Agricultural intensification impairs behavioral abilities and the expression of genes associated with social responsiveness in honey bees
Data files
Aug 12, 2024 version files 210.25 KB
Abstract
The honey bee Apis mellifera is one of the main pollinators in agroecosystems, and consequently, its colonies are exposed to agricultural intensification. These practices lead to increased use of agrochemicals, resulting in fragmented and more homogeneous habitats. In an agricultural setting located in the region of the Argentine Pampas, behavioral and molecular approaches were performed in honey bee colonies to compare their global state at different times of crop management. Our results show that foraging honey bees captured in these beehives impaired their gustatory sensitivity, cognitive abilities, and the brain’s expression of several genes related to metabolic, immune and neuronal processes associated with social behavior in bees after pesticide application and the flowering of the surrounding crops. To our knowledge, no previous study has reported the impaired effects of agricultural intensification on insect pollinators from an integrative neurobiological perspective under realistic conditions.
README
We have submitted our raw data related to colony activity (ColonyActivity_Data.csv) and collection of pollen (PollenTraps_Data.csv), gustatory responsiveness (GustatoryResp_Data.csv) and learning abilities (LearningScore_Data.csv), and gene expression (GeneExpression_Data.csv).
Descriptions:
ColonyActivity_Data
• Moment: Pesticide application moment
• Year: Sampling year
• Plot: Name of the plot where apiary is located
• Hive: Number of the hives studied
• Total: Total number of bees entering the hive per minute
• Pollen: Number of bees entering the hive with pollen loads per minute
• Nectar: Number of bees entering the hive without pollen loads per minute
PollenTraps_Data
• Moment: Pesticide application moment
• Year: Sampling year
• Plot: Name of the plot where apiary is located
• Hive: Number of the hives studied
• Pollen weight: Weight of pollen collected in the trap (g)
• Crop: Weight of pollen collected from the crop (g)
• Wild flora: Weight of pollen collected from the wild flora (g)
• Time: Trap set duration (min), used as an offset
GustatoryResp_Data
• Moment: Pesticide application moment
• Year: Sampling year
• Plot: Name of the plot where apiary is located
• Bee: Number of the individual forager bee evaluated
• Concentration: Sucrose solution concentration (%)
• Response: individual proboscis extension response (PER) of the bee to the stimulus (response=1; not response=0)
• Total: total number of forager bees evaluated
LearningScore_Data
• Moment: Pesticide application moment
• Year: Sampling year
• Plot: Name of the plot where apiary is located
• Bee: Number of the individual forager bee evaluated
• Score: Sum of PERs throughout the procedure was obtained for each bee
GeneExpression_Data
• Moment: Pesticide application moment
• Year: Sampling year
• Plot: Name of the plot where apiary is located
• Hive: Number of the hives studied
• Syx7: Relative expression of Syntaxin7
• Syx1a: Relative expression of Syntaxin1a
• AchE: Relative expression of Acetylcholinesterase
• Pelle: Relative expression of Pelle
• Brahma: Relative expression of Brahma
• Enolase: Relative expression of Enolase
Methods
Colony activity
We evaluated as a colony activity indicator, the number of incoming bees (incoming rate) at the entrance of the hives at each sampling moment. In each apiary, we randomly chose eight hives. The same eight hives were measured in the two sampling moments. We counted the arrivals in the hives chosen for 1 min, twice a day for three consecutive days, once in the morning and once in the afternoon. According to the presence or absence of pollen loads on their hind legs, incoming bees were recorded as pollen or non-pollen (nectar) bees, respectively. Plot, Moment and year were taken as fixed variables, while hive was taken as a random variable in a Negative binomial model.
Pollen traps
Pollen traps were placed in three randomly selected hives (different from the eight hives used for colony activity) for about 6 h and for three consecutive days. The weight of pollen loads collected in each pollen trap was measured. Color-classified pollen loads of each trap were weighed separately. Then, pollen grains were identified under the microscope to the lowest taxonomic level possible using a palynological database. Once color-classified pollen loads were identified, the total weight and the weight of pollen collected from the crop and the surrounding wild flora were calculated. Plot, Moment and year were taken as fixed variables, while hive was taken as a random variable in a Normal model. To analyze the proportion of pollen collected from the crop and wild flora, a Non-parametrical model was used.
Gustatory responsiveness
To evaluate gustatory responsiveness at the apiary level, we captured incoming honeybees at the hive entrances from different colonies and for three consecutive days. This methodology was carried out in the three apiaries and both sampling moments, before and after crop flowering and pesticide application, and in two years. Each forager bee was exposed to a sucrose solution of a specific concentration chosen in a pseudorandomized way (0.1%, 0.3%, 1%, 3%, 10%, 30%, or 50% w/w). At the end of the experiment, a gustatory response proportion was obtained for each concentration of sucrose solution, based on the number of responsive bees over the total evaluated bees. As we obtained a response ratio for each sucrose solution concentration, the statistics analyses were independent among the different solution concentrations tested. Plot, Moment and year were taken as fixed variables in a Binomial model.
Learning score
To evaluate the association between a pure odor with a sucrose reward via classical olfactory conditioning by using the proboscis extension response procedure, we captured incoming honeybees at the hive entrances from different colonies and for three consecutive days. This methodology was carried out in the three apiaries and both sampling moments, before and after crop flowering and pesticide application, and in two years. At the end of the experiment, a score defined as the sum of PERs throughout the procedure was obtained for each bee (learning score). Plot, Moment and year were taken as fixed variables in a Poisson model. As we obtained a score of responses for each forager bee evaluated, individual bees were not taken as a random variable.
Gene expression
The expression of genes associated with social responsiveness in honeybees was measured to assess the impact of pesticide application and the changes in the food resources’ availability at the molecular level. In each apiary, we randomly chose four hives, and 10 brains of forager bees were pooled. Thus, we utilized four biological replicates per apiary. Plot, Moment and year were taken as fixed variables, while hive was taken as a random variable in a Gamma model.
Methods
Methods
Colony activity
We evaluated as a colony activity indicator, the number of incoming bees (incoming rate) at the entrance of the hives at each sampling moment. In each apiary, we randomly chose eight hives. The same eight hives were measured in the two sampling moments. We counted the arrivals in the hives chosen for 1 min, twice a day for three consecutive days, once in the morning and once in the afternoon. According to the presence or absence of pollen loads on their hind legs, incoming bees were recorded as pollen or non-pollen (nectar) bees, respectively. Plot, Moment and year were taken as fixed variables, while hive was taken as a random variable in a Negative binomial model.
Pollen traps
Pollen traps were placed in three randomly selected hives (different from the eight hives used for colony activity) for about 6 h and for three consecutive days. The weight of pollen loads collected in each pollen trap was measured. Color-classified pollen loads of each trap were weighed separately. Then, pollen grains were identified under the microscope to the lowest taxonomic level possible using a palynological database. Once color-classified pollen loads were identified, the total weight and the weight of pollen collected from the crop and the surrounding wild flora were calculated. Plot, Moment and year were taken as fixed variables, while hive was taken as a random variable in a Normal model. To analyze the proportion of pollen collected from the crop and wild flora, a Non-parametrical model was used.
Gustatory responsiveness
To evaluate gustatory responsiveness at the apiary level, we captured incoming honeybees at the hive entrances from different colonies and for three consecutive days. This methodology was carried out in the three apiaries and both sampling moments, before and after crop flowering and pesticide application, and in two years. Each forager bee was exposed to a sucrose solution of a specific concentration chosen in a pseudorandomized way (0.1%, 0.3%, 1%, 3%, 10%, 30%, or 50% w/w). At the end of the experiment, a gustatory response proportion was obtained for each concentration of sucrose solution, based on the number of responsive bees over the total evaluated bees. As we obtained a response ratio for each sucrose solution concentration, the statistics analyses were independent among the different solution concentrations tested. Plot, Moment and year were taken as fixed variables in a Binomial model.
Learning score
To evaluate the association between a pure odor with a sucrose reward via classical olfactory conditioning by using the proboscis extension response procedure, we captured incoming honeybees at the hive entrances from different colonies and for three consecutive days. This methodology was carried out in the three apiaries and both sampling moments, before and after crop flowering and pesticide application, and in two years. At the end of the experiment, a score defined as the sum of PERs throughout the procedure was obtained for each bee (learning score). Plot, Moment and year were taken as fixed variables in a Poisson model. As we obtained a score of responses for each forager bee evaluated, individual bees were not taken as a random variable.
Gene expression
The expression of genes associated with social responsiveness in honeybees was measured to assess the impact of pesticide application and the changes in the food resources’ availability at the molecular level. In each apiary, we randomly chose four hives, and 10 brains of forager bees were pooled. Thus, we utilized four biological replicates per apiary. Plot, Moment and year were taken as fixed variables, while hive was taken as a random variable in a Gamma model.