Data from: Thermal plasticity in development and diapause strategy in a temperate butterfly across a latitudinal gradient
Data files
Sep 21, 2023 version files 721.24 KB
-
README.md
5.86 KB
-
Verspagen2023_ClutchData.csv
103.73 KB
-
Verspagen2023_FatData.csv
161.18 KB
-
Verspagen2023_IndividualData.csv
421.48 KB
-
Verspagen2023_InstarData.csv
28.99 KB
Sep 21, 2023 version files 721.24 KB
Abstract
- Trade-offs among traits are central to life-history theory and often closely linked to an organism’s fitness. Understanding how these trade-offs vary among populations and across environments is therefore important to more accurately predict species’ responses to future climate change. However, the extent to which responses vary across populations remains unknown because few studies investigate intraspecific differences.
- We performed a full-factorial split-brood common garden experiment to test how variation in rearing temperature affects developmental timing and other traits important for survival during diapause in the Glanville fritillary butterfly (Melitaea cinxia). Pre-diapause larvae originating from four regions across a latitudinal cline across Europe were reared at four temperatures (25, 28, 31, and 34 °C), and we used a reaction norm approach to test for evidence of genetic differentiation and variation in developmental plasticity across regions.
- We found clear signs of genetic differentiation in multiple developmental traits, as well as differences in developmental plasticity. Northern larvae entered diapause in the fourth instar when the temperatures were low, whereas southern larvae did so in the fifth instar. As a result, development time is canalized with regards to temperature in northern larvae: due to entering diapause one stage earlier, they develop fast even in the cold, whereas southern larvae always develop slower, especially at low temperatures. As a trade-off, northern larvae have a lower body mass when reared at cooler temperatures compared to southern larvae, and they show increased plasticity in diapause mass. No clear clinal patterns were found in relative fat content.
- Our results show that trade-offs between body size, development time and growth rate can vary within species living across environmental clines, possibly as a consequence of natural selection to local environmental conditions or other genetic constraints. This variation highlights the importance of recognising the context-dependency of relationships between important life-history traits and their interactions with local environments in predicting species’ responses to climate change.
README: Data from: Thermal plasticity in development and diapause strategy in a temperate butterfly across a latitudinal gradient (Verspagen et al., Functional Ecology)
These datasets contain all data and code used in the analyses of this manuscript.
Description of the data and file structure
Data files (.csv) are named as "NameYear"_"DataIdentifier".csv
Column title abbreviations are largely self-explanatory.
More specific details for each dataset:
Verspagen2023_ClutchData.csv:
In this file we record the development of the experimental group as a whole every four days.
The first five columns together form a unique identifier for each experimental group. They contain information on the location of origin (Origin), family (Female.ID), Clutch (Clutch.ID), Temperature and replicate. The next column shows whether the experiment was done in 2020 (Finland, Estonia and Spain) or 2022 (Finland2 and Belgium).
The Day column identifies which day since the second instar (i.e. start of experiment for this experimental group) the measurement was done, and the Date column represents the calendar date of this day.
Mass is the weight in mg of the group and N the number of larvae weighed at this time. Mean.individual.mass is calculated as Mass/N and represents the mean mass of an individual of this group at this timepoint.
N was also used for testing survival (not included in this paper) as a measure for how many alive larvae were present over time. However, in certain cases a larva was missed during weighing on one day but found during the next round (i.e. four days later), indicating that it was alive also during the earlier timepoint. A corrected value of N is found in N.survival.
Verspagen2023_IndividualData.csv:
In this file we record development of each individual as measured at diapause.
The first six columns correspond to the first six columns in the ClutchData file and together identify the experimental group. The column Diapause.number adds a number between 1-15 for each individual in this experimental group.
Column Fat.diapause indicates if a larva was sacrificed for fat content analysis (F) or not, and thus left to diapause (DP).
Diapause.date and Diapause.mass represent the date and weight (in mg) at diapause. Start.date is the date that the clutch molted to the second instar, indicating the start of the experiment and is thus the same for all larvae in an experimental group. Development.time (days) is calculated by subtracting Diapause.date from Start.date.
Mean.start.mass correspons to Mean.individual.mass in the ClutchData file and is used together with development time to calculate growth rate according to the following formula: (ln(Diapause.mass) - ln(Start.mass))/Development.time
A PCA was performed on diapause mass and development time, and the values from PC1 are present in the column PCA. A threshold value (see methods section) was used to assign an individual to instar 4 or 5 (column Diapause.instar).
Verspagen2023_FatData.csv:
In this file we record fat content and it includes only the individuals for which this was measured (i.e. those with "F" in the column Fat.diapause in the IndividualData file.
The first 15 columns of this file are the same as in the IndividualData file.
In the column Dry.mass we record the weight of the larva (mg) after drying, and this mass thus excludes water. Water.content is calculated by subtracting the dry mass from the diapause mass. Relative water content is the ratio of water that was present in the larvae, and calculated as Dry.mass/Water.content.
The column Fat.free.mass contains the mass of the larvae after drying and extracting of fat. Fat content is calculated by subtracting the fat free mass from the dry mass. Relative.fat.content is the ratio of fat that was present in the dry larvae and calculated as Fat.content/Dry.mass
Verspagen2023_InstarData.csv:
In this file we record when a clutch transitioned to the following instar.
The first six columns correspond to those in the other files and together uniquely identify each experimental group.
Instar.2.date shows when the larvae in one clutch transitioned to the second instar, marking the start of the experiment for this group. As a rule, we recorded the date that 50% of the clutch transitioned, but in practice larvae in one clutch were well-synced and thus close to 90% of larvae transitioned on this day.
Columns Instar.3.date and Instar.4.date and Diapause.date show the dates that at least 50% of the experimental group transitioned to either third or fourth instar or diapause. For diapause date, this information was taken from the IndividualData file.
Columns Instar.2 and Instar.3 contain the amount of days the experimental group stayed in this instar and is calculated by subtracting Instar.2.date from Instar.3.date (Instar.2) and Instar.4.date from Instar.3.date (Instar 3). Instar.4 also shows this information but is calculated using the individual diapausing dates.
An example for experimental group Finland2, Female/family 148, Clutch 11, Temperature 25, Replicate a:
ID Diapause date Date instar 4 Instar.4 (days) Diapause instar
5 24.4.22 22.4.22 2 4
6 24.4.22 22.4.22 2 4
7 24.4.22 22.4.22 2 4
4 24.4.22 22.4.22 2 4
9 24.4.22 22.4.22 2 4
10 24.4.22 22.4.22 2 4
11 25.4.22 22.4.22 3 4
12 25.4.22 22.4.22 3 4
13 25.4.22 22.4.22 3 4
14 25.4.22 22.4.22 3 4
15 30.4.22 22.4.22 8 5
8 24.4.22 22.4.22 2 4
Average 2.833333333 days
Larvae that diapause in fourth instar go to diapause almost directly after transitioning to the fourth instar, whereas the larva that diapauses in fifth inster (ID 15) spends the normal amount of time of roughly one week in this instar. This results in an average of 2.83 days spent in instar 4.
Code/Software
All analysis were performed in R by using the script Analysis.R