agridat (version 1.8.1)

nass.corn: U.S. historical crop yields by state

Description

Yields and acres harvested in each state for the major agricultural crops, from approximately 1900 to 2011. Crops include: barley, corn, cotton, hay, rice, sorghum, soybeans, wheat.

source

United States Department of Agriculture, National Agricultural Statistics Service. http://quickstats.nass.usda.gov/

Details

Be cautious with yield values for states with small acres harvested. Yields are in bushels/acre, except: cotton pounds/acre, hay tons/acre, rice pounds/acre.

Examples

dat <- nass.corn

# Use only states that grew at least 100K acres of corn in 2011
keep <- droplevels(subset(dat, year == 2011 & acres > 100000))$state
dat <- droplevels(subset(dat, is.element(state, keep)))

# Acres of corn grown each year
xyplot(acres ~ year|state, dat, type='l', as.table=TRUE,
title="State trends in corn acreage")

# Plain levelplot, using only states
datm <- acast(dat, year~state, value.var='yield')
levelplot(datm, aspect=.7, col.regions=RedGrayBlue,
          main="nass.corn",
          scales=list(x=list(rot=90, cex=.7)))

library("mvtsplot")
mvtsplot(datm, levels=11, margin=FALSE, norm='global')

# Model the rate of genetic gain in Illinois as a piecewise regression
# Breakpoints define periods of open-pollinated varieties, double-cross,
# single-cross, and transgenic hybrids.

dil <- subset(nass.corn, state=="Illinois" & year >= 1900)
m1 <- lm(yield ~ pmin(year,1932) + pmax(1932, pmin(year, 1959)) +
         pmax(1959, pmin(year, 1995)) + pmax(1995, year), dil)
signif(coef(m1)[-1],3) # Rate of gain for each segment
plot(yield ~ year, dil, main="Piecewise linear model of Illinois corn yields")
lines(dil$year, fitted(m1))
abline(v=c(1932,1959,1995), col="wheat")