Farming a Flat Function AAE 320
Overview of Lecture “Farming a Flat Function” What is it? (Give Examples) What does it mean? (Implications) In my opinion, this problem underlies a lot of complaints about farmers and the environment It’s not a simple problem to fix
Farming a Flat Function For many crop production processes, yield becomes relatively unresponsive to inputs when they are used at near optimal levels Yield Input
Farming a Flat Function As a result, profit also has a “flat” response to the input Profit Input
Mitchell (2004) Assembled data from experiments examining corn response to nitrogen Most from late 1980’s and early 1990’s Seven states (IA, IL, IN, MN, NE, PN, WI) Almost 6,000 individual observations Analyzed to see if could statistically observe effect of nitrogen on yield when used at high/near optimal nitrogen rates
One Site-Year from Iowa
All Site Years from Iowa 2,200 observations
Average Yield by N Rate
Main Point Once N rates get above 85-100 lbs/ac, expected (average) corn yield very flat, but lots of variability around this average Makes identifying yield effects of nitrogen on corn statistically difficult/impossible Change in yield with changing N rate hard to see with all the noise from other factors
Online N Tool from Problem Set 1
Current WI Recommendations Source: C. Laboski, UW Soil Science
Main Point WI nitrogen recommendations for corn give the range of N rates that are within $1/ac of the maximum return Notice how wide the range of N rates is Over the range of application rates the recommendations give, expected net returns vary less than $1/ac Returns from applying nitrogen to corn are very flat when near optimal levels
What about other inputs? 1) Economic analysis of Seeding Rates 2) Economic analysis of processing and fresh market sweet corn and the value of insecticide sprays for controlling European corn borer (ECB) Mitchell et al. (2005)
Corn Returns for Seeding Density (Lauer and Stanger 2006) Source: http://corn.agronomy.wisc.edu/AA/pdfs/A044.pdf
Soybean Returns for Seeding Density and Seed Treatments (Gaspar et al http://www.coolbean.info/library/documents/SoybeanTreatmentRisk_2014_FINAL.pdf
Processing Sweet Corn Insecticides
Capture (bifenthrin) on Processing Sweet Corn (mean with 95% error bars)
Capture (bifenthrin) on Fresh Market Sweet Corn
Capture (bifenthrin) on Fresh Market Sweet Corn (mean with 95% error bars)
Main Point Farmers are “Farming a Flat Function” for several of main inputs Implications 1) Small profit change over wide input range 2) Impact of inputs on returns is hard to notice with all the variability from other factors 3) A wide range of input levels will be and will seem consistent with maximizing profit
And Another Issue Under use of inputs is often obvious See weeds, insects, diseased plants, yellow or purple crop, wilted crops Over use of inputs is often invisible How many farmers leave an untreated check strip? When Farming on a Flat Function, How do you know if you put on too much Fertilizer? Fungicide? Insecticide? Water? Over use of inputs often an invisible cost Can’t! Costly!
Main Point Implications of Farming a Flat Function when under use of inputs is obvious and over use is invisible Wide range of input levels consistent with profit maximization (and risk management) Some inputs turn out to not be used/needed afterwards, but cannot tell when put them on Some of the “extra” inputs end up as pollution Farmers accused of using too many inputs
In my opinion This problem underlies a lot of complaints about farmers and the environment It’s not a simple problem to explain to the public: Why do farmer use so much fertilizer, pesticide, water, etc.? Would farmers waste money on purpose? Most farmers make an honest effort to be good stewards But farming is messy and variable and it is expensive to find out where you are or will be on some flat function What’s the answer? More/better technology? Try to better match “natural” systems? Both? Something else?