Nicolas Tremblay
Research Scientist
Agriculture and Agri Food
Canada

Biography
Dr. Nicolas Tremblay is President-Elect of the International Society for Precision Agriculture (ISPA). He leads an important research program for the Government of Canada and is known for his ability to generate new knowledge for the benefit of the agricultural sector. Dr. Tremblay is the author or co-author of more than a hundred scientific peer-reviewed publications. He is currently involved in the management of nitrogen fertilizer applications using remote sensing, geomatics, geostatistics and meta-analyses. He has conducted research on fluorescence techniques for the detection of stresses affecting crops. Dr. Tremblay is adjunct-professor at Laval University, the University of Ottawa and the Université de Montréal. He is President of the Commission chimie et fertilité des sols, the official body for fertilizer recommendations in Quebec. He is a member of the International Society for Horticultural Science (ISHS), the American Society of Agronomy (ASA), the Crop Science Society of America (CSSA), the Soil Science Society of America (SSSA), and the Association québécoise des spécialistes en sciences du sol (AQSSS).

Abstract
Geospatial Information: A Bridge for the Transfer of Past and Future Science to the Farmers

Although geospatial information is the essence of agricultural crop production, it has not been duly integrated into research efforts. An overwhelming number of technological opportunities are now available to the agricultural sector, many of which are fully implemented and generate improvements in the use of resources in a sustainable intensification context. Nonetheless, there is still much room for improvement in cases where reliable decision rules must be applied in the context of a farmer’s specific soil-crop-atmosphere system. The problem is the historical divergence of focus of scientists and farmers in handling different types of uncertainties. Scientists have been trained to control primarily metric uncertainties, whereas farmers are required to deal more with structural, temporal and translational uncertainties. Indeed, as summarized by Cook et al (2013), in the early 20th century, Fisher and co-workers at Rothamsted developed methods of statistical analysis to clarify the experimental treatment effects within the field. This had a huge impact on agricultural research, one that continues to this day. Yet, as with all scientific methodologies, there is a risk that the method starts to define the problem. Scientists can place more value on the information about the quantifiable effects of factors and discount information about the farming environment in which they occur. Under this scenario, the effects of factors are known, while the interaction with the farmed landscape (and the changing seasonal factors) less so. That is why, still today, there is a gap between the bulk of agricultural research available and the actual acceptance of their conclusions by farmers. Fortunately, the opportunity exists to bridge this gap by bringing back geospatial information that was disregarded at the time of experimentation. In agriculture, pertinent geospatial information generally translates into soil and weather features. When trials are performed on experimental farms or a limited number of commercial sites, their soil or seasonal characteristics convey very little information, hence the difficulty of appreciating their influence on the results obtained. Thanks to meta-analyses, it is possible to assemble the results or data of past studies, and to couple them with meta-data related to soils and seasonal characteristics. When a sufficient number of them are put together, meta-analyses allow for a fresh look at results of studies with apparently conflicting conclusions. Very often, the apparent contradictions are explained by the geospatial context (soil/weather) in which the studies were conducted. Conclusions can be drawn for each geospatial context and therefore adapted to the farmer’s reality, resulting in increased adoption. An example of this process can be taken from the history behind the SCAN (Soil, Crop, Atmosphere for N [Nitrogen]) decision-support system for in-season fertilization in corn which is currently being transferred to users by Agriculture and Agri-Food Canada. In the future, the spread of precision farming, UAV technology and satellite imagery will provide opportunities for (big) data collection. With this will come the need for proper data stewardship and data and meta-data repositories that will allow for further data mining initiatives. Geospatial information in a precision agriculture context is key to greater coordination among scientists, policy-makers, technology providers and users that will benefit the agricultural industry as a whole by facilitating more evidence-based decision making among stakeholders.That is why, still today, there is a gap between the bulk of agricultural research available and the actual acceptance of their conclusions by farmers. Fortunately, the opportunity exists to bridge this gap by bringing back geospatial information that was disregarded at the time of experimentation. In agriculture, pertinent geospatial information generally translates into soil and weather features. When trials are performed on experimental farms or a limited number of commercial sites, their soil or seasonal characteristics convey very little information, hence the difficulty of appreciating their influence