Limnology and Oceanography: Methods

Offshore industry still largely relies on traditional approaches for regulatory compliance to environmental impact on the water column. Implementing environmental DNA (eDNA) workflow can offer several advantages, but early stages such as sampling and conservation of the samples require standardization and simplification before they can be routinely applied in offshore monitoring programs. In this study, we assessed the effect of several filter types (Durapore disc, Sterivex capsule and Wattera high-capacity capsule; all with 0.22 {micro}m pore size) allowing for different volume of filtration used for sampling eDNA. We also evaluated the effect of 25 days conservation of unfiltered water samples with different preservative solutions (Benzalkonium chloride -BAC, Longmires solution LONGI and a modified Longmires solution without SDS, LNoSDS) as a viable option when immediate filtration and cold storage are not possible. For downstream eDNA evaluation of filter types and preservation, we used quantitative digital PCR on selected target DNA and metabarcoding for qualitative assessment of marine prokaryotic and eukaryotic communities. Overall, filter choice had relatively less effects on quantitative and qualitative information from eDNA compared with water preservation. Sterivex and Durapore were better filter choices for biodiversity assessment. While the Wattera filter allowed processing of larger water volumes and improved quantification of metazoan DNA, handling and processing were more challenging. For water conservation, LNoSDS was the best option. Chemical agents of LONGI and BAC may provide favourable substrates for some tolerant bacterial strains, altering the microbial community composition, with consequences for the overall qualitative evaluation of conserved eDNA. For targeted metazoan eDNA, however, chemical preservation showed clear benefits. This research highlights key considerations and viable options for eDNA sampling and simple preservation workflows without cold storage for implementation in offshore water column monitoring. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=141 SRC="FIGDIR/small/733101v1_ufig1.gif" ALT="Figure 1"> View larger version (50K): org.highwire.dtl.DTLVardef@22a175org.highwire.dtl.DTLVardef@1960864org.highwire.dtl.DTLVardef@1010f49org.highwire.dtl.DTLVardef@92a2f6_HPS_FORMAT_FIGEXP M_FIG C_FIG HighlightsO_LINeed for standardization of eDNA workflow for offshore water column monitoring C_LIO_LIImportance of eDNA sampling (filters) and eDNA conservation (preservatives) C_LIO_LIFilter choice does not affect drastically the dominant eDNA communities C_LIO_LIConservation outside cold storage challenging for eDNA-based biodiversity evaluation C_LIO_LIViable options: Sterivex filter for sampling; Longmires (no SDS) for conservation C_LI

The advancement in deep learning techniques has made Convolutional Neural Networks (CNNs) a powerful tool for the fully automated classification of zooplankton images. In this study, we systematically investigate how network selection, colour information and differences in imaging instruments affect the classification of zooplankton images by comparing multiple state-of-the-art CNNs on images of zooplankton and marine snow from the in situ Continuous Particle Imaging and Classification Sensor (CPICS), Video Plankton Recorder (VPR), In Situ Ichtyoplankton Imaging System (ISIIS), and the on-board Plankton Imager (Pi-10). With differences between models of 7.8 to 19% in F1-score, we find that model selection strongly affects the classification performance, with EfficientNetV2S showing the most reliable overall performance. Moreover, differences between model architectures are largest for the least abundant classes (<100 labeled images), which implies that when these are present, careful model selection is most beneficial. The high image quality of the Pi-10 strongly increases the performance for the least abundant classes compared to the other instruments. In addition, we find a significant correlation (r = 0.597) between ImageNet the performance and F1-score on zooplankton images, which implies that more generally, a model that performs well on ImageNet will perform well for zooplankton classification. Colour information increases the F1-score of the best performing classifier with 2.8%, but provides a stronger benefit (25% F1-score) for classes with <100 images. The overall performance increase of colour information is less than expected and questions the advantage of recording colour information for zooplankton.

Sinking marine particles is a key process regulating carbon export through the biological carbon pump, yet direct measurements of sinking dynamics remain limited in many coastal environments. One barrier is that most existing approaches require expensive instrumentation and large research platforms. Here, we present a low-cost, modular method for concentrating fast-sinking particles and measuring their individual sinking velocities under controlled conditions. This combines large settling tanks (110 L) for field-based particle fractionation with a video-based tracking system that quantifies the sinking behavior of natural marine particles. The particle sinking speed chamber is surrounded on three sides by a temperature-controlled water chamber, minimizing the problem of advection during measurements. The post-processing Python script delivers sinking velocity, particle size, circularity, and RGB-based properties for large numbers of particles. The method accuracy was validated using reference beads with known theoretical sinking velocities derived from Stokes law. Field deployments in the Baltic Sea demonstrated successful enrichment of fast-sinking particles and stable operation from both a research vessel and a small boat. Compared to existing methods, the approach substantially reduces logistical and financial barriers while maintaining particle-resolved measurements and compatibility with complementary biogeochemical analyses. This enables a broader observational coverage of sinking particle processes across environments that are currently underrepresented in carbon export studies.

Chelating ion exchange resins are widely used to eliminate metal interferences in the analysis of ammonium (NH4+) in soil extraction solutions. However, their potential to co-adsorb NH4+ remains underexplored. Here, synthetic metal ion solutions containing 6-30 mg L-1 NH4+ and the metal cations Ca2+, Mg2+, Cu2+, Mn2+, and Zn2+ were treated with Amberlite IRC-748 resin. The resin efficiently removed Ca2+ (-42.2%), Mg2+ (-21.1%), Cu2+ (-99.9%), Mn2+ (-56.9%), and Zn2+ (-93.6%). However, NH4+ losses of 2.2-5.6% were observed, indicating concentration-dependent co-adsorption. While these losses may be acceptable for concentration measurements via routine assays such as photometric analysis, they may still affect the accuracy of high-precision N analyses that rely on quantitative NH4+ recovery. This highlights a methodological caveat for resin-treated samples, especially in low-NH4+ environments. We therefore recommend including recovery assessments and correction factors when using chelating resins to improve accuracy in NH4+ quantification.

Floating debris in the ocean recruits microbes and marine invertebrates to its surface, resulting in rafting communities. As biodegradable plastics increase in prevalence as alternatives to petroleum-based plastics, their properties may impact the dynamics of rafting communities by influencing the composition of bacterial biofilms and attached invertebrates. We compared attached biofouling communities of bacteria and marine invertebrates on biodegradable surfaces and nonbiodegradable petroleum-based plastics and naturally occurring substrata. Six surface types: polypropylene plastic (PP), polystyrene (PS), polylactic acid plastic (PLA), polyhydroxyalkanoates plastic (PHA), maple wood veneer, and propagules of the mangrove Rhizophora mangle, were examined to determine if plastic type affected biofilm composition on the surfaces and their degradation patterns. Biofilm analyses were conducted at twelve weeks, and degradation analyses were conducted at twenty-two weeks of immersion. Using digital image analysis, 16S rRNA sequencing, and metagenomic analyses, we found that microbial biofilms, marine invertebrate inhabitants, and degradation patterns differed across the various substrate types tested. Microbial communities on PLA were more similar to those on the two non-biodegradable plastics, while communities on PLA were more like those on the natural substrata. The biodegradable PLA showed signs of degradation within 22 weeks of immersion, suggesting that biodegradable plastics behave variably in the natural environment. Results of this study bring forth the importance of designing biodegradable plastics with careful consideration of the environmental conditions in which they are likely to persist.

Capelin (Mallotus villosus), a key forage fish in the Northwest Atlantic, links zooplankton to predators including commercial fishes, seabirds, and marine mammals, yet its life-cycle movements in the Gulf of St. Lawrence (GSL) remain poorly understood. Between 2022 and 2024, otoliths from 927 individuals collected during and after spawning were analysed by Laser Ablation Inductively Coupled Plasma Mass Spectrometry (LA-ICP-MS). Building on previous work on regional structuring, seven trace elements (Li, B, Mg, K, Zn, Sr, Ba) were used to discriminate three regions. Early-life regional signatures were inferred through an edge-to-core approach, assigning otolith core chemistry to one of these regions using quadratic discriminant analysis. The core was treated as an integrated early-life signal (late-larval to early juvenile period) rather than a strictly natal signature. Spatial variation in core chemistry was consistent across cohorts, mirroring the stability documented on the otolith edge. Results revealed widespread dispersal alongside partial regional residency: individuals with northeastern early-life signatures showed the strongest correspondence between early-life and capture regions, whereas other regions were more connected. Fish sampled during spawning were more often reassigned to their inferred early-life region than post-spawning fish, a regional-scale homing-like pattern consistent with regional spawning fidelity. This coexistence of dispersive and resident strategies likely generates a portfolio effect buffering the population against environmental variability and localised reproductive failures.

The expansion of offshore renewable energy introduces artificial hard substrate, but the ecological effects may change depending on design features. For example, water replenishment holes, implemented for internal water refreshment and corrosion control, also allow colonisation of previously inaccessible monopile interiors, creating a novel semi-enclosed habitat. This study compared epifaunal communities on the interior and exterior walls of four water replenishment hole-equipped monopiles in the southern North Sea and modelled internal water quality to better understand factors shaping these communities. Vertical video transects were used to quantify percentage cover along depth gradients, while a coupled hydrodynamic-water quality model predicted vertical patterns in dissolved oxygen and particulate organic carbon over one year. Monopile interiors function as semi-enclosed, cave-like habitats with distinct environmental conditions, including darkness, restricted water flow, and vertical gradients in dissolved oxygen and particulate organic carbon. Compared to the external cover, interior communities showed reduced dominance of typical North Sea hard-substrate taxa and increased heterogeneity, with higher contributions of sponges, calcareous tube worms, and brittle stars, resembling communities reported from marine cave environments. Overall epifaunal cover was lower on the interior walls and broadly reflected vertical patterns in water quality. Organic matter accumulated on the interior seafloor, with indications of microbial mat formation. These findings suggest that the internal environmental conditions influence community development. Water replenishment hole design may therefore shape community composition inside monopiles, with implications for possible use as nature-inclusive design and environmental management. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=78 SRC="FIGDIR/small/729839v1_ufig1.gif" ALT="Figure 1"> View larger version (43K): org.highwire.dtl.DTLVardef@17eaed4org.highwire.dtl.DTLVardef@7e631org.highwire.dtl.DTLVardef@99b085org.highwire.dtl.DTLVardef@142376_HPS_FORMAT_FIGEXP M_FIG C_FIG

Inland waters in the Northern Hemisphere are experiencing increased annual runoff due to higher overall precipitation as well as intensified short-term events such as heavy rainfall, floods and storms. These events affect the total loading and variability of inputs of allochthonous, coloured dissolved organic matter (cDOM) and inorganic nutrients into lakes. Previous studies have shown that increased total cDOM and inorganic nutrient loads affect phytoplankton biomass and metabolic rates, but it is unknown how the effects of different cDOM and nutrient pulse scenarios are modified by spatial and seasonal differences in lake characteristics. Here, we conducted a coordinated, standardized mesocosm experiment across three lakes with different ambient cDOM and nutrient concentrations. In two of these lakes, the experiment was implemented in two seasons. The same total amounts of cDOM, nitrate and phosphate were added to all mesocosms, but in pulses that differed in intensity and frequency. We found that pulse intensity and frequency affected chlorophyll a and phycocyanin concentrations and metabolic rates, i.e. gross primary production and respiration, differently. Specifically, more pronounced effects were found in response to the extreme pulse scenario compared to those with more frequent, smaller pulse additions. Furthermore, the effects were mainly temporary and varied more among lakes than between seasons. The clearest differences between the extreme and more gradual runoff scenarios were found in the lake with the lowest background cDOM and nitrate concentrations, likely because lower light limitation and possibly stronger initial N-limitation caused a faster response to the nutrient addition. Our results highlight that both antecedent lake conditions and characteristics of runoff events can affect phytoplankton biomass and metabolic rates and that comparative experimental approaches are needed to reveal the complexity of the responses.

Metagenomic characterization of low-biomass Yellowstone National Park (YNP) hot spring waters remains challenging because microbial recovery is influenced by filtration methodology, sample preservation, DNA extraction, and sequencing strategy. We characterized thermophilic microbial communities in alkaline YNP hot spring waters (62-90.5{degrees}C) using three high-temperature-compatible filtration systems (Sterivex, Supor, and polycarbonate membranes), automated onsite DNA extraction ({micro}Titan), and shotgun metagenomic sequencing with Illumina short-read and Oxford Nanopore Technologies (ONT) long-read platforms. Across all filtration systems and sequencing workflows, microbial communities were consistently dominated by Bacteria ([~]90% of reads), whereas Archaea represented <10% of recovered sequences. Dominant microbial populations were reproducibly recovered across all approaches; however, recovery of lower-abundance taxa varied among methods. This variability was most evident in polycarbonate-filtered samples, which exhibited greater replicate-to-replicate variation and less consistent detection of microbial species. Thermocrinis ruber and related Aquificae-associated thermophiles dominated the hottest waters (78.5-90.5{degrees}C), whereas warmer effluent-channel waters (63.5-66.5{degrees}C) contained T. ruber together with photosynthetic taxa, including Synechococcus spp. and Candidatus Thermochlorobacter aerophilum. Archaeal communities were primarily represented by Pyrobaculum- and Thermoproteus-related taxa. Non-metric multidimensional scaling analyses indicated that overall community structure was largely unaffected by filtration or sequencing methodology, whereas alpha-diversity metrics showed that filter selection influenced richness and diversity estimates. These findings identify field-deployable workflows for metagenomic characterization of low-biomass thermophilic aquatic systems and demonstrate the importance of integrating filtration and sequencing strategies for studying extremophile microbiomes under remote sampling conditions. IMPORTANCEAccurate characterization of low-biomass geothermal water microbiomes remains challenging because microbial recovery is strongly influenced by sample handling, filtration efficiency, DNA extraction chemistry, and sequencing methodology. This study demonstrated that Yellowstone National Park alkaline hot spring water microbiomes were consistently dominated by Bacteria (>90% of recovered reads), whereas Archaea represented <10% of community abundance across all filtration systems. Although dominant microbial populations were reproducibly recovered, filtration-device selection influenced the recovery of microbial diversity and low-abundance taxa. By integrating field-deployable onsite DNA extraction with ONT shotgun metagenomic sequencing, this work evaluates practical workflows for studying thermophilic planktonic microbial communities under remote field conditions. These findings are relevant, not only to geothermal microbiology, but also to low-biomass environments in medical, pharmaceutical, and aerospace industries, where rapid onsite processing and contamination-aware workflows are essential for preserving authentic microbial signatures in extreme environments.

Freshwater salinization is of increasing concern for integrity and functioning of freshwater habitats worldwide. Experiments so far often have studied drastic salt additions, while gradient designs have been performed less commonly. We tested the effect of freshwater salinization in a mesocosm exposing the plankton community of the oligotrophic Lake Lunz, Austria, to a four-fold salinization gradient (control, 0.2, 1, a 5 ppt salt). Salinity was manipulated in a factorial design with enrichment, with 10 g L-1 and 30 g L-1 phosphorus, resulting in 8 treatments with 3 replicates each. We followed the effects of salinization on diversity, community composition and resource use over 36 days. Community composition was assessed by amplicon sequencing, Diversity loss and community turnover followed upon salt addition. All levels of salinization caused pronounced changes in community composition, with 5 ppt causing the most drastic changes. Salinization caused trophic downgrading by kicking out especially protistan consumers and rotifers, while some green algae and chrysophytes were especially tolerant, resulting in reduced phylogenetic and functional diversity with increasing salinization. In line with reduced top down control, salinization affected temporal variability in chlorophyll-a (chl-a) and resource use (RUE), with higher salinity causing more extreme fluctuations in chl-a and RUE. Enrichment overall aggravated salinization, enhancing temporal turnover and temporal fluctuations in resource use.

Fixed larval specimens often shrink and curve, making length measurement labor-intensive. Although recent studies have demonstrated efficient fish-length estimation from images using deep learning, methods for estimating curved length remain limited. Furthermore, although deep learning is a powerful method for object detection in images, an essential step for length measurement, it requires preparing large amounts of training data, which can hinder practical implementation. In this study, we used a zero-shot model that requires no training to detect fish in an image. The curved length was then estimated using an image-processing approach that combines image thinning with Bezier curve approximation, and its accuracy was evaluated. We analyzed 1,040 larvae from five tuna species captured in stereomicroscope images. Manual measurements (notochord length or standard length; 1.5-8.5 mm) were conducted by two measurers and served as reference values. Fish regions were detected using GroundedSAM, and curved body centerlines were extracted through image thinning and approximated with Bezier curves. The curve length was used as the estimated body length. Estimation accuracy was assessed using bias and standard deviation between estimated and manual measurements. GroundedSAM detected all 1,040 fish, although there were 49 overdetections. Overdetection was caused by the double-detection of a single fish or by the misidentification of debris and light reflections as fish. Although the standard deviation of the differences between manual measurements and the image analysis-based (IAB) method was larger than the inter-measurer differences, the bias for [&le;]5 mm was comparable to or smaller than the inter-measurer bias. According to the strength-of-agreement criteria for the concordance correlation coefficient, the IAB method demonstrated substantial agreement in the [&le;]5.0-mm range. The IAB method accurately measured most curved tuna larvae without prior training, particularly in the [&le;]5.0-mm range. Combining the IAB method with manual remeasuring can improve the efficiency of curved-length measurement tasks.

The prevalence of nitrogen limitation and nitrogen-phosphorus co-limitation (henceforth referred to as nitrogen-related limitation) in Norwegian lakes and their relationships with atmospheric nitrogen deposition, climate, dissolved organic matter (DOM), and catchment characteristics were assessed across space and time. Routine monitoring data from 1,529 lakes in the national Vannmiljo database were analyzed for two multi-year periods (1995-2009 and 2010-2025). Limitation was inferred using the molar NO--N/TP ratio as an indicator of dissolved inorganic nitrogen availability. Nitrogen-related limitation was widespread in both periods and exhibited strong regional structure, with highest prevalence in northern regions and lowest prevalence in southwestern Norway. Overall prevalence increased from 31% to 38% between periods, with significant increases in western regions. Regional-scale models identified climate, forest cover, DOM, agriculture, and atmospheric nitrogen deposition as predictors of limitation probability, whereas study period per se and bog/peatland cover were not significant. At the local scale, atmospheric nitrogen deposition and DOM were the only consistent predictors, with substantially lower explanatory power than at the regional scale. These results indicate that large-scale environmental gradients play a major role in shaping nutrient stoichiometry in Norwegian lakes. Because the monitoring dataset primarily represents lakes affected by human activities, the findings are particularly relevant for water management. The widespread occurrence of nitrogen-related limitation suggests that nitrogen availability may influence phytoplankton growth in many systems and that dual-nutrient management strategies addressing both nitrogen and phosphorus may be required in many regions.

Environmental DNA (eDNA) has the potential to be a powerful tool in blue carbon science for characterizing and quantifying the contribution of marine macrophytes; but its complex, dynamic relationship with bulk biomass is poorly understood. Here, we used eDNA to examine the degradation dynamics of sugar kelp (Saccharina latissima) in muddy, anaerobic marine sediment. This involved three 16-week incubations; with additions of lyophilized sugar kelp alone, a mix of lyophilized marine macrophytes including sugar kelp, and sugar kelp holdfasts buried in sediment. We used species-specific digital polymerase chain reaction assays for mitochondrial, chloroplast and nuclear markers, and metabarcoding for the 16S and 18S ribosomal RNA genes. In the former two incubations, all sugar kelp eDNA markers showed rapid log exponential declines (up to 98-99%) to asymptotes greater than the unamended controls, even as part of a more complex mix of macrophytes. In contrast, for the buried kelp holdfasts, sugar kelp eDNA increased to an asymptote (by up to [~]15X), which may be reflective of the different nature of added biomass. Overall, we demonstrate substantial preservation of environmental DNA and total organic carbon under anaerobic conditions, and the potential to use environmental DNA to quantify biomass in a blue carbon context.

Seagrass ecosystems are experiencing rapid global decline, and current seed-based restoration efforts are frequently hindered by severe early-life bottlenecks such as decapod predation and bioturbation. To bypass these ecological feedbacks, encasing Zostera marina seeds in protective clay matrices (seed balls) presents a promising, low-cost intermediate technology for scalable deployment. To refine this approach, a mesocosm experiment was conducted to evaluate the physical integrity, germination success, and early shoot development of Z. marina seeds across various clay formulations (fireclay, bentonite, and Kettering loam) and drying treatments. Seed balls were deployed either on the sediment surface or buried, and their performance was evaluated using a hurdle-style generalised additive modelling approach. Results indicated a clear trade-off between structural robustness and biological performance. While naked seeds (70.3% emergence) and buried fireclay balls (67.2% emergence) exhibited the highest overall emergence and post-emergence shoot length, surface-deployed seed balls formulated with higher proportions of bentonite and loam (e.g., 2:4:1 ratio) demonstrated the superior structural integrity necessary for resisting degradation during deployment. However, these denser matrices significantly reduced emergence probabilities and restricted shoot development compared to naked seeds. Formulations lacking loam (1:3) performed the poorest across all biological and structural metrics. Despite reduced emergence relative to buried treatments, surface-deployed seed balls still achieved high viability, outperforming typical field germination rates. Although further field-based refinement is required to test formulations under dynamic hydrodynamic and bioturbation pressures, seed balls offer a scientifically robust, scalable mechanism to bypass early-stage predation and support widespread seascape recovery.

Coastal wetlands of the Laurentian Great Lakes support abundant populations of fish, invertebrates, and vegetation, though the trophic linkages connecting primary production and lower consumers is not well understood in these systems. We implemented a multiple-tracer approach to evaluate trophic pathways, pairing traditional food web isotope tracers like carbon ({delta}13C) and nitrogen ({delta}15N) with total mercury concentrations (THg). We predicted that filamentous algae would be the dominant energy resource in the diet of lower trophic-level invertebrates in the Grand River Estuary, a network of riverine coastal wetlands adjacent to Lake Michigan. In addition, we predicted that adding THg as a tracer would improve the resolution of our food web models by clarifying trophic levels and relationships between wetland species. Four basal energy sources were sampled, including filamentous algae, emergent macrophytes, submersed macrophytes, and phytoplankton, along with organic detritus. Aquatic invertebrates were sampled across multiple functional guilds to represent primary and secondary consumers and included amphipods and odonates. Our findings suggest that organic detritus is the dominant resource responsible for energetically supporting these lower trophic levels in the Grand River estuary, although submersed macrophytes were important alternative energy sources for secondary consumers. THg concentrations enhanced the resolution of dietary contribution estimates in MixSIAR models applied to consumer and source data. Isotope biplots revealed that THg concentrations were a more reliable predictor of trophic position than {delta}15N in Grand River Estuary (GRE) sites. This methodology has important implications for future food web studies in complex ecosystems such as coastal wetlands and demonstrates the novel use of mercury as an ecological tracer in a Bayesian mixing model approach.

Freshwater monitoring programmes like the European Unions Water Framework Directive (WFD) provide a wealth of data on European lake status, including water quality and macrophytes (aquatic plants) as critical habitat features that support health of humans and wildlife. Easier WFD data access can enable external management and research to better safeguard human and natural freshwater use. We demonstrate a replicable workflow to easily download and process multi-year (2007-2024) observations of lake macrophytes (425 sites) and complementary water quality variables (202 sites) from Swedish WFD data. Then, we illustrate the value of improved data access to address ecological questions that drive conservation, investigating how spatial scales influence macrophyte richness and associated water quality relationships using a spatial random intercept model. Decomposing the spatial intercept links small scales (<10 km) to site-level gradients and large scales (>100 km) to biogeographical drivers. Stochastic and environmentally-structured processes coexisted at intermediate scales (10-100 km). Adding water quality rarely improved overall predictive performance of macrophyte diversity models but consistently influences the role of different spatial scales. Water quality variables showed consistent spatially structured variation at intermediate scales and unique spatial patterns in tandem, overlapping with large-scale biogeographical influences. Altogether, we show context-dependencies for spatial model interpretation and provide guidance in accounting for spatial confounding to improve inferential and predictive performance. Our workflow and results show a clear way forward for accessing high-quality macrophyte and water quality data sets and their utility for addressing ecological questions that guide macrophyte protection under the WFD. HighlightsO_LIyears Swedish of macrophyte and water quality monitoring data were extracted. C_LIO_LIrichness showed scale-specific patterns linked to geographic gradients. C_LIO_LIbest predictive models for richness had no water quality at all. C_LIO_LIoverlap in their spatial scales and must be carefully separated. C_LIO_LIpen access data and multiscale analysis can apply to many ecological questions. C_LI

Missing data is a fundamental challenge in space biology, where high experimental costs, limited sample availability, and tissue allocation constraints produce datasets that are sparse, multimodal, and heterogeneous. We present a systematic four-stage framework for diagnosing, implementing, and validating data imputation strategies tailored to these characteristics, and demonstrate its application to retinal imaging and omics data from the NASA Rodent Research 9 (RR9) mission. Using logistic regression-based missingness diagnosis, we identify a Missing At Random (MAR) mechanism driven by experimental design constraints across nine assay modalities. We implement and optimize three imputation strategies: K-Nearest Neighbors (KNN), Multiple Imputation by Chained Equations with weak ElasticNet regularization (MICE-Elastic), and a per-column hybrid strategy, evaluated against a random sample imputer baseline. Validation across seven complementary metrics including supervised classification, unsupervised clustering, correlation structure preservation, masked value recovery, cross-dataset generalization, and permutation testing reveals that MICE-Elastic and the Hybrid strategy preserve genuine biological signal in both RNA-seq and TUNEL modalities, while KNN and the random sample imputer do not despite achieving comparable cross-validation accuracy. A critical finding is that imputation substantially improves supervised classification performance while consistently degrading unsupervised clustering structure, a trade-off researchers must understand before applying these methods. This framework provides practical, actionable guidance for space biologists and data scientists managing sparse multimodal datasets, and represents a foundational step toward digital twin development for space medicine.

Environmental DNA (eDNA) surveys are increasingly used to assess marine biodiversity and inform deep-sea environmental decision-making, including mineral resource management and fisheries oversight. Yet standard low-volume protocols inherited from coastal work may be inadequate at depth, and no quantitative framework links depth and ecosystem context to defensible filtration volume targets. We compiled 841 eDNA samples from eight expeditions across the North Atlantic, Wider Caribbean, and Pacific (surface to 4000 m) to quantify how recoverable eDNA scales with depth and surface productivity, and to derive depth- and productivity-aware sampling targets. Total eDNA concentration declined with depth as a power law, with attenuation exponents (b) modulated by surface productivity: most gradual in eutrophic waters (b = 0.67), intermediate in mesotrophic (b = 0.90), and steepest in oligotrophic systems (b = 1.25); volume-weighted models explained 66-88% of the variance. At a fixed extract-concentration target, required filtration volumes diverged ~7-fold between oligotrophic and eutrophic systems at 200 m and ~38-fold at 4000 m. Conventional Niskin sampling, therefore, undersamples deep-sea biodiversity, particularly in mid- to low-productivity systems. Among laboratory parameters, the assay-specific extract-concentration target exerted greater leverage on required volume than extraction efficiency or elution volume. Volume-aware sampling paired with optimized recovery should be routine in deep-sea eDNA surveys.

Arctic ecosystems are critically endangered by rising temperatures and changing hydrography, especially the intrusion of increasingly warm water from the Atlantic Ocean known as Atlantification. In addition to housing fragile biodiversity, Arctic copepods and their lipids play a crucial role in cycling carbon by transporting carbon into the deep ocean through their diapausing behaviors. Here, we explored the lipidomes of the Arctic copepod Calanus glacialis, collected from three fjords around Svalbard during November 2022 when C. glacialis are known to be in diapause. These three field sites provide a natural laboratory experiment, as they are influenced by different water masses with varying degrees of Atlantic water, and experience vast differences in sea ice coverage over the year. These environmental differences were clearly reflected in the lipidomic analysis, with stations influenced most by Atlantic Warm Water having the lowest total lipid concentrations and the lowest accumulation of storage lipids necessary for entering diapause. Membrane lipids were a significant proportion of the Svalbard copepod lipidomes, with the highest ratios observed at the Atlantified site. The high membrane lipid and high triacylglycerol concentrations were interpreted as signs of active feeding. This was further corroborated by fatty acid composition analysis, which revealed dietary biomarkers of carnivory at Atlantified sites. The copepods from the site most insulated from Atlantic influence had more than double the amount of storage lipids per individual and fatty acids associated with diatom biomass, indicating assimilation in the spring. Ultimately, the decrease in lipid content observed in association with Atlantification around Svalbard will impact diapause patterns, as Calanus species need 20-30% more WE to successfully complete diapause. In turn, this will impact the magnitude of carbon sequestration through the seasonal lipid pump, not to mention having radiating effects through the Arctic food web where Calanus glacialis plays an important role. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=131 SRC="FIGDIR/small/738257v1_ufig1.gif" ALT="Figure 1"> View larger version (49K): org.highwire.dtl.DTLVardef@fc9321org.highwire.dtl.DTLVardef@1f6336org.highwire.dtl.DTLVardef@aaa3a1org.highwire.dtl.DTLVardef@dcd2a8_HPS_FORMAT_FIGEXP M_FIG C_FIG

Artificial intelligence (AI)-based image analysis can efficiently estimate fish length, but differences in devices, imaging conditions, operators, and AI models limit comparability among surveys. We propose a standardization framework that estimates a bin-specific error matrix from paired reference measurements and AI-derived lengths and applies it to standardize (correct) AI-derived length-frequency distributions. The Richardson-Lucy expectation-maximization algorithm was used, with the number of iterations selected via cross-validation. Simulations based on empirical length-frequency data from 110 species showed that standardization reduced relative bias and distributional discrepancy; median relative-bias and root mean square error ratios were below 1, and the performance was more affected by the amount of paired data than by the number of cross-validation folds. In real data from 957 Japanese jack mackerel, standardized AI-derived distributions approached human-observer histograms, although discrepancies remained in the range of 160-230 mm. The proposed framework provides a practical approach for improving the comparability of image-derived length-frequency data using paired calibration data, without retraining the underlying AI model.