Common Garden Assay
To determine the sources of phenotypic variation in the field, we brought samples back to McGill University to use in a common garden growth assay. Whereas phenotypic variation in the field is due to a mixture of environmental and genetic sources, growth in a common garden removes environmental variation, isolating genetic variation. A single colony of L. minor was collected from each of 3 microsites for each site and preserved in tubes filled with natural sample water and stored in the dark during transport.
Back in the lab, all tubes were placed under artificial grow lights (100 µmols/m2/s) until the plants had doubled in number, consisting of at least two detached colonies which would be used to found two clonal replicates for each microsite. The common garden assay was done using 500mL Erlenmeyer flasks. There was a total of 204 flasks (34 sites x 3 microsites x 2 replicate flasks). Flasks were filled with 350mL of growth media, diluted Hoagland’s E-media ([N]=5000 µgL-1, [P]= 780 µgL-1, pH=7.0 +/-0.05) (recipe in Supplementary information, Table S2), plugged with a foam stopper, and then autoclaved. A single colony (3-4 attached fronds) was used to inoculate each flask. These initial fronds were first marked on their ventral surface with a small dot with a permanent marker to later track generations. This was to ensure that phenotypes were only measured on fronds at least two generations younger than those sampled from the field.
All flasks were placed in one of two identical controlled growth chambers of the McGill phytotron (200 µmols/m2/s light, 200C, 70% relative humidity, with a 14/10 light-dark cycle). The two replicates were blocked, with one replicate of each microsite in each chamber. The 102 flasks in each growth chamber were randomly positioned, leaving a 15cm boundary from the chamber wall on all sides. The common garden assay was broken into three 10-day phases, separated by two transfers. Transferring the plants to fresh media every 10 days prevented nutrient depletion, all-the-while limiting the growth of phytoplankton whose differential abundance among flasks could influence nutrient availability and plant growth. To remove any maternal or carry over effects, we tracked generations to ensure that we only measured the phenotypes of plants at least two generations younger than the initial plants brought back from the field. The first, 10-day preliminary acclimation phase served to ensure an equal physiological starting point of all plants before we began to track population growth rates, and to ensure the removal of all fronds initially present in the assay. After these 10 days of growth, all fronds marked with a black dot were discarded, and a single younger colony was randomly selected and used to inoculate identical flasks with fresh growth media, after the oldest frond in this colony was again marked. After a second 10 days of growth, flasks were again removed, and all plants were transferred to fresh media before being returned to the chambers for a final 10 days of growth. After each of the two transfer dates, all flasks were returned to the same growth chambers, but their positions within each were independently randomized. At the end of the experiment, the total number of fronds was counted in each flask and used to calculate rates of exponential population growth (over the final 20 days). From each flask, we randomly sampled 10 individuals (on average ~15% of the population) for whom we measured frond area and root length by imaging (plants were pressed onto a sheet including a reference ruler and photographed at a standard 20cm distance) and subsequent image analysis using Image J (Abràmoff et al. 2004). Second generation fronds (marked with a black dot) were excluded, as were immature fronds (that didn’t yet have two daughter fronds budding from them).