An 8-year record of phytoplankton productivity and nutrient distributions from surface waters of Saanich Inlet

Phytoplankton are the base of nearly all marine food webs and mediate the interactions of biotic and abiotic components in marine systems. Understanding the spatial and temporal changes in phytoplankton growth requires comprehensive biological, physical, and chemical information. Long-term datasets are an invaluable tool to study these changes, but they are rare and often include only a small set of measurements. Here, we present biological, physical and chemical oceanographic data measured periodically between March 2010 and November 2017 from the euphotic zone of Saanich Inlet, a temperate fjord on the west coast of British Columbia, Canada. The dataset includes measurements of dissolved macronutrients, total and size-fractionated chlorophyll-a, particulate carbon, nitrogen and biogenic silica, and carbon and nitrate uptake rates. This collection describes phytoplankton dynamics and the distribution of biologically-available macronutrients over time in the upper water column of Saanich Inlet. We establish a baseline for future investigations in Saanich Inlet and provide a data collection protocol that can be applied to similar productive coastal regions. Measurement(s) chlorophyll a • biogenic silica • carbon uptake • nitrogen uptake • particulate carbon • particulate nitrogen • inorganic macronutrient dissolved in ocean water • concentration of silicic acid in ocean water Technology Type(s) fluorometer • ultraviolet-visible spectrophotometry • elemental analysis and isotope ratio mass spectrometry • nutrient autoanalyzer Sample Characteristic - Organism Bacillariophyta • Dinophyceae • Chlorophyceae Sample Characteristic - Environment estuary • coastal inlet Sample Characteristic - Location Vancouver Island • North East Pacific Ocean coastal waters of Southern British Columbia, Washington and Oregon • Province of British Columbia Measurement(s) chlorophyll a • biogenic silica • carbon uptake • nitrogen uptake • particulate carbon • particulate nitrogen • inorganic macronutrient dissolved in ocean water • concentration of silicic acid in ocean water Technology Type(s) fluorometer • ultraviolet-visible spectrophotometry • elemental analysis and isotope ratio mass spectrometry • nutrient autoanalyzer Sample Characteristic - Organism Bacillariophyta • Dinophyceae • Chlorophyceae Sample Characteristic - Environment estuary • coastal inlet Sample Characteristic - Location Vancouver Island • North East Pacific Ocean coastal waters of Southern British Columbia, Washington and Oregon • Province of British Columbia

One of the dominant groups of pelagic primary producers in Saanich Inlet are diatoms 12 , which tend to proliferate in productive, temperate ecosystems, such as major upwelling regions and coastal fjords [13][14][15][16] . As a relatively large and nutritionally high-quality phytoplankton, diatoms are important for supporting marine food webs with high animal biomass in Saanich Inlet and elsewhere. These animals include several economically important species such as Pacific herring (Clupea pallasii) and Pacific salmon (Oncorhynchus sp.). Due to their obligate silicon (Si) requirement to build their silica (SiO 2 ) frustules, diatoms need to take up dissolved Si in the form of silicic acid (Si(OH) 4 ). Measurements of suspended particulate biogenic silica (bSiO 2 ) in seawater provide estimates of siliceous phytoplankton biomass (primarily diatoms) and their contribution to total phytoplankton biomass. By comparing changes in diatom biomass to oceanographic conditions and nutrient availability, we can better understand how processes such as eutrophication and climate change may affect marine ecosystems in the inlet.
Fluctuating deep-water hypoxia, high pelagic primary productivity, restrictive bathymetry, and exposure to human activities result in a seasonally variable biogeochemical environment in Saanich Inlet 2,5,17,18 . This fjord provides a useful natural laboratory to investigate the relationships between primary productivity, concentrations and distributions of macronutrients, and abiotic oceanographic factors. The 2010-2017 data presented in this paper can be particularly useful when combined with other publicly available Saanich Inlet datasets such as those generated by the Ocean Networks Canada cabled observatory 19 , and by other groups 20,21 . Previous measurements (2005)(2006) conducted in Saanich Inlet are not included here, but are available in an earlier manuscript from our research group 4 . The data from 2016 to 2017 were collected as part of the Saanich Inlet Redox Experiment (SaanDox) lead by Dr. Roberta Hamme at the University of Victoria. All samples and data were obtained and analyzed by members of Dr. Diana Varela's laboratory at the University of Victoria over almost eight years using consistent methodologies. This long time-series of observations can be extrapolated to understand how primary producers are involved in larger oceanic processes such as the formation of OMZs, eutrophication, carbon export, and fisheries production in other marine systems.

Methods
Sample collection and hydrography. Sampling was conducted aboard the University of Victoria's MSV John Strickland either weekly, biweekly or monthly between 11 March 2010 and 15 November 2017 in Saanich Inlet at 48.59°N, 123.50°W (Fig. 1). To standardize measurements and due to biological significance, seawater was collected from the euphotic zone. Sampling depths corresponded to approximately 100, 50, 15, and 1% of the ). CTD measurements include depth, temperature and conductivity, with the addition of photosynthetically active radiation (PAR), fluorescence and dissolved oxygen when available. Data presented for the first time in this publication are represented by black squares, while previously published data 20,37,38 included in this data record are shown with grey squares. The year labels are positioned under the tick marks corresponding to January. Table 1 lists all measurements available in the associated data file and figures.
www.nature.com/scientificdata www.nature.com/scientificdata/ photosynthetically active radiation (PAR) at the surface (I o ). These "light" depths were either determined using a CTD-mounted PAR sensor or a Secchi disk. CTD profiles were performed prior to each seawater cast to measure depth, temperature and conductivity of the water column, and PAR, fluorescence, and dissolved oxygen (when available).
Seawater from each light depth was collected using Niskin or GO-FLO bottles on either a rosette sampler or an oceanographic wire. When possible, individual samples were collected directly from the Niskin or GO-FLO bottles. When time was not sufficient to allow direct sampling, bulk samples of seawater from each depth were collected into acid-washed polyethylene carboys, kept cold in the dark, and homogenized before sub-sampling for the individual measurements.
Dissolved nutrients. For the measurements of nitrite (NO 2 − ), nitrate and nitrite (NO 3 − + NO 2 − ), phosphate (PO 4 3− ) and Si(OH) 4 , seawater samples from each light depth were syringe-filtered through a combusted 0.7 µm (nominal porosity) glass fibre filter into acid-washed 30-mL polypropylene bottles and immediately frozen. All nutrient samples were stored at −20 °C until analysis. Concentrations of NO 2 − , NO 3 − + NO 2 − , PO 4 3− , and Si(OH) 4 were determined using an Astoria Nutrient Autoanalyzer (Astoria-Pacific, OR, USA) following the methodology of Barwell-Clarke and Whitney 22 . During 2014 and 2015, samples for the measurement of Si(OH) 4 were collected separately from those for the other nutrients, filtered with a 0.6 µm polycarbonate membrane filter and stored at 4 °C. During this period, Si(OH) 4 concentrations were determined manually using the molybdate blue colorimetric methodology 23 . Replicate (2 or 3) nutrient samples were taken at each depth; average data are presented in the published dataset and the figures (Fig. 2).

Suspended particulate matter. Total chlorophyll-a.
Chlorophyll-a (Chl-a) was used as a proxy for phytoplankton biomass (Fig. 3A). For total Chl-a analysis, seawater samples (0.25-1 L) were gently vacuum filtered onto 0.7 µm (nominal porosity) glass fiber filters, which were then stored at −20 °C until analysis. Chl-a Chl-a (> 20 µm) Contribution to total chlorophyll-a by the >20 µm size fraction % contribution mg m −2 Chl-a (5-20 µm) Contribution to total chlorophyll-a by the 5-20 µm size fraction % contribution mg m −2 Chl-a (2-5 µm) Contribution to total chlorophyll-a by the 2-5 µm size fraction % contribution mg m −2 Chl-a (0.7-2 µm) Contribution to total chlorophyll-a by the 0.7-2 µm size fraction % contribution mg m −2 Chl-a (0.7-5 µm) Contribution to total chlorophyll-a by the 0.7-5 µm size fraction % contribution mg m −2 Total Chl-a Total concentration of chlorophyll a (phytoplankton larger than 0.7 µm in diameter) Rate of carbon uptake derived from using 13 C tracer (ρC) µg C L −1 day −1 and µmol C L −1 day −1 mmol C m −2 day −1 rho_NO 3 Rate of nitrate uptake derived from using 15  www.nature.com/scientificdata www.nature.com/scientificdata/ concentrations were determined using the acetone extraction and acidification method 24,25 . Acidification of samples decreased the likelihood of overestimation of Chl-a concentrations due to the presence of chlorophyll degradation products 26 . Filters were submerged in 10 mL of 90% acetone, sonicated for 10 minutes in an ice bath, and left to extract at −20 °C for 22 h. Following the extraction period, samples were allowed to equilibrate to room temperature (~2 h). Fluorescence of the acetone solution containing the extracted Chl-a was measured before and after acidification with 1.2 N hydrochloric acid using a Turner 10-AU fluorometer. The final concentrations of total Chl-a were calculated from measurements made before (Fo) and after (Fa) acidification using Eq. (1) 25 . The coefficient (τ) of Eq. (1), adapted from Strickland and Parsons 25 , was derived from a calibration of the Turner 10-AU fluorometer with known pure chlorophyll standards ( Table 2).
Particulate carbon and nitrogen. Particulate C and N measurements were obtained from seawater samples incubated for carbon (ρC) and nitrate uptake (ρNO 3 ) rates (see section on "Uptake rates of carbon and nitrate" for methodology) (Fig. 3B,C). PC and PN measurements presented in this dataset were taken at the end of ρC and ρNO 3 incubations; however, original ('ambient') values can be back calculated by subtracting the amount of C and N taken up during the incubation period from the final PC and PN values. The differences between after-incubation PC and PN data and back-calculated ambient values were not significantly different than the measurement error.   Table 2. Fluorometer (Turner Designs ® , Model 10-AU) calibration dates and value of coefficient (τ) used in Eq.
(1) for the calculation of chlorophyll-a during this study.
www.nature.com/scientificdata www.nature.com/scientificdata/ analysis was performed on the supernatant. The transmittance of the samples, standards, and reverse-order reagent blanks were read at 820 nm using a Beckman DU 530 ultraviolet-visible (UV/Vis) spectrophotometer 27,28 . Uptake rates of carbon and nitrate. Seawater samples (~0.5-1 L) were gently collected into clear polycarbonate bottles. One additional sample was collected from the 100% light depth, into a dark polycarbonate bottle, which did not allow light penetration. After the addition of the isotopic tracers (see below), bottles were placed into an acrylic incubator with constant seawater flow to maintain surface seawater temperature. Three acrylic tubes wrapped in colored and neutral density photo-film (to obtain 50, 15, and 1% of surface PAR) were used to incubate sample bottles under the same in-situ light conditions from which samples were collected. Samples from the 100% light level were placed inside the same acrylic incubator, but outside of the film-covered tubes.
A LI-COR ® LI-190 Quantum sensor was installed next to the incubator and continuously recorded incoming PAR for the entire incubation period. During sampling in 2010 and 2011, all experiments were performed using a shipboard incubator. For sampling from 2012 onwards, all experiments were done using an incubator on land (University of Victoria Aquatic Facility), which was connected to a seawater system maintained at local surface seawater temperature (approximately 9-12 °C depending on the time of year).
Rates of C (ρC) and NO 3 (ρNO 3 ) uptake were determined using a stable isotope tracer-technique 29,30 (Fig. 4). A single seawater sample from each light depth received a dual spike, with NaH 13 CO 3 (99% 13 C purity, Cambridge Isotope Laboratories) for the determination of ρC and Na 15 NO 3 (98 + % 15 N purity, Cambridge Isotopes Laboratories) for the determination of ρNO 3 . Isotope additions were made at approximately 10% of ambient dissolved inorganic carbon (DIC) and NO 3 − concentrations. Spiked seawater samples were incubated for 24 h, except from 2010 to 2013 when the incubation period was 4 to 6 hr. After incubation, the entire sample was gently vacuum filtered onto a combusted 0.7 µm (nominal porosity) glass fibre filter. Filters were dried for 48 h at 60 °C and kept in a desiccator at room temperature until analysis. Filters were packed into pellets and sent to the Stable Isotope Facility at the University of California (UC) Davis for analysis of 13 C and 15 N enrichment, and total C and N content by continuous flow isotope ratio mass spectrometry and elemental analysis, respectively. For these measurements, UC Davis uses either an Elementar Vario EL Cube or Micro Cube elemental analyzer (Elementar Analysensysteme GmbH, Hanau, Germany) interfaced to either a PDZ Europa 20-20 isotope ratio mass spectrometer (Sercon Ltd., Cheshire, UK) or an Isoprime VisION IRMS (Elementar UK Ltd, Cheadle, UK). For samples incubated for less than 24 h, the daily C or NO 3 − uptake rates (ρX) were calculated using a PAR extrapolation method shown in Eq. (2): Additionally, to account for NO 3 − uptake occurring under no light, ρNO 3 − was measured in dark bottles and this rate was added to the ρNO 3 − of each sample incubated for less than 24 h. The ρNO 3 − DARK was calculated following Eq. (3): PAR data used in Eqs. (2) and (3) came from the LI-COR ® LI-190 Quantum sensor that was mounted beside the incubator. The seawater DIC value for each sample was calculated using a regression equation relating water density to DIC for Saanich Inlet 31 . Ambient NO 3 − concentrations were measured as described above.

Data records
The Saanich Inlet biological dataset is accessible as a tab delimited data file "Saanich_BioOcean_Data.tab" (https://doi.org/10.5683/SP2/6BATWK) on Dataverse 32 and contains the data fields outlined in Table 1. Data is presented for discrete depths of all measurements and as depth integrated values for dissolved nutrient, biomass and nutrient uptake rates.

Technical Validation
Sample collection and analyses. All seawater samples were collected by qualified individuals aboard the University of Victoria's MSV John Strickland. Personnel were trained either by experienced members of Dr. Varela's lab or by Dr. Varela herself at the University of Victoria before participating in research cruises, conducting sample collection or analysing samples using well-established protocols. Seawater was always collected into clean, acidwashed bottles to prevent contamination, except for Chl-a samples for which clean, non-acid washed bottles were used. Samples were kept cold and in the dark immediately after collection, and were processed as soon as possible. Calibration of fluorometer for chlorophyll-a measurements. The same Turner 10-AU fluorometer was used to measure all Chl-a samples over the course of the study. The fluorometer was calibrated three times during the study period using standards of pure Chl-a extracted from the algae Anacystis nidulans and the working coefficient (τ) was updated after each calibration ( Table 2). Calibration of the fluorometer was frequently checked between calibration dates by measuring solid Chl-a standards provided by Turner Designs ® .

Calibration of CTDs. Two
Precision of particulate biogenic silica data and sample digestion. At each sampling time triplicate (n = 3) samples of bSiO 2 were collected at 100% I o and were used to calculate a coefficient of variation (CV). The CV of 12.9% ± 7.9 is representative of the bSiO 2 values presented in this dataset. Single measurements were conducted at depths below 100%.
The NaOH digestion method used for the determination of bSiO 2 in this study may result in the digestion of 10-15% of lithogenic SiO 2 from the samples 34 . Digestion times were chosen as to limit the likelihood of lithogenic SiO 2 leaching while at the same time, completely digesting bSiO 2 .

Carbon and nitrogen techniques: isotopic ratios and particulates. Vienna Pee Dee Belemnites and
Air were used as the international reference standards for 13 C and 15 N measurements, respectively. Details on the www.nature.com/scientificdata www.nature.com/scientificdata/ specific protocols for instrument calibration at UC Davis can be found here: https://stableisotopefacility.ucdavis. edu/carbon-and-nitrogen-solids.
The material collected on the filters at the end of the incubations was not acidified before measuring the stable isotope ratios and PC/PN concentrations at UC Davis. Therefore, the total PC after the incubation could potentially include inorganic CaCO 3 material that can overestimate the total organic C content of the sample. Similarly, 13 C uptake rates could also be overestimated by the activity of coccolithophores. Although we did not conduct direct measurements of calcifying material in the water column or compare rates before and after acidification, it is known that Saanich Inlet is a diatom dominated system and that there is little to no contribution from calcifying phytoplankton 12,35 . Incubation methodology. To minimize the loss of particulate matter due to adsorption on bottle walls during the incubation, samples were vigorously rinsed three times with filtered seawater and the rinse water was also filtered through the same filter at the end of the incubation period 36 . Additionally, "blank" and "dark" rates were measured with each set of incubations. Blank samples were enriched with isotopes but were immediately filtered onto combusted 0.7 µm glass fibre filters. Dark samples were enriched with isotopes and incubated in completely black bottles. Data from the blank samples provided the natural isotopic values of 13 C and 15 N used in the calculation of uptake rates. Dark bottle 13 C values were used as a validation tool in order to confirm that there was no measured 13 C uptake under no-light conditions. Dark bottle 15 NO 3 values were used to correct for daily rates of 15 NO 3 − uptake for <24 h incubations (see above). Incubations were maintained at approximately surface seawater temperature for the length of the incubation periods. The average temperature difference between the surface and the deepest depth sampled was <2 °C.

Usage Notes
Data are presented in a single tab delimited excel file. Discrete depth concentrations are presented in the excel file followed by integrated total euphotic zone values. Values below the limit of detection are presented as 0 in this dataset but may be interpreted as any number between 0 and the limit of detection defined in this manuscript for each parameter measured.
Partial CTD, nutrient and biomass data from 2010, 2011, and 2015 have been included, as averages, in two previously published studies 37,38 . Here, we include the complete data for each depth for the 2010-2011, and 2015 sampling period. Additionally, some CTD data collected between 2012-2015 in collaboration with Dr. Hallam's lab at UBC, have been previously published 20 but are included here for completeness of the dataset. CTD data from 2016 and 2017 presented here were collected as part of the Saandox project and have not been previously published.

Code availability
The majority of data processing was done using Microsoft Excel 2010 ® version 14.0.4734.100. Python v3.8 was used for the calculations of the percentage size fractions of Chl-a, seasonal averaging (i.e. binning values into a monthly average for each sampled depth) and scaling the figure colormaps. The specific code written for this manuscript can be found within the plotting script at the following open source GitHub repository: https:// github.com/bjmcnabb/Saanich_Inlet.