Abstract Tomatoes spaced 18 in apart were transplanted on 29 Jan in 9-row plots 24 ft long on 9 raised soil beds (6 ft centers). Beds were fumigated at formation with 13.5 gal/acre Vorlex and covered with whiteface polyethylene mulch. Directly adjacent were 3 beds of 100% geminivirus-infected tomatoes planted the previous November and partially protected from a December freeze. A completely randomized split block design was used with 4 replications and 3 main plot treatments; JMS Stylet Oil, insecticide, and an unsprayed check. Insecticide plots were split into 3 subplots, each 3 rows wide. Concentrations of all treatments were fixed throughout the experiment but rates per acre increased as nozzles were added to accommodate larger plants. Manzate 200 (1.5 lb/acre) and Tri-Basic Copper (3 lb/acre) were applied weekly, Bravo (2 lb/acre) and Javelin SC at 1.2 qt/acre were applied twice to control Spodaptera spp. Treatment applications began 12 Feb and continued twice weekly until harvest using a tractor-mounted boom sprayer with 4 drop lines holding a maximum of 24 nozzles fitted with Albuz ceramic hollow cone tips (ATR brown delivering ca. 0.26 gal/min @ 400 psi). Whitefly immatures were counted at weekly intervals on the terminal leaflet of the 5th, 6th, or 7th leaf from the top (depending on plant maturity) from 6 randomly selected plants per subplot. Adults were sampled by carefully inverting the 3rd leaf from the top of 3 randomly chosen plants per subplot. Plants affected by geminivirus were marked as symptoms appeared. All fruit of marketable size were picked and graded on 19 Apr. Culls were removed and the remaining fruit separated into medium, large, and extra-large categories with a mechanical grader. Dollar value was calculated on the basis of $5.00/25 lb box for medium and large fruit and $7.00/box for extra-large fruit. A repeated measure analysis of variance for each whitefly lifestage was used over all sample dates and a separate analysis of variance was calculated for adults on each of the last 4 sample dates. A log10(x + 1) transformation was used to normalize counts of immatures because of the large number of zeros and high numbers in the data set.
Read full abstract