The purpose of this experiment, performed as part of the McMurdo Dry Valleys Long Term Ecological Research (MCM LTER) program, was to investigate the impact of lake level rise and moat expansion on microbial community diversity and function in the East Lobe of Lake Bonney, located in Taylor Valley, Antarctica. The “tLICE” experiment tested the following MCM5 Hypotheses: H3-Disturbance increases connectivity and accelerates shifts towards homogeneity, and H4-Decreased heterogeneity reduces community resistance and resilience. Lake water from the East Lobe of Lake Bonney was collected at depths representing maximum phytoplankton productivity (5 m, 17 m, 23 m), and transferred to dialysis bags. The bags were then attached to PVC frames and transplanted to new locations within the lake in two experiments: 1) a moat transplant, in which 5 m communities were transplanted to the open water moat; and 2) shallow to deep transplant, in which communities from all three sampling depths were moved 3 m down the water column to mimic approximately one decade of lake level rise. Transplanted samples were incubated under these conditions for two weeks, and then frames were recovered and samples were processed for a number of measurements. This data package includes the following measurements: nutrient concentration, chlorophyll-a content, chlorophyll fluorescence, cell density, amplicon gene sequences (16S and 18S rRNA genes), and physical parameters.
Lake water samples were collected at specific depths with a five-liter Niskin bottle. One 100 mL sub-sample was filtered through a combusted (475 degree centigrades for 4 hours) and acidified (the acidification step began with the 2008-2009 season) Whatman 25 mm GF/F using a bell jar apparatus. The filtrate was collected in a 125 mL acid washed Nalgene bottle and stored at -20 degrees centigrades until analysis at the Crary Analytical Lab. Nutrient assays are performed with a Lachat Quickchem AE Autoanalyzer using standard methods. Ammonia was analyzed using the phenol hypochlorite method. Nitrite was determined by diazotization with sulfanilamide and NED. Nitrate was determined by reducing nitrate to nitrite by passing a sample through a copperized cadmium column and analyzing for nitrite, which yields nitrate + nitrite (N+N) concentration. The nitrate concentration was determined by subtracting the nitrite concentration from the N+N concentration. Soluble Reactive Phosphorus was determined using ammonium molybdate methods.
Lake water samples were collected at specific depths with a five-liter Niskin bottle. Two-100 mL were filtered through a combusted (475 C for 4 hours) Whatman 25 mm GF/F using a bell jar apparatus. The filter was folded in half (organic material inside), placed in a glassine envelope, covered with aluminum foil, and frozen immediately for later analysis at McMurdo Station. All of the following laboratory methods were performed in a darkened environment. At McMurdo Station, a chlorophyll-a stock concentrate was prepared by dissolving 1 mg of chlorophyll-a standard (Sigma, Anacystic nidulans) in a 100 ml volumetric flask and diluting it to mark with 90% acetone (~10,000 micrograms per liter chlorophyll-a). A Beckman DU-640 UV/vis spectrophotometer was used to determine the actual chlorophyll-a concentration of the stock concentrate and the concentration of the stock dilutions. The absorbance of the stock concentrate was measured at 665 nm and 750 nm (non-acidified readings are denoted by a subscript "o"). The stock concentrate was acidified in the cuvette using 2-4 drops of 3 N HCl. The absorbance after acidification was measured at 665 nm and 750 nm (acidified readings are denoted by a subscript "a"). Chlorophyll-a content was determined using the following equation (Strickland and Parsons 1972; Parsons et al. 1984): Chl-a (micrograms per liter)= [26.7*((ABS665o - ABS665a) - (ABS750o - ABS 750a))*1000]/l where:
The stock solution was used to prepare six to ten standard dilutions of chlorophyll-a ranging from ~1.5 micrograms per liter to 500 micrograms per liter, plus a blank of 90% acetone. 90% acetone was used to dilute the standards. Fo and Fa were obtained for each standard by collecting initial fluorescence data on the Turner 10-AU fluorometer, and then acidifying the standard in the cuvette with 4 drops of 3 N HCl. The cuvette was briefly vortexed before determining Fa. The actual concentrations of the working standards were computed from the spectrophotometrically determined concentration of the stocks. A standard curve of chlorophyll-a concentration vs Fo-Fa was prepared. Each filter was placed into a 20 ml scintillation vial and extracted with 10 ml of 90% acetone. The extract was incubated for ~12 hours under cold (<0C), dark conditions. After incubation, the extract was briefly vortexed and 4 ml dispensed into the cuvette; the cuvette was inserted into the Turner 10-AU fluorometer. After Fo was determined, the sample was acidified with 4 drops of 3 N HCl, vortexed, and Fa was determined. The cuvette was rinsed three times with DI water and three times with 90% acetone to ensure there was no cross contamination of sample or acid between samples. The cuvette exterior was wiped with Kimwipes. The Fo-Fa was determined for each sample and chlorophyll-a concentration (microgram per liter) was calculated by comparison with the standard curve as below: Chlorophyll-a (micrograms per liter) = (((Fo-Fa) - y-intercept)/slope) * (ml extracted/ml filtered), where:
Samples were collected from various depths with a 5-L niskin bottle or dialysis bags. Chlorophyll fluorescence measurements were also conducted using the PHYTO-PAM II phytoplankton analyzer (Heinz Walz, Effeltrich, Germany). Prior to deployment in the field, algal references were constructed in the laboratory using pure cultures of Lake Bonney algae, Chlamydomonas sp. ICE-MDV (chlorophytes), Isochrysis sp. MDV (mixed/brown group), and an Antarctic cryptophyte, Gemigera cryophila. These references were then used to deconvolute the algal classes and estimate Chlorophyll-a/L. In addition, the following PSII photochemical parameters were measured for each algal class: i) FV/FM – maximum PSII photochemical efficiency, ii) Y(PSII) – effective quantum yield of PSII, iii) Y(NPQ) – non photochemical quenching, iv) Y(NO) – non-regulated nonphotochemical quenching.
Lake water was collected with a 5-L niskin bottle. 4.5 mL of lake water was transferred to a 5 mL cryovial and 500 μL of TE glycerol buffer (5% glycerol + 1XTE buffer) was added. Samples were flash frozen in liquid nitrogen and shipped back to the US on dry ice. Flow cytometry was performed on a BD Acurri flow cytometer. Samples were passed through a 40 μm filter, various stains were applied, and measurements were taken for a minimum of 30,000 events. Analysis of the events consisted of first gating for singlet cells by graphing FSC-H vs FSC-A. SYBR® Green I was used to count bacteria by adding 10 μl of working stock (1:100 dilution in DMSO) to 500 μl of sample in black tubes, inverting 10 times, and incubating on ice for 10 minutes before measuring on FL1. Chlorophyll autofluorescence (~430/670) was monitored on unstained samples on channel FL3 (488/>670).
Amplicon gene sequences
Samples collected from the water column (500 mL) or dialysis bags (200-300 mL) were gently vacuum filtered (0.3 mBar) onto 47 mm Pall Supor® 450 polyethersulfone membranes (Pall Corporation, NY). Filters were immediately cryopreserved in liquid nitrogen and stored at -80 until nucleic acid extraction. DNA from filters was extracted with the MP FastDNATM SPIN DNA kit (MP Biomedicals, CA) following the manufacturer’s instructions. The V4 of the 16S rRNA gene and the V9 of the 18S rRNA gene were amplified using the primer sets which encode F515/R806 for bacteria and F1391/R1501 for eukaryotes. Both PCR and MiSeq sequencing reactions were performed following protocols provided by the Earth Microbiome Project. We sequenced samples in-house using a 300-cycle MiSeq Reagent Kit v2 (Illumina, CA) on a MiSeq platform with a 2 X 150 bp paired-end run in the presence of 25 % PhiX sequencing control DNA. Sequences are embedded in a single fastq file (one each for 16S and 18S) which pools together all the raw sequences from the project.
A LI-COR LI-193SA spherical quantum sensor and a LI-190SA flat sensor were attached to a LI-COR datalogger to record instantaneous under-ice PAR and PAR incident on the surface of the lake ice. Data were recorded in an outside incubation hole covered by an opaque tarp, unless otherwise noted. The upper 3-6 meters of each profile, depending upon ice thickness for each lake, represents conditions within the ice melt hole and is not representative of actual lake water.