TITLE:
Testing of Precision Agriculture Technologies in Irrigated Cotton at AG-CARES, Lamesa, Texas, 2000.
AUTHORS:
K.F. Bronson, J.W. Keeling, T. Wheeler, R.J. Lascano, R. K. Boman, E. Segarra, J.
Booker, J.D. Booker, J. Mabry, Assistant Professor, Professor, Associate Professor, Professor, Extension Specialist, Professor, Research Assistants and Technician.
METHODS AND PROCEDURES:
Experimental Design: Randomized complete block with 3 replications
Plot size: 53 ft wide (16 40-inch rows) and > 500 ft long.
Experimental area: 27 ac
Soil type: Amarillo sandy loam to sandy clay loam
Variety: Paymaster Roundup® Ready 2326
Soil sampling: Half-ac grid (Fig.1 and 2)
P fertilizer rate: Blanket-rate of 30 lb P2O5/ac,
Average Variable-rate of 38 lb P2O5/ac
Planting date: May 10, 2000
Harvest date: October 4, 2000
Irrigation: LEPA on a 3.5 day schedule at 75% estimated cotton ET
replacement
Cotton responded to P fertilizer in all three landscape positions of the precision agriculture site at AGCARES (Table 1 and 2). Historically, the greatest yields have been observed in the bottomslope where re-distribution of water and nutrients occurs. Variable-rate (VRT) and blanket-rate resulted in lint yields greater than zero-P in both sideslopes. In the bottomslope, only the variable-rate treatment affected lint yields. The south-facing sideslope had the lowest lint yields in 2000. This may be due to the greater amount of blowing observed there and to faster soil water evaporation. The Micro-Trak® yield data was less variable than the hand-picked lint data and only the machine data showed the P fertilizer response (Table 1 and 2).
Temik® was applied at planting at a rate of 5 lb/ac to the entire 27-ac area. Greater nematode numbers were observed in the bottomslope, and for this reason, the VRT strips of this area received an additional 5 lb/ac sidedress of Temik. However, yields in this area were not affected by Temik®.
Preliminary cost and returns economic analysis of the VRT technology is shown in Table 3. This analysis does not consider the greater cost of grid-soil sampling or of VRT equipment. The
average P fertilizer rate applied in the VRT plots was 38 lb P2O5/ac, compared to 30 lb P2O5/ac in the blanket-rate plots. Although statistically there was no difference between the VRT-P and blanket-P treatments we did this analysis by calculating a “return to P fertilizer” for each. This preliminary analysis indicates that up to $24/ac return of VRT-P is possible. Extra or variable-rate Temik was not economical.
Table 1. Micro-Trak® cotton lint yields (lb/ac) for variable-rate, blanket-rate and zero-rate P
fertilizer application, Lamesa, TX, 2000.
| Treatment | North-facing sideslope | Bottom-slope | South-facing sideslope | Mean |
| Variable-rate P fertilizer | 536 a1 | 590 a1 | 485 a1 | 537 a1 |
| Blanket-rate P fertilizer | 540 a | 544 b | 479 a | 521 a |
| Zero P fertilizer | 493 b | 521 b | 434 b | 483 b |
| Mean | 523 a2 | 552 a | 466 b |
1 Means in a column followed by similar letter are not different by pairwise comparisons, p>0.05
Table 2. Hand-picked cotton lint yields (lb/ac) for variable-rate, blanket-rate and zero-rate P fertilizer application, Lamesa, TX, 2000.
| Treatment | North-facing sideslope | Bottom-slope | South-facing sideslope | Mean |
| Variable-rate P fertilizer | 679 a1 | 759 a1 | 570 a1 | 670 a1 |
| Blanket-rate P fertilizer | 634 a | 673 a | 564 a | 623 a |
| Zero P fertilizer | 596 a | 665 a | 523 a | 594 a |
| Mean | 636 a2 | 699 a | 552 b |
1 Means in a column followed by similar letter are not different by pairwise comparisons, p>0.05
Table 3. Input application rates and cost and returns of input applications, Lamesa, TX, 2000
a Assumes no gain in yield
b Does not consider capital costs of variable-rate application equipment or the greater cost of 0.5-ac grid soil sampling and laboratory analysis for the VRT treatments.
Fig. 1. Half-acre grid soil sampling locations and Mehlich-3 P, AGCARES, Lamesa, TX 2000
Fig. 2. Variable-rate inputs experimental layout and P fertilizer rates applied (V = VRT,
