One of the new trends in agriculture is the idea of "precision farming" or "site-specific management" of crop production fields. The goal of precision farming is to identify sources of variability within a field and then manage that variability. Two types of variability are most prevalent in a field -- spatial and temporal. Spatial or field variability is the difference in crop performance on numerous sites within the field. Temporal or season variability is the difference in crop performance on one same site within a field. Agronomists have long recognized the problems associated with field and season variability and deal with these sources of variability through statistics.

Many technologies are converging which allow growers to manage spatial variability. The question is largely one of economics. The techniques and technology required to identify field variation include soil surveys, yield monitors, positioning systems, soil sampling, sensors, computers and equipment with the capability of applying variable rates of agricultural inputs. Examples of management inputs that "even up" field variability include drainage and variable fertilizer, seeding, and pesticide rates.

Season variation is much more difficult to assess and manage because it involves predicting the weather and extrapolating these effects on crop production. Other interactions, such as diseases, insects, and other pests, can influence season variation. One example of a management factor that is often used to "smooth out" season variability is irrigation.

Variability is measured by evaluating discrete sites in a field over time. Yield monitors along with global positioning systems do a very good job of measuring field variability. Yield maps help assess the amount of variability and help to define a management area from which various input levels can be recommended. Field yield maps over many growing seasons will estimate season variability.How much variability can be expected at one site within a field over a number of seasons? Below is the data for four sites in a six-acre field over the last ten seasons. The soil is a Plano silt loam and the field is located near Arlington, WI. Each site measured 30 feet by 35 feet and had been cropped continuously to corn since 1983. The corners of each site were located in the exact same place every season.

Table 1. Corn yields over time at four sites in a field near Arlington, WI.
Site	1986	1987	1988	1989	1990	1991	1992	1993	1994	1995	Season range *	Season deviation
1	165	190	89	148	151	144	144	90	142	125	101	31
2	180	161	63	141	158	135	156	81	165	113	117	38
3	169	157	55	148	143	132	138	86	167	126	114	36
4	176	159	44	154	152	133	151	71	174	119	132	44

Field range	15	33	45	13	15	12	18	19	32	13
Field deviation	7	16	19	5	6	5	8	8	14	6
* Range = Maximum - minimum. Deviation = Standard deviation.

During 1986, the average yield of all four sites in a field was 173 bushels per acre (Table 1). The range between the highest and lowest yielding sites was 15 bushels per acre while the field standard deviation was 7 bushels per acre. Every site produced the greatest yield within a season, at least one time during the 10-season period. Site 1 produced an average yield of 139 bushels per acre over the ten season period. On site 1, the range between the highest and lowest yielding season was 101 bushels per acre, while the season standard deviation was 31 bushels per acre.

The field range between the maximum and minimum yielding sites within a year averaged 23 bushels per acre (Table 2). Field yield variability as measured by standard deviation averaged 9 bushels per acre over the ten-season period. The season range between the maximum and minimum yielding seasons for a site averaged 116 bushels per acre. Season yield variability as measured by standard deviation averaged 37 bushels per acre over the four sites.

Table 2. Average field and season (1986-1995) variability of corn at four sites in a field near Arlington, WI.
Variability source	Average	Range average	Deviation average
Yield (bushels per acre)
Field	135	23	9
Season	135	116	37
Grower return when corn price at $2.00 per bushel (dollars per acre)
Field	270	43	19
Season	270	232	75
Grower return when corn price at $2.75 per bushel (dollars per acre)
Field	371	59	26
Season	371	319	103
Grower return when corn price at $3.50 per bushel (dollars per acre)
Field	472	75	33
Season	472	406	131

Yield monitors will help producers evaluate "precision farming" technology and assist in identifying field management zones that need to be modified in order to "smooth out" field variability. They also have the potential to assist the grower in making better management decisions. In this example, if all field variability could be removed by utilizing precision farming techniques and bringing yield up to the level of the maximum site, there would be an average maximum gain of 23 bushels per acre. However, season variation was four times that of field variation (37 v. 9 bushels per acre). An average seasonal range of 116 bushels per acre was observed for the same site in a field. This relationship between field and season variation held regardless of corn price. Depending upon corn price, grower return could increase between 43 and 75 dollars per acre by eliminating site differences within a field. Grower return due to season differences ranged from 232 to 406 dollars per acre depending upon corn price. Season variation is the overriding source of variation for any site in a field. It is also the most difficult variation to manage in a production field.