Event Title
Big Data: Big Possibilities for the Cranberry Industry?
Start Date
30-8-2017 11:45 AM
End Date
30-8-2017 12:00 PM
Description
Abstract:
Cranberry yield and quality vary significantly within a season among production beds, marshes and geographic locations. Additionally, production is inconsistent among years. In anecdotal observation, the variability in cranberry yield is much greater than in many other crops. Why do some beds produce 200 barrels/a while nearby beds or marshes produce over 700 barrels/a in the same growing season and with the same variety? Consistent, high-quality berry production would aid individual growers in terms of long-term planning and the industry relative to crop forecasting and utilization. Previous research efforts have focused on individual parameters, such as fertilizer quantity or herbicide choice, yet it is commonly accepted that production levels are a result of a multitude of factors.
Consistent cranberry production is challenged by several multi-variable issues that are often lumped in the general category of a “stressed” crop. Multi-variable issues require a systems approach with robust data to lead to confidence in the solutions. Additionally, an economic component can be included to help determine which parts of consistent yield are financially reasonable to address and which can’t be solved (such as soil type), thus eliminating spending on inputs that don’t add yield or quality.
Using a “big data” approach, we can determine the relationship among crop production parameters and berry yield and quality. The more data that is included, the more certain we can become about those relationships. We conducted a pilot project with Wisconsin cranberry growers using 2016 production year data (November 1, 2015 through October 31, 2016). Growers were engaged in developing the list of inputs where data collection was anticipated to affect berry yield and quality and in providing pilot data for initial analysis. Sixteen pilot growers entered intensive data from over 500 cranberry beds. Forty-one variables were included that fall broadly in 6 categories: broad production characteristics (such as soil pH), water management, pest management, fertility management and tissue tests, pollinator management, and cultural practices. The relationship among these characteristics and cranberry yield and quality (brix, color, firmness, fruit size, useable fruit and rot) are currently being explored.
In preliminary analysis, drivers of berry yield and quality were identified that were not anticipated but very feasible to modify with reduced grower costs, such as through reducing pre- and post-season irrigation and flood events that result in saturated soil that reduces vine productivity. Other relationships were expected, such as a positive correlation between %N and fruit yield. Our next goals are to expand the initial pilot work, automate an analytical methodology that will be self-sustaining and refine cranberry production such that input levels are optimized for berry yield and quality.
This presentation is not available for downloading.
Recommended Citation
Colquhoun, Jed, "Big Data: Big Possibilities for the Cranberry Industry?" (2017). North American Cranberry Researcher and Extension Workers Conference. 16.
https://scholarworks.umass.edu/nacrew/2017/papers/16
Big Data: Big Possibilities for the Cranberry Industry?
Abstract:
Cranberry yield and quality vary significantly within a season among production beds, marshes and geographic locations. Additionally, production is inconsistent among years. In anecdotal observation, the variability in cranberry yield is much greater than in many other crops. Why do some beds produce 200 barrels/a while nearby beds or marshes produce over 700 barrels/a in the same growing season and with the same variety? Consistent, high-quality berry production would aid individual growers in terms of long-term planning and the industry relative to crop forecasting and utilization. Previous research efforts have focused on individual parameters, such as fertilizer quantity or herbicide choice, yet it is commonly accepted that production levels are a result of a multitude of factors.
Consistent cranberry production is challenged by several multi-variable issues that are often lumped in the general category of a “stressed” crop. Multi-variable issues require a systems approach with robust data to lead to confidence in the solutions. Additionally, an economic component can be included to help determine which parts of consistent yield are financially reasonable to address and which can’t be solved (such as soil type), thus eliminating spending on inputs that don’t add yield or quality.
Using a “big data” approach, we can determine the relationship among crop production parameters and berry yield and quality. The more data that is included, the more certain we can become about those relationships. We conducted a pilot project with Wisconsin cranberry growers using 2016 production year data (November 1, 2015 through October 31, 2016). Growers were engaged in developing the list of inputs where data collection was anticipated to affect berry yield and quality and in providing pilot data for initial analysis. Sixteen pilot growers entered intensive data from over 500 cranberry beds. Forty-one variables were included that fall broadly in 6 categories: broad production characteristics (such as soil pH), water management, pest management, fertility management and tissue tests, pollinator management, and cultural practices. The relationship among these characteristics and cranberry yield and quality (brix, color, firmness, fruit size, useable fruit and rot) are currently being explored.
In preliminary analysis, drivers of berry yield and quality were identified that were not anticipated but very feasible to modify with reduced grower costs, such as through reducing pre- and post-season irrigation and flood events that result in saturated soil that reduces vine productivity. Other relationships were expected, such as a positive correlation between %N and fruit yield. Our next goals are to expand the initial pilot work, automate an analytical methodology that will be self-sustaining and refine cranberry production such that input levels are optimized for berry yield and quality.
This presentation is not available for downloading.