Authors: Nick Andrews, Small Farms Program, Oregon State University & Len Coop, Integrated Plant Protection Center, Oregon State University
Publish Date: Fall 2011
Temperature controls the development of many organisms that don’t have complex thermoregulatory systems. This allows us to predict the development of different organisms by accurately using development or phenology models which are based on the accumulation of heat units (i.e. degree days) during a growing season. These sorts of models have been developed for many plants, insects and plant pathogens.
There are other factors that can affect the development rate of organisms such as moisture, day length and competition. However, simple degree day (DD) models can accurately predict development rates within a few days, especially if some assumptions can be made about other factors affecting development rates, such as the presence of adequate moisture as provided in an irrigated field.
Nearly all seed catalogs report the time to maturity of different vegetable varieties in number of days to maturity. If all vegetables were bred in our region, that could be fairly accurate despite considerable variation in weather from year to year. However, a month in Florida or California provides a lot more DDs than a month in Oregon or Washington. Reported days to maturity can give a relative idea of which varieties take longer to mature, but they normally don’t help schedule plantings or harvests with much accuracy.
Many farmers get a rough feel for this after years of experience with a variety, and some farmers and agricultural companies have developed their own DD models for the main varieties they grow. They can use this information to schedule plantings to provide the volume of harvest they want at different times of the year. When inclement weather interrupts the planting schedule, they can use this information to select different varieties that will get their harvest schedule back on track. As harvest time approaches, they can also use this information to communicate with buyers when their crop is ripening a bit earlier or later than expected. If climate change impacts local weather in the Pacific Northwest, this sort of information will help farmers adapt to years with poor growing conditions and take advantage of years with better growing conditions.
Crops have upper and lower development thresholds outside of which they don’t develop physiologically. One of the simplest ways of counting DDs for one day when the crop’s lower temperature threshold is known is: [(Tmax + Tmin) / 2] – T lower. For example, assume the crop’s lower threshold is 50°F and upper threshold has not been determined or is very high and therefore not used. On a day with a Tmax of 90°F and a Tmin of 40°F, the DDs for that day = [(90 + 40)/2] – 50 = 15. This method of calculation is known as the “Simple average DD method.” Since many crops such as corn are only responsive within a range of temperatures defined by the thresholds, substitutions are made; if the daily Tmax or Tmin are above or below the thresholds, they are reset to the threshold.
For example, using the same daily temperature values as above, and adding an upper threshold (used for corn) of 86°F and a lower threshold of 50°F, we would reset the Tmax from 90 to 86 and Tmin from 40 to 50 and calculate: [(86+50)/2]-50 = 18 growing degree days (GDD). A slightly more accurate method is to calculate the area under the curve between the maximum and minimum thresholds as shown by the shaded area in Fig. 1. Most insect DD models used in Oregon use a version of this known as the “single sine DD method.” Most models use daily maximum and minimum temperatures, but some have been developed using hourly temperatures divided by 24. Some instruments such as “Bioaccumulators” and weather stations with custom software can accumulate DDs with precision to the minute or less. Unfortunately, the different methods of calculating DDs are NOT interchangeable; thus users of DD models must pay strict attention to the calculation method that was used to develop the model, and to adhere to that method when using the model. So, for example if a model specifies a “single sine method”, then only that method can be trusted to be unbiased.
Temperature thresholds are somewhat complicated to determine, and require independent research to estimate the maximum and minimum temperatures which limit physiological development. However, these temperature thresholds don’t usually vary very much within a crop species or group of closely related species such as sweet corn or Brassicas
. Lower development thresholds for many crops are published in Knott’s Handbook for Vegetable Growers and the scientific literature. If these thresholds are known for your crop, you can use weather information from your farm or a nearby weather station (see below) to observe how many DD are needed before harvest or other event of interest (i.e. canopy closure or flowering) for your specific varieties. You will have generated a simple DD model for your variety that can be validated and used in subsequent years.
OSU’s Integrated Plant Protection Center (IPPC) has a collection of more than 90 DD and hourly-driven plant disease risk and chilling unit models, all pre-configured for individual insects, plants, and plant diseases, plus a generic DD calculator, that are freely available online at http://uspest.org/wea/. The models on the website link to a large network of more than 15,000 weather stations across the United States. In Fig. 3 we display an example output from running the Jubilee sweet corn model (reference above) at the website, using May 1, as the planting date over the years 2009-2011, and the AgriMet weather station ARAO located in Aurora, Oregon. This particular model has been run over 2,600 times at the website and has been used for scheduling planting and harvest dates for processed sweet corn.
The sweet corn DD model example in Fig. 3 shows how DDs may be used to improve the scheduling of harvest dates in years with differing temperature regimes. It illustrates how you can begin to develop models for your own needs by careful record keeping of crop development for selected events and then later running a DD calculator that will help you determine the average DD requirements for each event. If funded, a new proposed project (see below) will provide training and instruction on how to do this step-by-step. This online system is widely used in the tree fruit and nut industries to manage pests by estimating phenological events (e. g. egg laying) that can be used to reduce risk and time treatments. From the website (Fig. 4) go to Quick Start then select the crops you are interested in. When you enter your zip code the program automatically selects the closest weather station and produces a graphical display of all the available models that are relevant to your crops. Currently, this system has about 11 models of interest to vegetable growers including the late blight, early blight, cabbage maggot and Jubilee sweet corn development models. If you know the maximum and minimum thresholds and DD accumulation required for the varieties you grow, you can use the full featured DD calculator at this site: http://uspest.org/cgi-bin/ddmodel.pl. In addition to the models hosted by the IPPC, there are some other sources of DD models. For example, Crookham Company (www.crookham.com/) has DD models for their sweet corn varieties and UC Davis has an extensive collection of DD models (http://www.ipm.ucdavis.edu/MODELS/index.html). We are working with several vegetable growers on a new project proposal to increase the number of development models available to vegetable growers and develop a new interface for vegetable crop scheduling and management. If funded, Jim Myers (OSU Vegetable Breeder) would develop DD models for the varieties he works on, Dan Sullivan (OSU Soil Scientist) would work on a N-mineralization model, Ed Peachey (OSU Weed Scientist) would work on weed development models and Nick Andrews (OSU Small Farms Extension) would work on cover crop development models. Len Coop (IPPC) would lead development of the website and modeling system. The project team would work with growers to help them develop DD models for their own vegetable varieties.
We would also develop a new webpage that would help vegetable growers use these models to plan successive plantings, manage crops and schedule harvests. We are already working with a group of about 10 local producers on this proposal, but are looking for other interested growers. If you are interested in this project please send your contact details to Nick Andrews (email@example.com). Your interest may help us secure funding, and if the project is funded, we’ll get in touch to let you know how you can participate.