Description of data

data_local_adapt_traits.csv

This dataset incorporates all plant traits measured in this experiment, as well as data on bumblebee visitation on plants, and on the reproductive success of the plants. The dataset includes plants from genotypes which were grown for 6 generations at 23°C or at 27°C plus 6 “hot” days of 30°C and visited by Bombus terrestris bumblebees or hand pollinated (control). It also includes Generation 1 (G1) plants. On the last generation, all plants were grown at 23°C or 27°C plus 6 “hot” days of 30°C.
Among the phenotypic measurements there are visual, morphological, reward traits and the composition of flower scent, obtained through volatile collection and GC-MS. To acquire bumblebee behavioral data and plant reproductive success, plants were set in groups of 24, in a 5 x 5 matrix, where we allowed up to 6 bees to visit plants sequentially (until 50-70% of plants were visited). Each matrix contained plants from the same pollination genotype (bumblebee genotype or control) and compared the attractiveness and reproductive success of 12 hot-genotype plants versus 12 ambient-genotype plants.
There are a few mismatches between the phenotypic data and the behavioral/fitness data. Not all plants we phenotyped could be used in behavioral assays and vice versa. This is the main source of NAs in this document.

Description of variables

• temp_environment: The temperature at which we grew all plants on the last generation, during which we made all measurements. H means hot environment (27°C plus 6 “hot” days of 30°C) and C means ambient environment (23°C).
• temp_genotype: The temperature at which plant genotypes were kept during 6 generations. C means ambient-genotype (kept at 23°C) and H means hot-genotype (kept at 27°C plus 6 “hot” days of 30°C).
• poll_genotype: The pollinator visitation regime under which plant genotypes were kept during 6 generations. B means bumblebee-genotype (plants visited by Bombus terrestris bumblebees) and H means control (plants that were hand pollinated). G means G1 plants.
• rep: All treatments were divided into replicates A and B. A and B were composed of 49 plants belonging to different full-sibling families. At the start of the experimental evolution, plants from the same replicates belonged to the same 49 families across treatments.
• cohort: planting week. Plants from all treatments were grown and tested over a six-week period.
• plant: Individual ID for each plant.
• matrix: code indicating whether plants were tested in the same matrix during bumblebee visitation bioassays. For instance, HB1 means that bumblebee-genotype plants were tested in the hot environment on week one, and CH4 that control plants were tested in the ambient environment on week 4.
• height: plant height measured in cm, to the nearest 0.5 cm. Plants were measured using a ruler, from the base to the top of the highest inflorescence.
• leaves: total number of leaves on the week that the plants were tested.
• flowers: total number of open flowers, counted on the week when plants were tested, and before bioassays.
• nectar_per_flower: estimation of nectar amount per flower, in microliters (µl). Nectar was collected from three flowers using microcapillaries, and the resulting amount was then divided by 3.
• nectar_per_plant: estimation of nectar amount per plant, calculated by multiplying the nectar amount per flower by the number of open flowers. As we only could measure nectar amounts on three flowers per plant, when nectar per flower was 0, we added n/a on nectar per plant, as we cannot know for sure whether the plant had absolutely no nectar.
• flowering_time: days from sowing until opening of the first flower.
• petal_area: area per flower petal. We measured the petal area of all four petals from three flowers by scanning the flat, excised petals and calculating the area on ImageJ.
• uv_absorbance: The relative area of UV absorbance per petal. Calculated by dividing the area of UV absorbance per petal (in cm2, obtained by photographing flattened petals using a camera with a UV tranmission filter) by the petal area.
• uva: Relative diffuse reflectance of UVA color, measured 0.5 mm below the top edge of the petal with a spectrometer.
• uvb: Relative diffuse reflectance of UVB color, measured 0.5 mm below the top edge of the petal with a spectrometer.
• yellow: Relative diffuse reflectance of yellow color, measured 0.5 mm below the top edge of the petal with a spectrometer.
• chosen: Indicates whether plants were visited (1) or not (0) by bumblebees in the pollination bioassays.
• seed_number: Total number of seeds per plant, collected and counted after the bioassays, and once fruits were mature.
• fruit_number: Total number of siliques per plant, collected and counted after the bioassays, and once fruits were mature.
• seeds_per_fruit: number of seeds per plant divided by the number of fruits per plant.
• fruit_set: number of fruits divided by the number of open flowers at the time of pollination.
• bee_flower_visits: number of flower visits by bumblebees received by each plant.
• bee_plant_visits: number of overall plant visits by bumblebees received by each plant.
• bee_seconds_per_flower: average seconds spent per flower by a bumblebee, when visiting a plant. Measurement obtained by dividing the time spent per plant by the number of flower visits.
• bee_seconds_per_plant: cumulative time spent by all bumblebee visitors on a plant. Calculated by measuring the time elapsed between arrival on a plant and departure from it for each bumblebee visitor, and then adding the times together for an individual plant.
• run: code indicating when the scent samples from the plant were measured in the GC-MS. Plants sharing the code were measured on the same run.
• Benzaldehyde - ZZ.alpha.Farnesene_D: Individual flower scent compounds identified and quantified by the GC-MS. Scent amounts expressed as pg / flower / l / h, transformed by ln(1+x).
• scent_sum: Sum of the amounts from all identified flower scent compounds, expressed as pg / flower / l / h, transformed by ln(1+x).
• sum_aromatics: Sum of the amounts from all identified aromatic flower scent compounds, expressed as pg / flower / l / h. Aromatic scent compounds detected in this study are: benzaldehyde, methyl Benzoate, phenylethyl alcohol, 2-amino benzaldehyde, p-anisaldehyde, methyl anthranilate, phenylacetaldehyde, benzyl Nitrile, methyl salicilate and indole.
• lnBenz - lnaromatics: ln(x + 1)-transformed values of scent emission for the individual scent compounds, their sum (lntotal), and the sum of all aromatic compounds (benzaldehyde, methyl Benzoate, phenylethyl alcohol, 2-aminobenzaldehyde, p-anisaldehyde, methyl anthranilate, phenylacetaldehyde, benzyl Nitrile, methyl salicilate, and indole). Before transformation, scent amounts are expressed as pg / flower / l / h.

Due to a formatting error, in a previous version of this document, values from Benzaldehyde to sum_aromaticswere already ln(x + 1)-transformed, and values from lnBenz to lntotal were doubly ln(x + 1)-transformed. We have now corrected this issue and the data displays the intended non-transformed (Benzaldehyde to sum_aromatics) and ln(x + 1)-transformed (lnBenz to lnaromatics) scent amounts.

reaction_norms.csv

This dataset includes the temperature reaction norms of all plant phenotypic traits measured, and from plants of all treatments. For every treatment, we grew pairs of half-sibling plants, one at an ambient (23°C) and one at a hot (27°C plus 6 “hot” days of 30°C) temperature, which allowed us to obtain temperature reaction norms by calculating the difference between the trait values of sibling plant pairs grown at ambient and hot temperature environments.

Description of variables

As we display the temperature reaction norms from the plant phenotypic traits present in the dataset above (data_local_adapt_traits.csv), all the column headings have the same meaning as in that dataset. Except for the fact that here, instead of showing individual measurements, we are showing temperature reaction norms of traits, obtained from half-sibling plant pairs grown at different temperature environments.
N/a are present if data was missing from one (or both) plants in the half-sib pair.

haldanes.csv

This dataset contains the evolutionary rates, measured in Haldanes, of each trait per replicate, for all treatments. Haldanes were obtained by calculating the absolute difference between the replicate trait means at Generation 8 vs. 1, dividing that by the pooled SD per trait per replicate, and dividing the result by the number of generations under selection (6). A more detailed description and the pertinent formulas are available in the methods section of the main manuscript.

Description of variables

• temp_genotype: The temperature at which plant genotypes were kept during 6 generations. C means ambient-genotype (kept at 23°C) and H means hot-genotype (kept at 27°C plus 6 “hot” days of 30°C).
• poll_genotype: The pollinator visitation regime under which plant genotypes were kept during 6 generations. B means bumblebee-genotype (plants visited by Bombus terrestris bumblebees) and H means control (plants that were hand-pollinated).
• rep: All treatments were divided into replicates A and B. A and B were composed of 49 plants belonging to different full-sibling families. At the start of the experimental evolution, plants from the same replicates belonged to the same 49 families across treatments.
• traits: All of the phenotypic traits which were measured from plants of all treatments. As mentioned above, haldanes were calculated per trait per treatment replicate.
• haldane: Evolutionary rates measured in Haldane's index for synchronic comparisons.
• trait_cat: column indicating whether the traits used to calculate haldanes consist of flower scent compounds (scent) or all other phenotypic traits measured (non-scent).