This dataset includes macroinvertebrate and water quality data from 32 lakes sampled in the Northwest Territories, Canada, from the Dempster and Inuvik-Tuktoyaktuk Highways during July and August of 2017 and 2018. It was collected as part of a project to determine how physicochemical variables affect macroinvertebrate and fish communities in small Arctic lakes.

Study sites

Our 32 study lakes were located within the boreal to tundra transition area. Eight of the lakes were sampled in August 2017 in the boreal forest region within the Gwich'in Settlement Area (GSA), and 24 of the lakes were sampled in August 2018 in the Inuvialuit Settlement Region (ISR) where tundra vegetation dominates. Our selection of lakes was not random, but was determined by a combination of logistical, licensing, and scientific considerations. As our project included fish community sampling, we had heavy gear, including gill nets and motorboats to transport to each lake, limiting the distance we could travel from the highway. Peat hummocks in the tundra and boggy terrain in the boreal region made it especially difficult to traverse long distances carrying heavy gear. Licenses were required for the research project as a whole (Northwest Territories Scientific License), and for fish collection in particular (Department of Fisheries and Oceans Scientific Collection Permit). These licenses required community consultation in the choice of lakes sampled so that lakes of cultural or recreational value to communities could be excluded. Therefore, for logistical reasons we considered all lakes within 300 m of the highways as possible sites. We also included five sites further away from the highway (0.6-3.5 km from the highway) that were accessible from the Trail Valley Creek Field Station (68.741°N, -133.499°W) north of Inuvik. For licensing reasons, and to respect the wishes of the communities, we excluded culturally significant lakes such as the Husky Lakes near Tuktoyaktuk or Dolomite Lake near Inuvik. Finally, for scientific reasons, we deliberately attempted to select lakes that varied in latitude and surface area. Our rationale was that we were hoping to examine environmental gradients, so lakes at different latitudes and of different sizes would be likely to vary in temperature, depth, fish presence/absence, and other important environmental variables.

Biological data

We collected macroinvertebrate samples using a modified version of the Ontario Benthic Biomonitoring Network protocol (OBBN) (Jones et al. 2007). We collected three replicate samples using a 500 μm D-net to kick and sweep macroinvertebrates. The modifications to the original OBBN protocol were that each replicate was collected over three minutes, and replicates were taken along parallel transects from the shore until 1 m depth was reached. We used a three-minute traveling kick and sweep per replicate rather than ten minutes because of the abundance of organic matter on the lake bottom that quickly clogged the D-net, preventing further sample collection. To conduct a traveling kick and sweep, the researcher “vigorously kicks the substrate to disturb it to a depth of ~5 cm” and sweeps the net “back and forth and up and down” to capture disturbed invertebrates (Jones et al. 2007). We did not take replicates at random locations in the lake as in the conventional OBBN protocol due to inaccessibility of different parts of the shoreline. A boat was not always available when we were collecting macroinvertebrate samples, and the boggy terrain often made it difficult to walk the shoreline without sinking into the peat. We were able to discern three types of substrate in our study lakes: mats of sphagnum moss, sphagnum moss with floating and emergent vegetation, and mixed gravel/cobble. For most lakes we visited, the littoral zone was quite homogeneous, consisting primarily of sphagnum moss, below which was a layer of fine silt. Many lakes also had floating vegetation, such as water lilies, and emergent sedges or grasses nearshore. A smaller number of lakes had cobble/gravel habitats in addition to sphagnum substrate. In cases where there were no clear indications of differing habitat types along the shore, we collected replicate samples at least 10 m apart. In cases where obvious habitat heterogeneity existed (e.g. sphagnum bottom versus rocky bottoms), we collected at least one sample in each habitat type. Each sample was preserved in 95% ethanol and brought back to the laboratory for identification.

Using a 500 µm sieve and a dissecting microscope, we identified macroinvertebrates to the order and family level in the laboratory according to the OBBN tally sheet (Jones et al. 2007). We performed quality assurance and quality control on taxonomic accuracy and sorting efficiency following the Canadian Aquatic Biomonitoring Network (CABIN) protocol (McDermott et al. 2014). We counted full samples from the eight lakes collected in 2017 from the GSA, while we used subsampling by weight for the 24 lakes sampled the next year in the ISR due to large sample sizes and high abundances of macroinvertebrates. For the ISR samples, we weighed an entire sample, and then weighed out a small amount of sediment representing approximately 10% of the sample for analysis. If we did not count at least 100 individuals in this subsample, then we added more sediment until at least 100 individuals were identified for that particular subsample. We counted two more subsamples for each lake in the same manner, such that a minimum of 300 individuals were identified from each lake. If fewer than 300 individuals were present in the whole sample, we processed the entire sample. This methodology resulted in a mean of 437 individuals identified for each ISR lake with a standard deviation of 148. We conducted subsampling without replacement, meaning that we did not return sediment to the sample after each subsample.

We collected fish community data in both sampling years. In 2017, we performed gillnetting according to the Ontario Broadscale Monitoring (BsM) protocol (Sandstrom et al. 2013). In 2018, we modified the BsM protocol in that we checked nets every 45-60 minutes and deployed nets an average of 11 hours in small to medium lakes (<500 ha) and 37 hours in larger lakes (>500 ha). We changed the sampling methodology to address community concerns regarding the more lethal method of the BsM protocol where gillnets are left overnight (16-22 hours). Because of differences in sampling between field seasons, we used the occurrence of fish species instead of catch-per-unit-effort. In our statistical analysis, we used occurrence data for the most common fish that were caught, including whitefish (Coregonus clupeaformis or Coregonus nasus), northern pike (Esox lucius) and least cisco (Coregonus sardinella).

Water quality and chemistry data

We measured a suite of water quality and chemistry parameters in lakes where macroinvertebrates were sampled. We obtained water clarity measurements using a Secchi depth at the deepest point of the lake by lowering the Secchi disk over the shady side of the boat. We used a Eureka Manta multiparameter probe (Eureka Water Probes) to take water quality measurements in the littoral zone at the same locations where macroinvertebrates were collected. The measurements were taken before collection of the invertebrates to avoid changes that might result from disturbing the sediment. The probe measured pH, conductivity, turbidity, and water temperature. Additionally, we collected a nearshore water sample to measure total suspended solids (TSS), chlorophyll-a¸ total phosphorus, total nitrogen, dissolved organic carbon, and calcium. To measure TSS, we followed standard operation procedures according to method 2540 D (Rice et al. 2017). We measured chlorophyll-a concentrations by filtering 250 mL of each water sample through Fisherbrand G4 glass fiber filters. We then used methanol to extract the chlorophyll-a from the filters and measured the concentration dissolved in the methanol using a fluorometer (Turner TD700) (Symons et al. 2012). We used a Perkin Elmer Optima 8000 Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES) to measure calcium concentrations, and a Shimadzu TOC-LCPH carbon and nitrogen analyzer (Shimadzu Corp.) to measure dissolved organic carbon and total nitrogen. We measured total phosphorus by first digesting a portion of the sample in an autoclave using ammonium persulfate and sulfuric acid according to EPA method 365.1. Then, we used a SEAL Continuous Segmented Flow Analyzer (SEAL Analytical, Inc.) to measure total phosphorus colorimetrically.

In addition to the described water quality parameters, we collected sediments, as they have been shown to be important in determining the structure of macroinvertebrate communities in lakes (De Sousa et al. 2008; Namayandeh and Quinlan 2011). We determined sediment size distribution by drying sediment samples at 105 °C for eight hours and then using a sieve shaker for ten minutes per sample to separate grain sizes using seven different sieve sizes (4 mm, 2 mm, 1 mm, 500 µm, 250 µm, 125 µm, 63 µm). To calculate % organic matter and % CaCO3 in sediments we used 5 g of sediment from each sample that was <2 mm and followed the standard operating procedure for loss on ignition (Santisteban et al. 2004).

References

De Sousa, S., Pinel-Alloul, B., and Cattaneo, A. 2008. Response of littoral macroinvertebrate communities on rocks and sediments to lake residential development. Can. J. Fish. Aquat. Sci. 65(6): 1206–1216. doi:10.1139/F08-031.

Jones, F.C., Somers, K.M., Craig, B., and Reynoldson, T.B. 2007. Ontario benthos biomonitoring network: protocol manual. Available from https://desc.ca/sites/default/files/OBBN2007finalapril18c.pdf.

McDermott, H., Paull, T., and Stachan, S. 2014. Canadian Aquatic Biomonitoring Network (CABIN) laboratory methods: Processing, taxonomy, and quality control of benthic macroinvertebrate samples. In Environment Canada. doi:10.1089/dna.1998.17.321.

Rice, E.W., Baird, R.B., and Eaton, A.D. (Editors). 2017. Standard methods for the examination of water and wastewater, 23rd edition. American Public Health Association, American Water Works Association, Water Environment Federation, Washington, D.C.

Sandstrom, S., Rawson, M., and Lester, N. 2013. Manual of instructions for broad-scale fish community monitoring; using North American (NA1) and Ontario small mesh (ON2) gillnets. Ontario Minist. Nat. Resour. Peterborough, Ontario. Version 2013.2 35 p. + Append.

Symons, C.C., Arnott, S.E., and Sweetman, J.N. 2012. Grazing rates of crustacean zooplankton communities on intact phytoplankton communities in Canadian Subarctic lakes and ponds. Hydrobiologia 694(1): 131–141. doi:10.1007/s10750-012-1137-6.