Data from: Variation in season length and development time is sufficient to drive the emergence and coexistence of social and solitary behavioral strategies

Ruttenberg, Dee1 ; Levin, Simon1 ; Wingreen, Ned1 ; Kocher, Sarah1

Published Sep 04, 2024 on Dryad. https://doi.org/10.5061/dryad.rjdfn2zmr

Data files

Sep 04, 2024 version files 2.17 MB

Abstract

Season length and its associated variables can influence the expression of social behaviors, including the occurrence of eusociality in insects. Among bees, ants, and wasps, social behaviors can vary widely across environmental gradients, both within and between different species. While numerous theoretical models have been developed to examine the traits in a species’ life history that underlie the emergence and maintenance of eusociality, the impact of seasonality on this process is largely uncharacterized. Here, we present a theoretical model that builds on these foundational models by incorporating season length and offspring development time. The integration of these environmental and life history traits into a single, individual-focused model enables detailed examination of how these factors can shape the costs and benefits of social living. We find that longer season lengths and faster brood development times are sufficient to favor the emergence and maintenance of a social strategy, while shorter seasons favor a solitary one. We also identify a range of season lengths where social and solitary strategies can coexist. Moreover, our theoretical predictions are well-matched to the natural history and behavior of two, flexibly-eusocial bee species, suggesting our model can make realistic predictions about the evolution of different social strategies. Broadly, this work reveals the crucial role that environmental conditions can have in shaping social behavior and its evolution and underscores the need for further models that explicitly incorporate such variation to study evolutionary trajectories of eusociality. The above data is needed to reproduce Figure 4 in this manuscript, showing the matching between our theoretical predictions and the weather data.

README: Weather Data

https://doi.org/10.5061/dryad.rjdfn2zmr

Description of the data and file structure

Data available for all locations for the period 1965 to 1975 using the Center for Environmental Data Analysis (CEDA) (UnitedKigdomRawData.csv) and the Climate Data Online Search (CDOS) (ColoradoRawData.csv) for England (Dartmoor, Hexham, Inverness, Sussex) and the US (Almont and Crested Butte), respectively.

Variables used in analysis:

ColoradoRawData.csv

STATION -- Research station the data was collected

NAME -- Location the data was collected

DATE -- Date (From 1965-1975) the weather data was collected

TMAX -- Maximum air temperature on DATE in degrees Celsius

TMIN -- Minimum air temperature on DATE in degrees Celsius

TOBS-- Temperature at the time of observation

UnitedKingdomRawData.csv

ob_end_time -- End of range of data analysis

src_id -- id of weather station (Sussex = 803, Inverness = 297, Hexham = 115, Dartmoor = 1350)

*max_air_temp -- *Maximum air temperature in range of range of ob end time in degrees Celsius

*min_air_temp -- *Minimum air temperature in range of range of ob end time in degrees Celsius

TemperatureTemplate.csv

This is a generic format to standardize our loaded data into a range of exactly 10 years (1965-1975 for our analysis)

month, day, year -- three columns date from 1965-1975

Code/software

R is required to run BritishDataAnalysis.R and PlotClimateData.R; the script was created using version 4.4.0. https://github.com/kocherlab/HalictidClimateModeling/tree/main.

Methods (from Ruttenberg et al. 2024)

To contextualize our model, we sought to compare the conditions associated with social and solitary outcomes in the model to the climatic conditions associated with social versus solitary nesting in two well-characterized sweat bee species. First, we examined the season lengths of the social and solitary populations of *Halictus rubicundus. *This species occurs both in Europe and in North America; it is typically eusocial, but solitary populations have been documented at high elevations in the United States (4) and at high latitudes in the UK and Scotland (2, 31). We also examined the locations of social and solitary populations of a second, socially polymorphic sweat bee, *Lasioglossum calceatum *(32). This species is distributed throughout the palearctic, and it is most commonly eusocial (33). Solitary populations of *L. calceatum *have been documented in the northern UK and Ireland (32) as well as at high elevations in Hokkaido, Japan (3).

To estimate season lengths for the solitary and social populations of each of these species we found data available for all locations for the period 1965 to 1975 using the Center for Environmental Data Analysis (CEDA) and the Climate Data Online Search (CDOS) for England and the US, respectively (44, https://www.nrcs.usda.gov/wps/portal/wcc/home/). We elected to limit our data to these years in order to minimize the impacts of climate change. Some of our data contained many missing days. In order to allow us to use this data, we generated 100 uncorrelated seasons from this data using RMAWGEN (https://rdrr.io/cran/RMAWGEN/man/RMAWGEN-package.html, version 1). This also reduced the impact of local weather conditions and correlation between days. We used the generated weather data to estimate the season length for each of these locations by calculating the length of time between the end of the last day of the first 5-day interval where the max temperature each day was above 14°C, approximately the minimum temperature required for Halictus rubicundus *and *Lasioglossum calceatum to forage (3,13) and the end of the last day of the first 5-day interval after that where the max temperature each day was below 14°C.

Access information

Other publicly accessible locations of the data:

https://github.com/kocherlab/HalictidClimateModeling/tree/main

Data was derived from the following sources: