Studies by Dr Caroline Brophy, Department of Mathematics & Statistics in Maynooth University, can help farmers grow more grass and thereby feed more animals.    

    More than half of Ireland is covered in grassland. Economically this “crop” is important for our beef and dairy cattle farmers.  Brophy’s research results suggest getting the seed recipe right has positive impacts on yields and could lessen the need for added fertiliser.

    The main aim of a large study that Dr Brophy worked on was straightforward: whether planting two grasses and two legumes (e.g. clover), one fast and one slow to establish, was better than just one grass.   
     

    But to get a definitive answer to this question is not so simple.  It meant trying out different mixtures in many different places.  Across Europe, 31 sites took part in the study, which ran for three years. The reams of data collected meant that statisticians were needed to wade in and help understand it all, a big data project which Dr Brophy, lecturer in the Department of Mathematics and Statistics, relished.

    “We found strong diversity effects, with better yields from mixtures,” says Dr Brophy, who was part of a large group of statisticians and biologists working on the study. This “mixing effect” was above and beyond what you might expect from growing any plant on its own.

    In many fields, farmers plant one variety of grass; but this research shows monoculture may not be the best option.

    Bonus yield
    Dr Brophy was tasked with finding out why the bonus yield occurs by using her statistical training to tackle the big data pile. “We can measure the way two species interact together and use that information to consider how best we could maximise outputs,” Dr Brophy explains. Monocultures were planted, but also various proportions of the four-species mixtures, which meant lots of measurements.

    “What we saw from the study was that mixing gave a strong diversity effect,” says Dr Brophy, and so higher yields. Why four types of plants gives more feed for animals can be partly explained by the extra nitrogen fixed by the legumes, she says. This can replace expensive mineral nitrogen. 

    Variation in plant strategies helps too. “Think of putting two types of species together. If, for example, one is deeper rooted than the other, it is able to go further down for nutrients and water, so you utilise the system better,” says Dr Brophy. Her stats acumen allowed her to crunch the data and reveal how different species interacted.

    “Diverse systems are also better at resisting weeds. This is because on its own one species might be susceptible to certain conditions. For example, a tall weed might invade and take over, but diversity helps protects against that,” Brophy explains. Better insight here will allow for fields with more growth.

    The study also needed to look at how long the positives would last. Legumes wane, and the slow growing grass becomes more dominant, but yield benefits remained even into the third year.  “The benefit of mixing persists and that is really powerful and provides additional motivation for farmer’s to consider mixture grassland swards,” Dr Brophy explains.

    Resilience against setbacks
    Diversity also can make for a farm that is more resilient to change or unexpected knocks. The next step for the Maynooth statistician is to look at how diversity might protect against climate extremes, which is a growing concern. Climate change suggests Ireland might have water shortages, more intense rainfall events or colder winters, and farmers need to think about coping strategies.

    “If you end up with a summer without much rain or a winter that is extremely cold, how do you cope with that? Can diversity help protect against climate extremes?” Dr Brophy questions.  

    These critical questions have both statistical sides and biological sides. “I’m learning all the time,” says Dr Brophy, who had not studied biology for her science degree. “There has been so much more about biodiversity in recent years,” she says. “And there are lots of statistical challenges in this area. I can combine my statistical expertise with an area I feel passionate about.”

    She reads lots of ecology literature and is an associate editor for the Journal of Ecology. But equally, she has converted some plant scientists to her love of statistics. The learning for statisticians and biologists is a two-way street.  She recalls a training workshop when the project scientists from across Europe came to hear how they could apply statistics to the data they had collected. 

    Getting outside
    Ask someone would they like to learn a statistical model, and you might get a ho-hum response, but ask them if they would like to learn how to better understand their own data, well that changes things, Dr Brophy says.

    She travelled to Sweden, Norway, Austria, Iceland and Italy during her work on the 31-site European project. Though not from a farming background, she was keen to don wellies and get out into the field to see where the data was coming from. This gives an advantage when it comes to looking at numbers on the spreadsheet.

    “If you don’t understand the process by which the data were generated, it can lead you up the wrong path in terms of how you analyse,” she explains. “And it helps if you understand the practicalities involved.”

    For this study, legume plants had to be separated from grasses, and large fields might be cut five times in one year.  So subsamples and grab samples are used and then researchers extrapolate up. You could not separate all species every time the field is cut. It's a simple example, but the message, says Dr Brophy, is it’s always helpful to get out and see the experiment in action.

    She recently embarked on a new project looking at the reliability of Met Éireann weather forecasts and how to improve predictions about grass growth for farmers. The idea would be to predict, using weather forecasts, what kind of grass growth a farmer might expect in a coming week. This might be accessed via an online tool or app, but whirring underneath it would be statistical insights from Maynooth.