Data on macrobenthic prey from an essential western gray whale feeding habitat, Sakhalin Island, Russia, 2001–2015

Data in this article presents data (means and standard deviations) for prey biomass from essential feeding habitats for the endangered western gray whale. Prey include Actinopterygii (primarily the sand lance Ammodytes hexapterus), Amphipoda, Bivalvia, Cumacea, Isopoda, and Polychaeta. Total prey biomass (sum of the six prey groups) is also presented. Statistical analyses document spatial and temporal trends in prey biomass concentrations. Multivariate analyses using canonical correspondence analysis characterize relationships of potential drivers of community changes.


Data
Macrobenthic community biomass was determined from the northeastern Sakhalin Island gray whale feeding area, Sea of Okhotsk, Russia [1]. Bottom samples were collected from 2001 to 2015 to determine prey biomass characteristics and distributions. Biomass data presented here include 6 prey (Actinopterygii (primarily sand lance Ammodytes hexapterus), Amphipoda, Bivalvia, Cumacea, Isopoda, and Polychaeta), and total prey (sum of the six prey categories). Average biomass, sample size, and standard deviations are presented for the nearshore (2001e2015) and offshore (2001e2015) study areas ( Fig. 1; Tables 1 and 2). Biomass values are also presented for feeding points (opportunistically sampled locations where whales were observed feeding) to determine if feeding areas had different biomass characteristics than other stations (Table 3). ANCOVA and Tukey multiple comparisons document long-term differences for benthic prey among years for the nearshore and offshore surveys (Tables 4e6). Canonical correspondence analyses were conducted to investigate relationships between environmental predictors and community biomass structure using all macrofaunal groups ( Fig. 2; Table 7) (see: Table 8).

Sampling
Bottom sampling was conducted in the summers of 2001e2015 to measure biomass concentrations and environmental characteristics in the GW feeding area adjacent to northeastern Sakhalin Island (the Sakhalin feeding area; Fig. 1). Sampling was conducted aboard the R/V Okean (2001), R/V Nevelskoy Specifications table   Subject area Biology, marine ecology More specific subject area Marine macrobenthic ecology; gray whale prey biomass, spatial and temporal variability of marine benthic communities Type of data Value of the data Data summarize benthic prey biomass concentrations from essential habitat for the endangered western gray whale. Temporal variations in prey biomass concentrations are large and a significant source of variability in prey biomass. Amphipoda biomass, the primary prey for gray whales, demonstrated significant declines over time.
Understanding prey dynamics contributes to a better understanding of sources for change in gray whale populations, particularly for the endangered western gray whale. The Sea of Okhotsk is undergoing significant climate-related ecosystem changes and the macrobenthic data provide biomass concentrations that support a more complete understanding of the extent of climatic influences. Research vessel drafts limited operations of van Veen grabs to water depths !9 m though some slightly shallower samples (7e8 m) were occasionally collected. Onboard, samples were sieved over a 1.0-mm-mesh screen and the organisms preserved in 4% formalin. In the taxonomic laboratory, biological material was sorted from the sediment residues and animals were identified, counted, and weighed. Animals were grouped into classes or higher taxonomic categories. Sediment grain-size analyses were used to determine standard grain-size categories. Initial investigation of the nearshore feeding area was conducted in 2001 by divers to explore prey habitat (5e30 m water depth) and provide a basis for designing the nearshore Piltun survey grid. The current nearshore sampling grid consists of 72 cells along the northeastern coast of Sakhalin Island encompassing the nearshore (<20 m) GW feeding habitat and extending to deeper waters to capture environmental gradients. The total area of the nearshore survey area is approximately 1100 km 2 . Sampling was initiated in the offshore feeding area in 2002 and the offshore survey currently includes 48 cells with a total area of approximately 2000 km 2 . Data records for diver sampling in 2001 consists of single biomass estimates for sampling location. Benthos and sediment sample collections from 2002 to 2015 comprised three replicates collected at randomly selected sampling points or by repeated sampling of locations selected in previous years. Samples from 2002 to 2015 were collected in water Table 1 Sample size, averages, and standard deviations (SD) for the nearshore study area for Amphipoda, Bivalvia, Cumacea, Isopoda, Actinopterygii, Polychaeta, and the total prey (T6: sum of 6 prey) biomass, 2001e2015.   Table 3 Average biomass (g m À2 ) of six prey groups and total prey biomass from feeding points in the nearshore and offshore surveys 2002e2015. depths ranging from 7 to 35 m nearshore and from 30 to 63 m offshore. Both survey grids in the Sakhalin Island coastal study area are adjacent to or overlap with areas of heightened anthropogenic activities including commercial fishing and oil and gas platforms (Fig. 1). During the course of the investigation, locations where gray whales were observed feeding were opportunistically sampled and three replicates collected at each point. These feeding points were identified by gray whale observers from shore and on vessels associated with oil and gas exploration and production activities. The feeding points provide further information characterizing specific locations where whales feed. Feeding areas were sampled differently in 2015 using a targeted sampling approach with six replicates collected along two transects at 9 m and 13 m for a total of twelve replicates at each location. Feeding areas from 2015 were not statistically-evaluated for differences here but were considered separately (Blanchard et al., unpublished data). Environmental variables included water depth, year of sampling, percent sand (sand: particles between 0.1 mm and 1.0 mm; other categories were colinear with sand), and the Aleutian Low Pressure Index (ALPI). The Aleutian Low [2] influences winter wind patterns and sea-level pressure throughout the Bering Sea and variations in its strength and position can directly influence water circulation [3,4]. The ALPI is available at https://open.canada.ca/data/.

Statistical analyses
Analysis of covariance was performed for surveys using mixed modeling to test for differences among years. ANCOVA's were performed separately for the nearshore (incorporating data from 2001 to 2015) and offshore surveys (using data from 2002 to 2015). The mixed-modeling package nlme [5] was used with the statistical program R [6] for analysis of as it allows incorporation of models for temporallycorrelated errors. Autoregressive and moving average correlation models were used in nlme to correct Table 4 Tukey multiple comparisons among years for faunal groups for nearshore surveys, 2001e2015.

Comparison
Amphipoda       Est. ¼ the difference between feeding point biomass e grid station biomass for transformed biomass data used in the mixed models, and P-values from mixed models. A positive estimate value indicates that average biomass was higher at feeding points.
The "Comparison" columns denote whether biomass was higher in feeding points (FP) or grid stations (GS). Years included in the ANOVA were 2002e2012 for the nearshore and 2002e2015 for the offshore. for temporal correlations in yearly averages. Correcting for temporal correlations among errors increases the precision of statistical tests by correcting variances. Here, we limited our consideration to models of at most 3 lags, or up to 3 years distant. We also presumed that any spatial correlation structures would be approximated by and incorporated in the correlation models. Models considered for adjusting errors were autoregressive (AR), moving average (MA), and combined models (ARMA). Model selection included determination of the variables appropriate for inclusion as well as the best correlation model. The available models included depth, year, and Depth and Year. Depth was a continuous variable and Year a fixed factor. Station was included as a random factor in mixed models. Akaike's Information Criterion (AIC) was used to determine the best model of the three for each faunal group analyzed. The choice for which correlation model to use was guided by likelihood ratio tests that compare the variance reductions among correlation models. Tukey multiple comparisons were performed using the lmerTestpackage in R [7].

Multivariate analyses
Multivariate analyses were applied to characterize changes in benthic community biomass concentrations related to environmental predictors. Canonical correspondence analysis (CCA) was used to test the hypothesis that environmental and temporal covariates were predictors of biomass  Mixed models were adjusted for time-series errors. F-statistics (F) and p-values are presented. community structure. The community data set was benthic biomass of all categories identified with rare animals excluded. Biomass data were ln(Xþ1)-transformed prior to analyses to reduce influences of extreme values on the ordination. The covariates were water depth, year, the ALPI (a measure of macro-scale climate variability), and percent sand. CCA was conducted using the vegan package [8].

Acknowledgments
We are grateful for the assistance of the captains and crews of the R/V Aademic OparinR/V Okean, R/ V Lavrentyev, R/V Igor Maksimov, and R/V Nevelskoyeand scientists at the National Scientific Center of Marine Biology (NSCMB, previously the AV Zhirmunsky Institute of Marine Biology) of the Far Eastern Branch of the Russian Academy of Sciences. Exxon Neftegas Limited. (ENL) and Sakhalin Energy Investment Company, Ltd. (SEIC) provided financial support for sampling and laboratory analyses through the joint western gray whale research and monitoring program, Sakhalin Island, Russia, 2001e2015. Funding for this publication was provided by ENL. Many people contributed to the successful implementation of the benthic surveys throughout the years, and we would like to thank the benthic field sampling teams and laboratory personnel of NSCMB, the science lead on the vessel Yuri Yakovlev (NSCMB), and the logistic support of Vladimir Efremov, Ervin Kalinin (ENL) and Igor Zhmaev (LGL ECO). We dedicate this paper to V. I. Fadeev (NSCMB), who was the lead benthic investigator during most (13) years and H. R. Melton who was instrumental in establishing and continuation of the western gray whale research program. . The conclusions of the paper are those of the authors and do not necessarily represent the views of Exxon Neftegaslink Limited. (ENL) and Sakhalin Energy Investment Company, Ltd. (SEIC) provided financial support for sampling and laboratory analyses through the joint western gray whale research and monitoring program, Sakhalin Island, Russia, 2001e2015. Funding for this publication was provided by ENL..

Transparency document
Transparency document associated with this article can be found in the online version at https:// doi.org/10.1016/j.dib.2019.103968.