The browser you are using is not supported by this website. All versions of Internet Explorer are no longer supported, either by us or Microsoft (read more here:

Please use a modern browser to fully experience our website, such as the newest versions of Edge, Chrome, Firefox or Safari etc.

Modelling for the survival of bumblebees

Embroidery of agricultural landscape. Photo.
Embroidery and photo by Maria Blasi

Lack of habitats and climate change – parameters that scientists know affect the number of pollinators in the world. But exactly how big is the effect and what else contributes to the decline of bees and bumblebees? This has Maria Blasi i Romero tried to find out in her dissertation which will be presented on November 26.

A sure sign of spring is the buttery yellow rapeseed fields that every year spread like a blanket over the southern parts of Sweden and in Europe. In addition to being a delight to the eye, the flowering rapeseed fields are also a buffet for pollinators such as bees and bumblebees who for a few weeks can eat as much as they want. But their joy doesn’t last long. The flowering only last a few weeks and if there is no other flowering crop or a flower rich meadow nearby, starvation awaits the pollinators.

The way we produce food, including large monocultures like rapeseed fields, relying heavily in pesticides and reducing the amount of pastures and grasslands, are an important cause of the decline in pollinators. In addition, the pressure of climate change, for example droughts, does not make the situation easier for the pollinators. A statement made by Maria Blasi i Romero, doctoral student at Center for Climate and Environmental Science (CEC), in her dissertation Wild bees in agricultural landscapes: Modeling land use and climate effects across space and time”. In a total of four different studies, she has investigated how land use and climate change is affecting wild pollinators such as bumblebees and solitary bees.

Different classification in different countries

In one of the studies, she studied how the number of bees during the flowering period of the rapeseed fields differed in different places and during different years. She compared data from 2011-2013 from four different locations in Europe: Sweden, Germany, the Netherlands and the United Kingdom. Unsurprisingly, she saw that the more rapeseed fields in an area, the fewer the number of wild pollinators. However, the opposite, semi-natural landscapes with a mixture of fields, forests and pastures, did not always mean more pollinators.

”There was a big difference between the countries, which means that what is good for the bees in Sweden is not necessarily good for the bees in the Netherlands. It is therefore not reasonable to assume that what is good for one population in one region is good for all regions”, says Maria Blasi i Romero.

This is partly due to different species preferring different crops or flowers and that countries have different classification systems of what is a natural habitat.

”What we in Sweden, for example, consider to be a pasture can look very different compared to a pasture in the United Kingdom or in the Netherlands. Therefore there are some limitations when using statistical models to predict the number of pollinators in a location”.

Follows the year of the bumblebee

The use of more detailed models, such as mechanistic models, are complementary to statistical models, and are necessary to understand the relationship between climate, land use and pollinators as well as being able to project future scenarios. Maria Blasi i Romero and her colleagues therefore developed a model based on the bumblebee's life cycle. The year of the bumblebee begins with the queen waking up in the spring, finding a suitable nest, gathering food and laying eggs which develops to become her workers. Later in the season, the queen stops laying working bumblebees and switches to produce males and future queens who will hibernate to the following year. And so it goes on year after year. The model also mimics how the landscape changes over seasonand what happens to the bumblebee population when one or more parameters change.

The difference from other pollinator models is that the new model captures detailed dynamics of what happens in the landscape every week. The resources in the landscape are driven by a simple phenology model that captures the flowering season of the habitats in the landscape.

”We were able to capture realistic patterns of bumblebee cycle in our model, and to test it with simulating drought events. Based on different combinations of simulated landscapes, we could find out which resources that are crucial to maintain a bumblebee community during different climates. We could also simulate how the number of workers and queens produced are affected by how the landscape is composed”.

Bumblebees wake up earlier

In her dissertation, Maria Blasi i Romero has also studied whether hibernated bumblebee queens are active at the same time as in previous years or whether there are any phenological changes. Several previous studies have shown that the queens seem to be active both earlier and later in the spring than usual, which according to the studies is probably due to climate change, although land change can also play a role. To conduct her own studies, Maria Blasi in Romero needed to compare data spanning several years, museum collections and observations from citizens, both current and from previous years.

"It seems that bumblebees in areas with very intensive agriculture and few natural environments, are active much earlier than usual. And early-emegent species are active even earlier. Late emerging ones, however, do not seem to change their activity period and their numbers are decreasing. This is probably because bumblebees that wake up late risk dying of starvation because there are not so many late-flowering crops. Many of them like red clover, which has decreased dramatically in recent years.

However, according to her, much can be solved for the bumblebees if more semi-natural landscapes were developed with a mixture of pasture, forest and flowering meadows. The importance is to ensure continuous floral habitats during all season.

Hopes to inspire model building

Maria Blasi i Romero believes that more systematic and refined data is needed to further investigate and monitor different types of pollinators and to be able to better classify different types of habitats. She advocates for the need to combine both statistical and mechanical models, and encourages other researchers to reflect over the need to transfer not only the knowledge driven by their results, but also how these models can be transferable.

”The theoretical model we developed is simple and based only on a few parameters, but is easy to adapt so that it can be used for many different types of pollinators and crops in the landscape. It offers great opportunities to keep exploring interactions between climate, land use and pollinators, she says.