Dataset: Mesocosm warming experiment seagrass metrics

Mesocosm experiment. 
We used a substitutive design to test the effects of eelgrass genotypic identity (eight genotypes), diversity (monocultures of 1 genotype vs. polycultures of 4 genotypes), and temperature (ambient or + 3.2° C) on the prevalence and intensity of the wasting disease parasite Labyrinthula over eight weeks in an array of flow-through 120-Liter mesocosms at the Bodega Marine Laboratory in Bodega Bay, CA. In July 2015, we created ten unique polyculture combinations of four genotypes (4 genotypes per experimental pot) randomly drawn from a pool of eight genotypes; all eight genotypes were also grown in monoculture (1 genotype per pot). We filled pots (8.9 x 8.9 cm) with coarsely sieved sediment collected from Bodega Harbor, and planted 4 shoots of eelgrass per pot, matching the lower range of average field densities reported for Bodega Harbor (Ha and Williams, 2018) to allow for growth during the experiment. Plants were originally collected in Bodega Harbor, CA in 2012, confirmed to be unique genotypes using 11 DNA microsatellite loci developed specifically for Zostera marina (Abbott et al. 2018), and propagated in separate flow through mesocosms at Bodega Marine Lab. We previously characterized traits of each genotype relating to growth rate, morphology, nutrient content, and chemical defense in common garden experiments at ambient temperature from July 2013 to August 2014 (Abbott et al., 2018). We selected the 8 genotypes used in this experiment to encompass the range of trait values determined for this population of eelgrass measured when the common garden was experiencing marine heatwave conditions (DuBois et al., 2019).

We assigned ten pots -- two unique polyculture combinations and each of the eight monocultures -- to each of ten mesocosms, with five mesocosms per temperature treatment (see DuBois et al. 2020 for a diagram of the experimental set up). All mesocosms received sand-filtered flow-through seawater at a rate of approximately 0.8-1.0 liters per minute (L min^-1). We allowed the plants to acclimate for one month prior to initiating the temperature treatments. We maintained an ambient temperature treatment by cooling flow-through seawater in a head tank by approximately 1˚C using an Aqua Logic Delta Star in-line titanium chiller. Seawater in the elevated temperature treatment was raised approximately 3˚C above the ambient treatment in a separate header tank using Process Technologies titanium immersion heaters. This level of warming mimicked the 2014 and 2015 extreme warming events in the Northern Pacific, where the mass of unusually warm water called “The Blob” raised summer ocean temperatures three standard deviations above the long-term average (Sanford et al., 2019).

At one-month intervals over the course of the experiment, we measured leaf growth rate of the terminal shoot of each transplant using the “hole-punch” method (Williams and Ruckelshaus, 1993). Labyrinthula zosterae infection can be affected by plant defenses (Steele et al., 2005; Trevathan-Tackett et al., 2015) and these defenses may trade off with plant growth rate, resulting in a positive relationship between growth rate and infection. Alternatively, L. zosterae infection can result in reduced leaf growth rate (Graham et al., 2021), leading to a negative relationship. L. zosterae prevalence can also be affected by plant size (Groner et al., 2016), due to greater leaf-to-leaf contact and resulting increased parasite transmission (Muehlstein, 1992), so we measured the length of the focal leaf of each transplant at the end of the experiment.

Labyrinthula zosterae DNA extraction and quantitative PCR Assay
We extracted L. zosterae DNA from dried leaf tissue using Omega Bio-Tek E.Z. Tissue DNA extraction kits at the Northeastern University Marine Science Center in Nahant, MA. For each sample, we separated the dried leaf tissue into 2-16 mg subsamples and homogenized the tissue in a ball mill (Retsch, Germany) at a frequency of 30 Hz for 5 min (Bockelmann et al. 2013). We lysed ground subsamples individually following the manufacturer’s instructions and added 1 microliter of 500 ng per microliter (ng*uL^-1) salmon sperm DNA solution (Invitrogen, USA) to the first subsample of each sample immediately before recombining all subsamples in the spin columns. Salmon sperm DNA was added to enhance extraction efficiency and ensure that even low amounts of target DNA are carried through the filter absorption steps (Bockelmann et al., 2013). We eluted all DNA extractions into 100 uL. Following elution, we used Zymo OneStep-96 PCR Inhibitor Removal kits to clean 50 uL sub-samples of each DNA extraction following the manufacturers instructions. We stored cleaned DNA extractions at -20˚C prior to quantitative PCR.

We used a TaqMan quantitative PCR (qPCR) assay with a forward primer: TTGAACGTAACATT-CGACTTTCGT, reverse primer: ACGCATGAAGCGGTCTTCTT, and MBG probe: TGGACGAGTGTGTTTTG that carries the fluorescence label 6-Fam at the 5’ end and the dark quencher FHQ at the 3’ end (Bio-Rad, USA) developed specifically for L. zosterae (Bockelmann et al., 2013, Bergmann et al., 2011). We made up qPCR reactions to a 10 uL reaction volume using standard conditions recommended by the manufacturer: 5 uL SsoAdvanced^TM Universal Probes Supermix 2x (Bio-Rad, USA), 1 uL template DNA, 0.4 uL 4:1 Primer:Probe Mix (final concentrations of 400 nM forward primer, 400 nM reverse primer, 100 nM probe), and 3.6 uL Milli-Q H₂O (Thermofisher, USA). Reactions were run on a CFX96 Real-Time System (Bio-Rad, USA) using the following thermo-cycling program: 3 min at 95˚C, followed by 40 cycles of 15 sec at 95˚C and 1 min at 60˚C. We tested all samples in duplicate and if replicates differed by greater than one cycle threshold (Ct), reactions were rerun in triplicate. We only used the data from reactions in analyses when replicates fell within one Ct. Our lowest detection was 1.76 copies per reaction or 0.15 cells per extraction.

We ran each 96-well plate of qPCR reactions with a set of nine standards: a dilution series of gBlock Gene Fragments (Integrated DNA Technologies, USA) designed based on the highly conserved sequence of the 5.8s ribosomal RNA gene of L. zosterae known as internal transcribed spacer 1 (ITS) targeted by the TaqMan qPCR assay; an L. zosterae cell standard consisting of a sample of DNA extracted from a know quantity of pathogenic L. zosterae cells; and an inhibition control consisting of a half volume of L. zosterae cell standard and a half volume of a haphazardly selected sample. We ran a total of 31 96-well plates of qPCR reactions with a mean efficiency of 97.4% ± 4.3 and R² 0.996 ± 0.004. To convert Ct values to to L. zosterae cell numbers, four equations were used. The equations and details are captured in the Supplemental Files section of this metadata within the file titled, "Equations for conversion of Ct values to Labyrinthula zosterae cell numbers."

We used a pure culture of the pathogenic L. zosterae isolate 316b provided by D. Martin in 2015 to make our L. zosterae cell standard (Martin et al., 2016; GenBank: KU559372.1). We cultured L. zosterae cells on serum seawater agar media (Muehlstein et al., 1991). We scraped cells from an actively growing edge of L. zosterae culture into serum seawater liquid media (D. Martin pers. com.). We mixed the liquid media + L. zosterae cell slurry vigorously on a bench top vortex for 30 sec and aliquoted immediately into three replicate subsamples for cell counts and extraction. In order to break up cell clumps for ease of counting, we added Tween80 (Sigma-Aldrich, USA) to a final concentration of 1:100 into the two subsamples used for cell counts, and mixed for 30 sec. We counted cells of four replicate aliquots per subsample on a hemocytometer. We calculated cell concentration by averaging over all replicates. Prior to DNA extraction, we centrifuged the third replicate L. zosterae cell solution at 6,000 g for 10 min and drew off the supernatant without disturbing the cell pellet. We then added a ~4 mg section of dried healthy Zostera. marina tissue to the cell pellet to account for possible interference of Zostera marina compounds in the extraction process. To extract L. zosterae DNA, we followed the DNA extraction and inhibitor removal protocols outlined above.

We designed the gBlock double stranded DNA fragments (Integrated DNA Technologies, USA) using published sequences of the ITS region of the L. zosterae genome (GenBank: JN121409-13).

5’-CTGTGATCTCTGAAAATACTTGTTT
(1)TTGAACGTAACATTCGACTTTCGTCGATT TTG
(2)TGGACGAGTGTGTTTTGT AAACCTACCC
(3)AAGAAGACCGCTTCATGCGT GTCGCTGACTAATGAAACAAACAAA-3’

The gBlock fragment sequences were a total length of 130 bp (base pairs), which included target regions for the forward (1) and reverse (3) primers and the MGD probe (2), underlined above, as well as 25 base pairs of additional sequence on both the 5’ and 3’ ends to increase fragment stability. We diluted gBlock fragments in Milli-Q H₂O (Thermofisher, USA) to seven concentrations: 2.24e¹, 1.12e², 5.61e², 2.81e³, 1.40e⁴, 7.02e⁴, 7.02e⁵ copies/µL and included this dilution series in each qPCR run as a standard curve (Bergmann et al., 2011). The range of the gBlock dilution curve: approx. 1-60,000 cells/extraction encompassed the range of most L. zosterae values observed in our samples: 0.15-450,000 cells/extraction or 1.84e²-5.52e⁸ copies/extraction.

Life Sciences Identifiers (LSID) for taxonomic names:
Zostera marina (urn:lsid:marinespecies.org:taxname:145795)
Labyrinthula zosterae (urn:lsid:marinespecies.org:taxname:395093)
Labyrinthula (urn:lsid:marinespecies.org:taxname:119090)