Dataset DOI: 10.5061/dryad.w3r22814f

Description of the data and file structure

Lifespan-dependent cognitive biases in honeybee workers

Summary

This study investigates how honeybee lifespan influences their cognitive biases and emotional states, focusing on decision-making processes. The research was conducted at two locations in Krakow, Poland—the Institute of Environmental Sciences and the University of Agriculture—between May 2018 and June 2023. A total of 14 honeybee colonies, each containing 20,000 to 40,000 workers, were used across three experiments.

The first experiment aimed to shorten the bees' lifespan through two methods: CO2 anesthesia and thoracic injury. Worker bees were collected from subunit B of their colonies during the afternoon, after being separated from non-foraging bees in the morning. The bees were then randomly assigned to three groups: untreated controls, CO2-anesthetized for 30 minutes, or injured with a needle puncture in the thorax. The bees underwent lifespan assessment in laboratory cages, followed by judgment bias tests, and control tests to observe spontaneous responses. These experiments took place at the apiary, with sample collection and testing occurring in designated laboratory settings.

The second experiment involved infecting worker bees with Nosema spores to induce a reduction in lifespan. Similar to the first, bees were collected from the colonies, and their survival was monitored after infection. The purpose was to evaluate how infection-related lifespan shortening affects cognitive biases.

The third experiment focused on "rebel workers," a naturally longer-lived group within the colony, to compare their behavior and decision-making patterns with those of shorter-lived bees. These rebel workers were also subjected to judgment bias assessments.

The judgment bias paradigm was employed to measure emotional states indirectly. Bees were trained to associate specific odors with rewards (positive stimuli) or punishments (negative stimuli). During testing, responses to ambiguous stimuli—intermediate odors—were recorded. Bees that responded optimistically (proboscis extension to ambiguous stimuli) were interpreted as exhibiting positive or risk-taking biases, while more cautious responses indicated negative biases. Results showed that bees with shortened lifespans, whether through injury or infection, exhibited more optimistic biases, tending to respond more readily to ambiguous stimuli. Conversely, rebel workers with longer lifespans displayed more pessimistic biases, suggesting a cautious foraging approach.

Overall, the findings demonstrate that honeybee decision-making and emotional states are closely tied to their lifespan. Shorter-lived bees tend to adopt riskier strategies, possibly favoring immediate gains, whereas longer-lived bees are more cautious, reflecting different affective states influenced by their life expectancy. These insights highlight the complex relationship between longevity and cognition in social insects.

How to read the files from the Supplementary Data:

1. Each file in the Supplementary Data corresponds to a separate experiment. Specifically, the data for: Experiment 1 is in data_file_Experiment 1_longevity.csv, data_file_Experiment 1_jugment_bias_test.csv, data_file_Experiment 1_spontanous_responses.csv, Experiment 2 in data_file_Experiment 2_longevity.csv, data_file_Experiment 2_jugment_bias_test.csv, data_file_Experiment 2_spontanous_responses.csv, data_file_Experiment 2_nosema_infection.csv, and Experiment 3 in data_file_Experiment 3_longevity.csv, data_file_Experiment 2_jugment_bias_test.csv, data_file_Experiment 3_spontanous_responses.csv, data_file_Experiment 2_ovariole_number.csv

2. In each file, the "Longevity" - data_file_Experiment 1_longevity, data_file_Experiment 2_longevity, data_file_Experiment 3_longevity  presents lifespan data for individual bees from different colonies and treatments. The columns include: the colony number from which each bee originated, the treatment applied (for Experiment 1: control, injury, and anesthesia with CO₂; for Experiment 2: control and infected workers; for Experiment 3: control [normal workers] and rebel workers), the individual number (with 50 bees per colony and treatment in each case), the lifespan in days and censored data were 0 means that bess alive at the end of the study or lost to follow-up, while 1 usually indicates that the event of interest (e.g., death) has occurred.

3. In each file the "Judgment_Bias_Test" - data_file_Experiment 1_jugment_bias_test , data_file_Experiment 2_jugment_bias_test, data_file_Experiment 3_jugment_bias_test presents data from the last conditioning session as well as the results of the judgment bias test. The columns include (from left to right): the colony from which the bees originated, the individual bee number, and the treatment to which the bee was subjected (see point above). The next column, "Odour CS+," describes the scent that was paired with the reward (ratio 9:1). The following two columns display the results from the last conditioning sessions, where "A+" (with a scent combination of 9:1) indicates a response to the signal associated with the reward, and "B-" (with a scent combination of 1:9) indicates a response to the signal associated with an unpleasant event. The columns labeled with ratios (9:1, 7:3, 1:1, 3:7, 1:9) show the bees' responses (proboscis extension reflex) to the exposed scents during the judgment bias test. Additionally, there is a side section showing the number of individuals responding to each scent by extending their proboscis in each colony and treatment, as well as the total responses across all colonies.

4. Each file contains a "Spontaneous responses” data_file_Experiment 1_spontanous_responses, data_file_Experiment 2_spontanous_responses, data_file_Experiment 3_spontanous_respons which serves as a control for the bias tests themselves. In this section, bees are exposed to scent mixtures without prior conditioning to CS+ and CS-. The purpose of this control is to determine whether the bees respond spontaneously to any of the scents used during conditioning, possibly due to previous exposure. The first column provides information about each individual bee (bee number), followed by the treatment it received (see point 3), and the colony from which it originated. The "Odour CS+" column describes the scent that was used in a higher proportion during the test (ratio 9:1). The subsequent columns, ranging from 9:1 to 1:9, show the responses of individual bees to the exposed scents in the form of binary data (0 or 1), indicating whether the proboscis extension reflex was observed.

5. The file "data_file_Experiment 2_nosema_infection" presents data on how individuals in both groups were infected with Nosema sp. The first column contains information about the colony of origin of the tested workers. The second column indicates the treatment (whether the individual was a control or infected with Nosema through feeding). The third column specifies the individual’s number. The last column reports the number of Nosema spores, which is an estimate of the spores per individual after normalization.

6. The file "data_file_Experiment 3_ovariole_number" includes data on the number of ovarioles in the tested control and rebel individuals. These data were collected to ensure that we are dealing with two distinct groups of bees. The table provides, from left to right, the individual bee's number, the colony it originated from, the treatment (whether it was assigned to control workers or rebel workers), and the number of ovarioles in the ovaries of each individual.

Files and variables

File: data_file_Experiment_1_longevity.csv

Description: 

Variables

• colony: number of colony
• treatment: the treatment applied: control, injury, and anesthesia with CO₂;
• no. of bee: number of tested bee
• longevity (days): the number of days the bee survived
• censored data: were 0 means that bess alive at the end of the study or lost to follow-up, while 1 usually indicates that the event of interest (e.g., death) has occurred.

File: data_file_Experiment_3_ovariole_number.csv

Description: 

Variables

• nr of bees : number of tested bee
• colony: number of colony
• treatment: normal or rebel workers 
• no. Of ovarioles: the sum of ovarioles from two ovaries

File: data_file_Experiment_2_longevity.csv

Description: 

Variables

• colony: number of colony
• treatment: the treatment applied: healthy bees and infected with Nosema apis spores
• no. of bee: number of tested bee
• longevity (days): the number of days the bee survived
• censored data: number of colony: were 0 means that bess alive at the end of the study or lost to follow-up, while 1 usually indicates that the event of interest (e.g., death) has occurred.

File: data_file_Experiment_2_nosema_infection.csv

Description: 

Variables

• colony: number of colony
• treatment: the treatment applied: healthy bees and infected with Nosema apis spores
• no. of individuals: number of tested bee
• no. Of nosema spores per digestive track: number of spores in the digestive track

File: data_file_Experiment_2_spontanous_responses.csv

Description: 

Variables

• no.of bees: number of tested bee
• treatment: the treatment applied: healthy bees and infected with Nosema apis spores
• colony: number of colony
• odour: type of odour tested as CS+ 
• 9:1: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 7:3: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 1:1: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 3:7: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 1:9: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)

File: data_file_Experiment_3_jugment_bias_test.csv

Description: 

Variables

• colony: number of colony
• no. Of bees:  number of tested bee
• treatment:  the treatment applied: normal and rebel workers 
• odour CS+ (9:1): type of odour tested as CS+ 
• CS+: show the responses of individual bees in the last session
• CS-:  show the responses of individual bees in the last session
• 9:1: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 7:3: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 1:1: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 3:7: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 1:9: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)

File: data_file_Experiment_3_longevity.csv

Description: 

Variables

• colony: number of colony
• treatment:  the treatment applied: normal and rebel workers 
• no. of bee: number of tested bee
• longevity (days): the number of days the bee survived
• censored data: were 0 means that bess alive at the end of the study or lost to follow-up, while 1 usually indicates that the event of interest (e.g., death) has occurred.

File: data_file_Experiment_3_spontanous_responses.csv

Description: 

Variables

• no.of bees: number of tested bee
• treatment: the treatment applied: normal and rebel workers
• odour: type of odour tested as CS+ 
• 9:1: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 7:3: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 1:1: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 3:7: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 1:9: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)

File: data_file_Experiment_2_jugment_bias_test.csv

Description: 

Variables

• colony: number of colony
• no. Of bees:  number of tested bee
• treatment:  the treatment applied:  healthy bees and infected with Nosema apis spores
• odour CS+ (9:1): type of odour tested as CS+ 
• CS+: show the responses of individual bees in the last session
• CS-:  show the responses of individual bees in the last session
• 9:1: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 7:3: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 1:1: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 3:7: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 1:9: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)

File: data_file_Experiment_1_jugment_bias_test.csv

Description: 

Variables

• colony: number of colony
• no. Of bees:  number of tested bee
• treatment:  the treatment applied: control, injury, and anesthesia with CO₂;
• odour CS+ (9:1): type of odour tested as CS+ 
• CS+: show the responses of individual bees in the last session
• CS-:  show the responses of individual bees in the last session
• 9:1: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 7:3: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 1:1: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 3:7: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 1:9: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)

File: data_file_Experiment_1_spontanous_responses.csv

Description: 

Variables

• no.of bees: number of tested bee
• treatment: the treatment applied: control, injury, and anesthesia with CO₂;
• odour: type of odour tested as CS+ 
• 9:1: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 7:3: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 1:1: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 3:7: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)
• 1:9: show the responses of individual bees to the exposed scents in the form of binary data (0 or 1)

Code/software

How the analyses were conducted in Statistica 13.

Longevity

The data were sourced from the following files: data_file_Experiment 1_longevity for Experiment 1, data_file_Experiment 2_longevity for Experiment 2, and data_file_Experiment 3_longevity for Experiment 3.

From the Statistica 13 interface, the following steps were taken: select the tab – Statistics – Survival Analysis – Multiple Sample Comparison – Variables (survival times, censor indicator, grouping variable), where we input data from the "longevity" column as the lifespan. The censored data is the censor indicator, and the grouping variable is the treatment or family for all three experiments – then click OK. We assign codes for complete observations – our censoring data as 1, and for censored observations (i.e., organisms that survived the experiment but are not included in our study) as 0. Then, click on "Codes for layers," select "All," and click OK. Next, click OK again to obtain the results.

If more than two curves were compared, a post hoc test should be performed, i.e., compare each curve with every other. In this case, go to – Statistics – Survival Analysis – Compare Two Samples. We provide the data similarly as described above, only adding group codes such as Control, Injury, etc., then click OK. From the available options, select the log-rank test for comparing the curves, and the results will be generated.

Jugment bias test

The data were sourced from the following files: data_file_Experiment 1_judgment_bias_test for Experiment 1, data_file_Experiment 2_judgment_bias_test for Experiment 2, and data_file_Experiment 3_judgment_bias_test for Experiment 3.

From the Statistica 13 interface, the following steps were taken: select the tab – Statistics – Advanced Models – Generalized Linear and Nonlinear Models – Logit Model – Variables – .Dependent Variable (PER responses at the appropriate concentration in the test, e.g., 9:1 or 3:7, depending on which we wanted to test at that moment); for qualitative predictor variables, select the two "colony" and "treatment" (as named in each experiment) – click OK. Next, select the codes for the dependent variable, click "All" and then OK, and do the same for the qualitative predictor codes – click "All" and then OK. Finally, click on "Summary of All Effects" to obtain the statistics.

Nosema spores infection

The data were sourced from the following files: Experiment 2, and data_file_Experiment 2_ nosema_infection

From the Statistica 13 interface, the following steps were taken: select the tab – Statistics – Nonparametric – Multiple Independent Sample Comparison. The dependent variable chosen was “Number of Nosema spores per digestive tract,” and the independent variables were colony and treatment. Then, click OK. Click on "Codes" – select all, then OK. Next, click on the ANOVA Kruskal-Wallis summary and the median test. When the results turn out to be statistically significant, go back to the previous interface and perform multiple comparisons of rank means for all samples, and obtain the results.

Ovariole Number

The data were sourced from the following files: data_file_Experiment 3_ ovariole number.

From the Statistica 13 interface, the following steps were taken: select the tab – Statistics – Advanced Models – Generalized Linear and Nonlinear Models – Log Poisson Model – then click OK. The variables included the dependent variable – number of ovarioles, and the qualitative predictors were colony and treatment. Next, select the factor codes – all, then click OK. Then, click OK again, provide a summary of all effects, and obtain the results.