted to new imaging conditions using a modest training set and a user‐friendly point‐labeling technique, supporting broader‐scale monitoring of one of the world's most invasive invertebrates.
Abstract The integration of high‐throughput image–based flow cytometry into routine phytoplankton assessment requires rigorous evaluation against established microscopy techniques. As method agreement is often site‐ and species‐specific, case‐by‐case validation is essential to successfully adopt emerging methods. In this study, we evaluated the agreement between two methods, FlowCam and inverted microscopy, for the analysis of phytoplankton communities from nine small freshwater kettle holes located in an agricultural landscape. Specifically, we compared automated FlowCam–derived particle properties with cell size measurements and derived biovolume estimates obtained by classical inverted microscopy across the whole phytoplankton size range. We found good agreement between one or more automated FlowCam dimensional estimates and microscopy measurements for most phytoplankton groups. However, morphologically more complex and particularly elongated taxa showed substantial discrepancies between methods. Accordingly, dimensional measurements translated into representative biovolume estimates for morphologically simple taxa, especially using prolate spheroid approximations, were similar between both methods. Divergent measurements of FlowCam‐derived widths introduced systematic biases, mainly for cylindrical approaches of biovolume calculations based on geodesics and markedly overestimated biovolumes of filamentous taxa. The findings demonstrate that the choice of FlowCam‐derived particle properties must be morphology‐specific and validated against microscopy. Although cylindrical biovolume estimates require refinement, Legendre‐based approaches performed well, supporting the integration of FlowCam into monitoring programs and experimental studies for diverse phytoplankton communities in small water bodies like kettle holes.
Increasing evidences show that fauna living in deep-sea ecosystems exhibit various complex behaviors that remain understudied. Although several in situ observatories have been set up around the globe, the use of their imagery data to gain insights into animal behavior across temporal scales remains constrained by the labor-intensive nature of the analyses. Here, we developed an automated deep-learning-based pipeline with multi-target tracking. Our pipeline integrates YOLOv8 object detection with category-specific approaches to characterize spatial-temporal behavior patterns (i.e., mobility and trajectory). This pipeline was applied to video imagery from the Lucky Strike hydrothermal vent field, Mid-Atlantic Ridge as a case study. We tailored tracking methods to three functional mobility groups: Sedentary, Simple Motile, and Complex Motile, represented by the mussel Bathymodiolus azoricus, the shrimp Mirocaris fortunata, and the crab Segonzacia mesatlantica, respectively. Tracking performance was evaluated using standardized multi-object tracking metrics: the Hungarian algorithm showed strong manual agreement for Sedentary taxa (r = 0.792), BoTSORT excelled for Simple Motile (HOTA = 0.956, 0.012% identity switches), and DeepSORT achieved best HOTA = 0.727 for Complex Motile. Our results reveal: (1) daily displacement of the mussels, (2) a potential semi-lunar rhythmic pattern in the movement of the shrimps (14.76-day periodicity) modulated by the proximity to vent emissions and seasonal factors, and (3) acute stress responses in the crabs linked to localized thermal exposure to hot vent fluids. Our pipeline provides a scalable detection-tracking-behavior analysis framework, enabling statistical identification of infra-monthly behavioral patterns that are difficult to quantify through manual annotations alone. As a result, it can be readily adapted to other deep-sea ecosystems for standardized multi-species behavioral monitoring.
Abstract The recovery of amplifiable DNA from formaldehyde‐fixed (FF) zooplankton samples has long been considered problematic, but advances in degraded‐DNA retrieval have renewed interest in FF samples. To access the information stored in long‐term zooplankton time series, we evaluated methods for extracting amplifiable DNA from community samples preserved for up to 28 yr in formaldehyde at room temperature. A method previously reported as successful in FF zooplankton stored at 4°C proved ineffective here, likely due to the differing storage conditions. In contrast, by adapting two protocols developed for FF museum specimens—a harsher (Hahn–hot alkaline [HHA]) and a gentler (Hahn–proteinase K column [HPC]) extraction approach—we were able to amplify and sequence a subset of samples. As expected, DNA integrity and sample pH decreased with preservation time, and only short DNA fragments were recoverable, ruling out standard ≥ 300 bp metabarcoding markers. While DNA integrity appeared to be a better predictor than DNA yield for amplification success, the presence of a gel band of the expected size did not always guarantee congruence with microscopy assessments. Although amplifiable DNA was recovered from most samples, including some of the oldest, community compositions concordant with microscopy were consistently recovered only from samples preserved for up to 2 yr. Beyond this point, the HHA and HPC methods produced divergent results, reflecting a trade‐off between removing formaldehyde‐induced cross‐linkages and avoiding additional DNA damage. Among the three small universal markers tested (~ 120–170 bp), only the 18S rRNA V9 region consistently amplified. We conclude by providing recommendations aimed at improving the methods assessed here.
Abstract While it is well recognized that suspended particles can bias titration‐based total alkalinity (TA), whether addition of minerals for alkalinity enhancement introduces systematic TA artifacts remains poorly quantified. To address this gap, we quantified the contribution of ultrafine particles (< 0.2 μ m) to TA titration introduced by limestone (0.1, 0.5, 1, and 2 g L −1 ) under two p CO 2 conditions (420 and 1495 ppm). We found that the TA increase is much higher than the theoretical increase determined by [Ca 2+ ] when added limestone is over 0.1 g L −1 . However, this excess TA is negligible when the added mineral particles exclude ultrafine particles. Also, the particulate TA decreases when there is more mineral dissolution under the high p CO 2 treatment. These findings suggest that even after 0.2 μ m filtration, traditional TA titration of mineral‐treated water may still overestimate true mineral dissolution, potentially leading to inflated carbon dioxide removal estimates.
