Height data were collected three times weekly between pinch and flowering to represent `Royalty' rose (Rosa hybrida L.) response to 15 unique treatment combinations of irradiation as photosynthetic photon flux (PPF: 50 to 300 μmol·m-2·s-1), day temperature (DT: 12 to 22 °C), and night temperature (NT: 15 to 25 °C) under constant growth chamber conditions. Combinations were determined according to the rotatable central composite design. A previous full quadratic model approach was compared with a revised approach using a nonlinear Richards function derivative form. This allowed a dynamic change of parameter values for each daily growth iteration by computer. The Richards function assumes nonconstant daily growth rates are proportional to current size; Euler integration enabled additive accumulation of these values. Ratios of the growth constant (k) to the theoretical catabolic constant (m = v+1) caused flexible changes in the growth curve, which were compared with the previous quadratic approach.
Douglas A. Hopper
One-year-old plants of four cut rose (Rosa hybrida L.) cultivars were grown under either natural or supplemental irradiance for 4 months during the winter in Colorado. Supplemental irradiance with high-pressure sodium (HPS) lamps was supplied at 100 μmol·m–2·s–1 for 10 h each night during off-peak electrical use periods. Total cut flower yield, stem length, and fresh weight of individual flowers were recorded. The number of flowers produced and fresh weight increased for all cultivars under the supplemental irradiance treatment. Flower count, stem length, and fresh weight showed significant differences among the four 4-week production periods; production differences were promoted through pinches of two stems per plant to time for holiday peaks. When production was highest, stem length and fresh weight were lower, most likely due to redistribution of the limited carbohydrate pool during the winter.
Douglas A. Hopper
Ninety-six uniform plants of each `Russell hybrid' and `Gallery' mix lupines sown 9 June 1995 were randomly assigned to 32 unique treatment combinations. On 14 Dec 1995, plants were either placed in a 17/13°C day/night temperature (DT/NT) greenhouse (COOL) or 22/18°C DT/NT greenhouse (WARM) as controls, or in a constant 4.5°C cooler in the dark for 6, 8 10, or 12 weeks. After cooling, plants were transplanted to #1 nursery cans (2.75 liter) using Sunshine mix #2 and were assigned randomly to the COOL or WARM greenhouse. Greenhouse control plants under natural days were transplanted at intervals similar to cooled plants. Days until visible bud and flowering were analyzed using SAS PROC GLM. Plants receiving long day (LD) flowered 7 to 10 weeks (46 to 70 days) after the start of LD forcing. Buds were visible in 30 to 35 days. Plants receiving natural days (ND) did not flower uniformly unless they were cooled for 12 weeks, yet flowering took longer (8 to 12 weeks) when compared with LD. Unfortunately, LD lighting for the entire forcing period caused excess stretching, so plants finished too tall for quality potted plants. Forcing in a COOL greenhouse delayed flowering about a week compared to the WARM greenhouse.
Douglas A. Hopper
Improvements to computer software and advancing technology made it necessary to convert the computer greenhouse simulation model, GHSIM, to a new application for operation across a greater number of platforms. Originally, economics and internal organization compatibility led to use of spreadsheet Quattro Pro. Standard features were relative and absolute references, multiple pages for topic organization, random event generation, and graphing of calculated trends. However, Quattro Pro contained many convenient features, yet proprietory, which were not readily converted: certain formats for graphing trends, recursive formulas, cross page referencing, buttons, macros for dynamic time execution, and floating toolbars that actually changed between old and newer versions (v.5.0 vs. v.7.0). Translation from Quattro Pro v.5.0 to Microsoft Excel 97 produced tedious page by page (worksheet) conversions, loss of buttons and macros, distorted/unreadable graphs, nonexistent toolbars, and, most troubling, obscure problems with recursive execution causing Excel to crash amid nondescript error messages and a core dump. All these were eventually resolved; current efforts seek to reach other platforms, including MacIntosh and the Internet.
Douglas A. Hopper
One should choose the simplest form of a model as a tool that adequately represents the processes and relationships of interest. ROSESIM was first developed in SLAM II and FORTRAN to run on a mainframe computer, where it had few users and it was cumbersome to learn and use. As use of models on a personal computer (PC) has become more popular for instruction and simulation, ROSESIM was translated first into the American Standard Code for Information Interchange (ASCII) to run in the Beginner's All-purpose Symbolic Instruction Code (BASIC) language in the popular Microsoft Disk Operating System (MS-DOS). As graphical user interface (GUI) Windows applications have gained increased popularity, ROSESIM has been translated into C++ as object-oriented programming (OOP) to run inside Microsoft Windows 3.1. This makes ROSESIM for Windows readily available to virtually every PC user. Features of ROSESIM for Windows are listed and discussed.
Douglas A. Hopper and Kevin Cifelli
An interactive simulation model of plant growth must be flexible to accept specific crop equations from the user. ROSESIM functions as a dynamic plant growth model based on `Royalty' rose (Rosa hybrida L) response to 15 unique treatment combinations of photosynthetic photon flux (PPF), day temperature (DT), and night temperature (NT) under constant growth chamber conditions. Environmental factors are assumed constant over an entire day. Coefficients are read from an external ASCII file, thus permitting coefficients from any linear, quadratic, or interaction terms to be input into ROSESIM up to a full quadratic model form. Nonsignificant terms are given a coefficient of zero. ROSESIM has been restructured into Borland C++ object oriented program (OOP) code to execute in the Microsoft Windows 3.1 operating environment. This enables ease of operation in the user friendly graphical user interface (GUI) provided with most IBM personal computers (PC). The user chooses a set of environmental conditions which can be altered after any selected number of days, allowing conditions to be changed and modeled daily for interactive comparison studies.
Douglas A. Hopper and Julie A. McIntyre
Research focused on alternative methods to control Western flower thrips (Frankliniella occidentalis Pergande), encompassing chemicals from varying classes, parasitic nematodes, microbial insecticides, and physical/mechanical deterrents. Chemical spray applications were applied weekly for 4 to 6 weeks. Experiment 1 made comparisons between fenoxycarb (Precision), bifenthrin (Talstar), and entomopathogenic nematodes (Biosafe). Experiment 2 compared abamectin (Avid), spinosyn A and D (Spinosad), azadirachtin (neem extract: Margosan-O), and diatomaceous earth (a physical control aimed at deterring pupation). Experiment 3 compared Spinosad, fipronil, and two microbial insecticides (Naturalis-O and Mycotrol). The number of thrips counted in flowers after treatments had been applied indicated that the strict chemical treatments (Avid, Spinosad, fipronil) provided quick knockdown and overall longer-term population control. Microbial insecticides, diatomaceous earth, and nematodes maintained populations at a lower level than the control, but were not as effective as strict chemical controls. Margosan-O, Precision, and Talstar controlled populations at medium levels. For periods when populations may cycle upward, more potent chemicals could be used (Spinosad, fipronil, and Avid) while still avoiding problems associated with more toxic chemicals.
Douglas A. Hopper and Steven E. Woerner
Three-year-old Rosa hybrida L. `Royalty' and `Red Success' plants were pinched 20 Oct. and 22 Dec. 1990 to time production for Christmas and Valentine's Day, respectively. Two greenhouses received ambient solar radiation while two additional houses had 50% reduction in irradiance using a shading thermal blanket. All temperature set points were 23C/17C day/night. Every 10 days and at flowering shoots were measured for leaf (node) number, stem diameter, stem length, and fresh weights of stem, leaves, and flower bud. Time to visible bud and to flowering from pinch were recorded.
A computer model, ROSESIM, had been formulated in SLAM II and converted to both FORTRAN and BASIC code to make the program more portable. Although ROSESIM closely predicted `Royalty' fresh weights when irradiance was high before Christmas, underprediction occurred in the lower irradiance before Valentine's Day, and time to flowering was predicted earlier than was observed under all conditions. Corrective coefficients were added to ROSESIM to improve accuracy of prediction under actual greenhouse conditions.
Steven E. Woerner and Douglas A. Hopper
A computer simulation model was developed to be used in evaluating irrigation scheduling techniques and assisting irrigation scheduling decisions under greenhouse conditions in Colorado. The model simulates variable greenhouse conditions and shows how each of four irrigation scheduling techniques responds to these conditions. Reports from the model detail numbers of irrigation events, sensitivities to parameters, and forecasts water usage. The model was also used to determine appropriate accumulation triggers for Colorado conditions.
Four techniques evaluated here include: time clock control; accumulated radiation; accumulated vapor pressure deficit; combination method (radiation and vapor pressure deficit). The model has shown the combination method to be the most sensitive to changes in environmental conditions, while the time clock method proved to be least sensitive (and most wasteful of water).
The model may evaluate additional irrigation scheduling techniques by including additional parameters in the model, and may readily be adapted to different climatic regions.
Douglas A. Hopper and P. Allen Hammer
Day (DT) and night temperature (NT), and irradiation (photosynthetic photon flux-PPF) treatment combinations were run from a rotatable central composite design. At flowering, response surfaces were plotted for `Royalty' rose for plant growth characteristics: node number, stem diameter, stem length, and dry weight of stem, leaves, flower bud, and total shoot. Overall development was recorded from pinch until visible bud, first bud color, and flowering. All characteristics except node number and stem length were significant (P=0.05) for a full quadratic model having ten estimated coefficients. Greatest dry weights were predicted at flowering for the lowest DT (15C) and NT (12C) combination, and required the longest time for development. Conversely, most rapid development and least dry weight accumulated for high DT (25C) and NT (22C). Thus, a compromise exists between rapid development and quality of the flowering stem. As PPF increased from 50 to 300 μmol m-2s-1, predictions for dry weights increased while developmental time decreased.