The Demonstration project was conducted on the Demonstration Farm at the Maricopa Agricultural Center. In this project all current guidelines and recommendations disseminated by the University of Arizona were integrated in a systems approach. The management decisions were made by the Extension Specialists in agronomy, entomology, irrigation management, weed sciences, and plant pathology following the University recommendations. On a 50.5 acre field 80% Bt and 20% non-Bt cotton was planted dry and watered up. Due to the cold spring and sand-blasting, only a stand of 30,900 plants/A could be established with 84% terminal damage. 72 acreinches of water were used with 41.3 acre-inches in postplant irrigations. Weed control could be achieved with one preplant application and three cultivations. Three sprays against Lygus and one spray against whiteflies were necessary after the thresholds were exceeded. A total of 4120 lb seedcotton per acre were harvested, with 32.7% lint turnout (2.81 bales/A) and 45.9% seed turnout (1891 lb/A). After harvesting a field budget was established. The variable costs per acre were $915, the total cost $1266/acre. In spite of the lack of replications this project validates the usefulness and compatibility of University recommendations and the potential for integration of all disciplinary guidelines in one system.

Upland cotton advanced strains and commercial check comparison varieties were evaluated in replicated field studies at three locations in 1998. The test sites include Buckeye, Az., Maricopa, Az., and Safford, AZ. Twelve seed companies submitted a maximum of five advanced strains entries per location. Three commercial check varieties were used at each site for comparison purposes and included DPL 5415, SG 125, and STV 474.

Nine defoliation treatments were applied to Pima and upland cotton to compare the treatment effects on percent leaf drop, trash sent to the gin, lint yields, percent lint turnout and percent first pick. All of the treatments were beneficial to leaf drop compared to the untreated check with the Ginstar treatments generally performing better than the others. The addition of nonionic surfactants and drift retardants seemed to reduce the activity of Ginstar. Yield differences on long staple treatments were notices and discussed in the paper.

A single field experiment was conducted in 1998 at Marana, AZ to evaluate a scheduled (based upon stage of growth) versus a feedback approach (based upon growth parameters and crop conditions) to nitrogen (N) and mepiquat chloride (PixTM) applications on upland cotton (Gossypium hirsutum L.). The parameters used in evaluating feedback applications for both N and Pix included fruit retention (FR) levels and height to node ratios (HNRs) with respect to established baselines for cotton grown in the desert Southwest. Scheduled and feedback Pix applications were made for a total of 1.5 and 2.5 pint Pix/acre, respectively, with the feedback treatments receiving a late season application at approximately 3100 heat units after planting (HUAP 86/550 F threshold). Scheduled Pix treatments received a single 1.5 pint Pix/acre application prior to peak bloom (approximately 2000 HUAP). Scheduled applications of fertilizer N totaled 205 lbs. N/acre from three applications. Feedback applications of N received a total of 100 lbs. N/acre from two applications. Treatments consisted of all combinations of feedback and scheduled applications of both N and Pix. The highest lint yields occurred in the treatment consisting of Pix feedback and N feedback (treatment two), however, there were no significant differences (P≥0.05) among any of the treatments with respect to yield.

A series of boll measurements were taken at numerous locations in cotton producing areas across Arizona in 1998 in an attempt to continue to develop a yield prediction model with a project that began in 1993. Results from 1995 showed the strongest relationship between final open boll counts and yield compared to a number of other measurements. Based on these results, data collection on boll counts began in 1996 and has continued in 1997 and 1998. Boll counts were taken as the number of harvestable bolls meter-1. All boll count measurements were made within one week of harvest. Number of bolls per unit area were then correlated to lint yield and an estimate for the number of bolls per area needed to produce a bale of lint was calculated. Estimates using all three years data combined indicate that approximately 38 bolls meter-1 are needed to produce one bale of lint per acre.

Ten field experiments were conducted in major cotton growing areas of Arizona in 1998 for the purpose of evaluating Upland cotton varieties in terms of adaptability and performance. Eight commercial cottonseed companies participated in the program. A maximum of two varieties were submitted by each company at each location. Experiments were conducted on a commercial level on grower-cooperator fields in most cases. Locations used in the program spanned the range of conditions common to cotton producing areas of the state from about 100 ft. to 4,000 ft. elevation. Each of the participating seed companies offer a compliment of varieties that can serve to match various production strategies commonly employed in the state. The 1998 cotton season was a very difficult one for many cotton producing areas in AZ below ~2,000 ft. elevation, characterized by a cool wet spring, late planting, a delayed crop, and a strong monsoon season that reduced fruit retention in many cases. Many varieties commercially available performed well at several locations demonstrating good adaptation to Arizona conditions.

Thirty six upland cotton varieties were grown at the Maricopa Agricultural Center as part of the national cotton variety testing program. Lint yield, boll size, lint percent, plant populations, plant heights and fiber properties are presented in this paper.

Forty eight short staple varieties were grown in a replicated field trial on the Safford Agricultural Center. Excellent yields were recorded, especially considering the late start caused by unseasonably cold weather in April. FM 989, the Australian variety formerly known as IF 1003, produced the highest lint yield of 1601 pounds per acre. Three other varieties, FM 975, AP 4103 and IF 1002, produced over 1500 pound of lint per acre. Agronomic values for the plants at harvest and HVI data for lint quality are tabulated in this paper.

Two replicated on-farm short staple variety trials were planted in 1998. Fifteen varieties were evaluated on both the Carpenter farm in Thatcher and the Colvin farm near Ft. Thomas. Several new varieties were planted in both studies, including 4 transgenic varieties: DP 90B, BXN 47, DP 90RR and Paymaster 1560BG, 2 varieties from Australia: FiberMax 989 and FiberMax 832, and seven other varieties seen for the first time. Two of the new varieties produced the highest yields; AgriPro 6101 and Phytogen 952 on the Carpenter and Colvin farms, respectively. Other agronomic data from the varieties and HVI values from the lint are also included in this report.

The next generation of Bollgard® cotton was evaluated for agronomic and insecticidal efficacy under central Arizona growing conditions. Two novel lines were compared with their recurrent parents, DP50 and DP50B. There were no seasonlong differences observed among the varieties in most plant development and insect parameters. However, DP50 had significantly lower emergence than the other lines tested (possibly related to seed quality). The lower plant population may have been responsible for greater whitefly abundance observed on two dates mid-season. During early-season ratings of secondary “pests” (15 DAP) (scaled on damage and/or presence), the two test lines received lower ratings for thrips and flea beetle when compared with DP50, DP50B and DP50Bu (untreated for Lepidoptera). However, these difference are likely as a result of the difference in seed treatments that the two test lines received (Gaucho®) and the others did not. This seed treatment does have known activity against thrips and beetle pests. In mid-season, the two test lines received lower ratings for beet armyworm when compared to DP50, DP50B and DP50Bu (although, not significantly different from DP50B or DP50Bu). Efficacy against pink bollworm (PBW) was assessed one time at the end of the season (we were limited to this time, so as to not affect yield), and DP50 was the only variety in which PBW exit holes were observed and PBW larvae collected. However, the low Lepidoptera pressure experienced during the season limited assessments of the two novel lines’ efficacy toward PBW. There was no significant difference in yield (bale/A) among the varieties. Although, one of the test lines had a lower lint turnout than each other variety. The two novel Bollgard lines performed well under our growing conditions, but continued evaluations will be necessary under more conditions and more insect pressures before “varietal” performance and gene efficacy can be assessed adequately.