Statistical Modelling for Precision Agriculture: A Case Study in Optimal Environmental Schedules for Agaricus Bisporus Production via Variable Domain Functional Regression

Author: mycolabadmin
2017-09-29
View Source

Summary

This research developed a statistical method to help mushroom farmers optimize their growing conditions using sensor data. The study analyzed how factors like temperature, humidity, and oxygen levels affect mushroom yields throughout the growing process. The findings provide practical guidance for farmers to adjust their growing environments to maximize production. Impacts on everyday life: – More efficient food production through optimized growing conditions – Potential reduction in agricultural resource waste and costs – Improved mushroom availability and quality for consumers – Application of similar methods could improve yields of other indoor-grown crops – Demonstrates how data analysis can improve traditional farming practices

Background

Modern agricultural production processes are evolving with sensor network technologies that improve monitoring of farming environments. While there is growing sensor network data available, there is a lack of statistical techniques accessible by the farming industry to combine this data with understanding of the farming process. Commercial mushroom cultivation has experienced explosive growth, increasing from 60,000 tons in 1978 to 25.7 million tons in 2011 to become a $24 billion industry. While research exists on growth substrate preparation, there has been limited research on identifying optimal environmental conditions for the growing process.

Objective

To develop and apply a variable-domain functional regression (VDFR) technique to quantify the effects of environmental factors like temperature, humidity, oxygen and CO2 levels on Agaricus Bisporus (button mushroom) yields across growing processes of different durations. The goal was to determine optimal environmental schedules that would maximize mushroom production yields.

Results

The VDFR model revealed clear effects of oxygen levels on total yield – lower oxygen was beneficial in the first two-thirds of growing while higher oxygen increased yields in the final third. For temperature, higher levels were beneficial in the first half of the process, while in the second half these were only favorable for longer duration grows. The model with CO2 production, Oxygen and Humidity deficit as functional covariates showed the highest explanatory power. When CO2 and Oxygen were both included, only one was found significant due to their physical displacement relationship.

Conclusion

The VDFR technique successfully enabled understanding of how different environmental variables affect mushroom yields across varying growth period lengths. The model provided meaningful interpretations of optimal growing conditions that can be used by farmers to improve yields. The approach shows promise for modeling yields of other crops grown in controlled conditions. Implementation of the framework by the industry partner has improved their crop performance and yield.

Published in:PLOS One,
Study Type:Observational Study,
Source: 10.1371/journal.pone.0181921