The site
The study area will be at one or two typical norwegian farms in Vestfold (Nøtterøy and Horten) where they produce raspberries (map to the right). Honeybees are set out in the area, close to the raspberry fields. The process of collecting data will start some time in May 2015, approximately one week before flowering of raspberry and end some time in late June or early July. Raspberry usually start flowering around June 1st in this area.
My hypothesis
The way insects react to temperature variation varies a lot. Different species might have very different tolerance limit with respect to temperature and it reason to believe that honeybees (originally a southern species) and bumblebees (a more robust northern species) show contrasting behavioural responses to temperature. How will climate change and the warming of the climate affect these two species? By studying honeybees and bumblebees, it might be possible to say something about the effect of increasing temperatures will have on these species and consequently the ecosystem services they provide. The question I try to answer in this master thesis is: “Will bumblebees struggle with the global temperature increase, and will honeybees benefit from this because they are adapted to a warmer climate?”
Data collecting
To investigate this fascinating question, I will collect insect activity data in the form of flower visitation. This will primarily be done for honeybees and bumblebees on raspberry, but also some other plants before and during the flowering of raspberry. All this will be done in field in Vestfold. Data on temperature, precipitation/humidity, wind and other environmental covariates will also be collected. This together with time of day and date will be a strong foundation for the dataset. It will be important to have several counts every day and during the whole flowering season of raspberry. In addition, pan traps (dishes) may be used to more objectively collect insects during the day, and can be used to compare with the counts of flower visits as a more unbiased estimate of population densities. It might also be possible to do a small temperature experiment on bumblebees in lab, but this is still in the early planning stage.
Statistics
The collected data will be compiled in a matrix to create the basis for further statistical analysis. My goal is to create a graphical representation of pollinator activity for the different variables, such as temperature, time of day, date and so on. The results might look like a bell-curve, with maximum activity at intermediate temperatures. This can then be interpreted and it might be possible to say something about the optimal temperatures for these species and how an increase in temperature will affect them. What statistical models will be used for this is still being discussed, but Generalized Linear Mixed Models (GLMM) looks like a good option. We will assume a Poisson distribution, and Bayesian inference will be considered if the models become too complex.
Results
This section will be updated in early august.