DOI: 10.5061/dryad.zs7h44jjx

Data from an experiment where 1000 L plastic mesocosms (ponds) were filled with water from a mesotrophic lake. The lake water was piped into a ~100,000 L reservoir near the experimental site, and was sampled at the same time as the ponds. The ponds were grouped into meta-ecosystems, each containing one or two ponds. Most meta-ecosystems were composed of two ponds, with different levels of connectivity (also referred to as Movement in the code and data) between the upstream and downstream.

Every week, a pre-determined percentage of the pond volume (10, 20, 30, or 40 %) was emptied from the downstream pond, moved from the upstream to downstream pond, and used to refill the upstream pond with lake water. In the code and data, the upstream and downstream is often coded as Node 1 and 2, respectively. A control connectivity (0 %) was also included in the experiment and did not involve any water transfer.
The control connectivity ponds were alone in their respective metaecosytems, and are coded as Node 1.

The experiment also featured a nutrient enrichment treatment, ranging from 0 to 200 μg P L^-1 per week. Nitrogen was also supplemented at the same molar ratio as present in the lake of 31:1 N:P. Nutrients were added after the connectivity treatment had been applied (same day).

For more details on the experimental design, see the associated paper, published in Royal Society B.

Description of the data and file structure

Unless otherwise specified, all .csv files use commas (,) to separate columns and periods as decimal points (.).

Data/Treatments.csv and Data/Treatments (alphabetical).csv contain a mapping between the ID assigned to each pond, the connectivity and nutrient enrichment treatments, and the position of the pond in the meta-ecosystem.

• Mesocosm: ID attributed to a pond.
• Movement: Connectivity treatment (% pond volume displaced per week).
• NutrientP: Nutrient enrichment treatment (μg P added per L per week).
• Node: Upstream (1) or downstream (2) pond.

Data/Metaecosystems.csv as above, but contains an additional column:

• Metaecosystem: ID attributed to a metaecosystem.

Data/Metaecosystems-Replicate.csv as above, but contains an additional column:

• Replicate: provides a (non-unique) ID ("A" or "B") to each replicate metaecosytem.

Chlorophyll a data

Data/FP-final/ExportMain/Data.bdb contains the data collected from the Fluoroprobe instrument in the native proprietary format.

Data/FP-final/ExportMain/FP.txt is a tab-separated file which contains the experimental chlorophyll a data, as exported from the Fluoroprobe instrument software. See Fluoroprobe manual for details.

• Comment contains [the pond ID (column named Mesocosm elsewhere)]-[filtered or unfiltered]. All samples were measured as filtered and unfiltered through a 22 μm sieve, but ultimately, only the unfiltered data was used for the analysis.

Data/Pilot/FP-dilution series SEPT292022.txt is a tab-separated file which contains the chlorophyll a data for the self-shading correction, as exported from the Fluoroprobe instrument software. See Fluoroprobe manual for details.

• Comment contains [the pond ID (here, a single sample was used, from the highly concentrated F4 pond)]-% dilution. The [% dilution] indicates the percentage to which the sample was diluted. As above, we also measured samples that were filtered through a 22 μm sieve, but ultimately did not use this data.

Data/FP-final/FP_clean.csv contains the tidy data used for the data analysis:

• Mesocosm: ID attributed to a pond.
• Date: Sampling date in YYYY-MM-DD format.
• Green.Algae, Bluegreen, Diatoms, Cryptophyta, Yellow.substances: different classes of phytoplankton (and yellow substances) as detected by the Fluoroprobe using the factory-calibrated parameters (μg L^-1). These values were not used in our analysis.
• Chla: total chlorophyll a (μg L^-1) corrected for self-shading at high concentrations (see paper, Supplemental Figure 3).

Nutrient data

Data/Nutrients/N-TN DN Gonzalez (McGill) - LEAP 2022.xlsx has the raw total and dissolved nitrogen data (analysed by technicians at the GRIL analytical laboratory at UQAM (Université de Quebéc à Montéral)). The different sheet names correspond to dates on which the samples were analysed, not collected.

• Sample ID: ID given to various controls and samples. The samples are named as follows: [Pond ID] [DN or TN, for dissolved or total nitrogen] ((X10)). Dissolved samples were filtered through 0.22 μm syringe filters and represent the dissolved fraction of the total nitrogen. Some samples, identified with (10X) were diluted by a factor of 10 to provide more accurate measurements.
• Concentration (ppm): The measured total nitrogen concentration in the sample (ppm; equal to mg/L).
• Corrected conc. (ppm): The measured concentration corrected for the negative control (Sample ID: EAU NANO + PERSULFAT) and, if necessary for the sample dilution.
• NOTES: Comments from the analyst.
• The smaller table on the right, with column names "Standard ID", "Concentration (ppm)", "Peak height", and "CV (%)" contains data that were used to calculate a calibration curve (see Supplemental Text 1 in the paper's electronic supplemental materials for a brief method summary and references).

Data/Nutrients/P Gonzalez (McGill) - LEAP 2022.xlsx has the raw total and dissolved phosphorus data (analysed by technicians at the GRIL analytical laboratory at UQAM (Université de Quebéc à Montéral)). The different sheet names correspond to dates on which the samples were analysed, not collected.

• Sample ID: ID given to various controls and samples. The samples are named as follows: [Pond ID] [DP or TP, for dissolved or total phosphorus]. Dissolved samples were filtered through 0.22 μm syringe filters and represent the dissolved fraction of the total phosphorus.
• Absorbance (@890nm): Measured absorbance at 890 nm wavelength.
• Absorbance Fe (@890nm): unused
• Concentration (ug/L): Total phosphorus concentration in the sample (corrected for dilution, if present).
• NOTES: Comments from the technician. Some samples, identified here, were diluted by a factor of four to provide more accurate measurements.
  • The smaller table on the right, with column names "Standard ID", "Concentration (ppm)", "Absorbance (@890nm)", and "CV (%)" contains data that were used to calculate a calibration curve (see Supplemental Text 1 in the paper's electronic supplemental materials for a brief method summary and references).

Data/Nutrients/Nutrients.csv contains the tidy nutrient analysis data used for the analysis:

• Date: Sampling date in YYYY-MM-DD format.
• Mesocosm: ID attributed to a pond.
• Nutrient: nutrient analysed (nitrogen (N) or phosphorus (P)).
• Fraction: Dissolved (filtered through 0.45 μm syringe filter).
• conc: the concentration measured (units specified in unit column - μg/L for P and mg/L for N)
• unit: specifies the units used for conc column.

Environmental data (YSI)

Data/YSI_raw/ folder contains files YSI1.csv through YSI6.csv along with YSI Aug 15.csv. WARNING: these files contain incorrect dates (see Data/YSI_date_correction.csv for the correction key). The columns are as in Data/YSI_clean.csv, except Site, which was renamed to Mesocosm in the clean version (see below).

Data/YSI_date_correction.csv contains the correction key for the files in the Data/YSI_raw folder:

• MeasuredDate: incorrect date in the Data/YSI_raw files.
• ActualDate: the true date on which the measurements were taken

Data/YSI_clean.csv contains the data with the corrected dates:

• Timestamp: Date and time at which measurements were taken. Note: time may be unreliable in some cases.
• Mesocosm: ID attributed to a pond.
• Specific Conductance (uS/cm), Dissolved Oxygen (mg/L), pH_1 (Units), Temperature (C) are self-explanatory.
• The remaining columns are not useful.

Zooplankton data

Data/Zooplankton/ZoopCounts.csv contains the raw zooplankton counts along with some comments:

• Sample Date Sampling date, in Month (abbreviated)-Day format (all in 2022).
• Pond ID: ID attributed to a pond.
• Date Counted: Date the sample was counted, in Month (abbreviated)-Day format (all in 2022).
• Transfer (%): Connectivity treatment (% pond volume transferred per week).
• Nutrients: Nutrient treatment (μg P added per L per week).
• Node: Upstream (1) or downstream (2) pond.
• Counter: One of two people that performed the task of counting (Michelle or April).
• Daphnia to Copepoda: Crustacean zooplankton counts, per taxonomic group (number per 4 L water volume).
• TOTAL: Total numbers of crustacean zooplankton (number per 4 L water volume).
• Nauplii to Rotifers: Other forms of zooplankton, and other organisms. Note that Rotifers were only counted by Michelle and not included in the final analysis (number per 4 L water sample volume).
• Notes: Notes about the sample, along with a qualitative assessment of the rotifer species present reported by Michelle.

Data/Zooplankton/ZoopCountsClean.csv is as above, but without the TOTAL and Rotifers, which were excluded from the analysis along with a new column:

• Date: The sampling date in YYYY-MM-DD format.

Data/Zooplankton/ZoopCountsCleanLong.csv is as above, but in long form:

• Taxa: The different taxa.
• Abundance: The number of individuals per 4 L of sample volume.
• Note that columns Daphnia and Other were excluded from the Taxa column due to a lack of individuals and irrelevance to the study, respectively.
• Additionally, the columns described for the file Data/Treatments.csv are included.

Data/Zooplankton/ZoopBiomass.csv is as above, but with a species-based estimate for the per-individual dry mass:

• Dry mass: Species-based estimate for the per-individual dry mass (mg/individual)
• Ref: Source for the dry mass estimate.
• Biomass: The total biomass per individual per sample (mg per 4 L of sample volume)

Data/Zooplankton/ZoopBiomass.csv is as above, but with a species-based estimate for the per-individual dry mass:

• Dry mass: Species-based estimate for the per-individual dry mass (mg/individual)
• Ref: Source for the dry mass estimate.
• Biomass: The total per-species biomass per sample (mg per 4 L of sample volume)

Data/Zooplankton/ZoopBiomassSummary.csv:

• Date: The sampling date in YYYY-MM-DD format.
• Mesocosm: ID attributed to a pond.
• Biomass: The total zooplankton biomass per sample (mg per 4 L of sample volume)

Data/ZooplanktonTraits/Hebert2016-short.csv is a csv2 file (comma for the decimal point and a semicolon for the separator) and contains data used for the species-specific individual-based dry mass estimates. This file contains a subset of the relevant columns published as a data paper by Hébert et al. (2016; see "Zooplankton Dry Mass Estimation" section below).

• Species: Scientific name of species. Note that "-(M)" indicates male individuals.
• Genus: Scientific name of genus.
• Group: Scientific name of a more general taxonomic group (Copepoda or Cladocera).
• Habitat: Type of habitat (Freshwater or Marine).
• Dry.mass: Dry body mass of a single individual (mg per individual).
• Ref.dm: Reference code for dry mass. See Hébert et al. (2016) for reference key.

Doc/MesocosmLayour.pdf is a diagram which shows the spatial layout of the experimental ponds.

R/ contains the code necessary to fit the models and produce the figures in the paper (see section "Code" below).

Rmd/ModelFits/ contains the fitted models, helpful for generating the figures (see section "Code" below).

DESCRIPTION contains a list of R packages used for this project (see section "Code" below).

LEAP 2022.Rproj is a RStudio project file helpful for executing code included in the project (see section "Code" below).

Zooplankton Dry Mass Estimation

Three references were used for estimating dry weight of various zooplankton (see R code in R/DataGetters/getZoopData, in the function getZoopBiomass()):

[1] Hébert, M.-P., Beisner, B.E. and Maranger, R. (2016), A compilation of quantitative functional traits for marine and freshwater crustacean zooplankton. Ecology, 97: 1081-1081. https://doi.org/10.1890/15-1275.1

[2] Rosen, R. A. (1981). Length-Dry Weight Relationships of Some Freshwater Zooplanktona. Journal of Freshwater Ecology, 1(2), 225–229. https://doi.org/10.1080/02705060.1981.9664034

[3] Anderson, D. H., & Benke, A. C. (1994). Growth and reproduction of the cladoceran Ceriodaphnia dubia from a forested floodplain swamp. Limnology and Oceanography, 39(7), 1517-1527. https://doi.org/10.4319/lo.1994.39.7.1517open_in_new

Code

To re-run the analyses for this paper, use scripts found in R/Figures/.
Here, ChlaMeta.R is used for meta-ecosystem level plotting, whereas ContourChla.R contains the code for plotting the upstream and downstream ponds separately.
These scripts automatically load the appropriate scripts from R/DataGetters which are used to load the appropriate and necessary data.

Additionally, some model fitting operations take some time, and are included in either R/Objects or Rmd/ModelFits. In these cases, the model fitting operation is commented out in the code. If you would like to re-fit some data, simply un-comment the model fitting section.

All file paths used are relative to the project directory ("LEAP2022").
If using RStudio, open "LEAP 2022.Rproj" to automatically set the correct working directory.

The DESCRIPTION file contains all required R packages.
Note: Unicode text in figures may not render properly on Mac computers.