Secret Site Map
Saturday, April 19, 2014

Roger E. Gold and Barry D. Furman

The authors are with the Department of Entomology at Texas A&M University, College Station, Texas. Email Gold at and Furman at


[Cover Story] Want to Predict This Year’s Termite Swarms?*

Cover Story

How do termite swarms foretell a successful termite season? Research out of Texas A&M University finds there are two simple — yet reliable — tips to determine when termites are going to swarm.

February 26, 2013

Those of us who have business and research interests in termites have always had an interest in the timing and intensity of termite swarms.

The alates that make up the swarms are the adult reproductive caste of the colony, and they play an important role in keeping the genetic makeup of the species as diverse as possible. One of the interesting things to observe in nature is the synchronization of the emergence of adult insects in time and space to complete the reproductive processes. It is this genetic diversity that is critical to survival of populations through time.

Termites are among the most ancient of animals and have been on the planet for millions of years. Subterranean termites (Rhinotermitidae) have primary reproductives (swarmers), but also can produce secondary and tertiary reproductives where workers, or nymphs, can mature sexually, while never developing wings and going through the swarming processes (Miller 1969; Thorne and Forschler, 1998; Gold and Jones, 2000). This alternative form of reproduction has certain advantages for immediate survival of a colony, but, through time, the lack of genetic diversity (inbreeding) has major limitations.

One of the challenges of living and working in Texas is that we have four families of termites, each of which has a specific period in which they undergo swarming behavior. In general, the swarms start in the late winter and early spring with native subterranean termites (family Rhinotermitidae). These are followed by Formosan termites in early May (Mother’s Day swarmers). In late spring to early summer we have drywood termites (family Kalotermitidae) and dampwood termites (family Termopsidae). Finally, the desert and agricultural termites (family Termitidae) are in the fall. In the summer months, there are overlaps with the swarming of different species of termites, and so proper identification is critical to the proper diagnosis and treatment of termite infestations.

We undertook an initial investigation (Furman and Gold, 2002) to determine if there was a correlation between temperature, precipitation and region of the state (Texas is approximately 1,000 miles across at the widest points), as it related to native subterranean termite (Reticulitermes spp.) swarming behavior. This information has applicability to termite swarms throughout the country.

Collecting Swarm Data. Homeowners generally recognize that they have a problem with termites during the swarming season. It is often at this time that a homeowner first contacts a pest management professional for an assessment of the problem and a possible solution. Our interest in studying termite swarming was in response to pest management professionals who wanted to be able to predict termite swarming in order to provide focused advertising just prior to swarms. There have been other studies of termite swarming based on collecting and trap results (Howell et al. 1987; Scheffrahn et al., 1988; Henderson 1996; Howell et al., 2009).

We wanted to expand our data on termite swarms, so we collaborated with personnel from Orkin and reviewed the firm’s telephone contact information (call-in sheets) and the date of sales (sales closure sheets) for subterranean termite control contracts for several years. We were able to collect this data from nine cities in Texas including Corpus Christi, San Antonio, Houston, College Station, Waco, Tyler, Dallas, Lubbock and Amarillo. The concept was to recover data representing termite swarming from the coastal region up to, and including, the high plains of Texas. The climatic differences between these regions were dramatically different in terms of mean temperatures, amount of precipitation, as well as other criteria such as elevations, soil types and density of people and structures in the areas.

The data on swarming was based on the date that termite swarming calls for service were received, and the date that a service contract was purchased by homeowners. Two computer models were developed to display and provide statistical analysis of each data set. Temperature and precipitation data were collected from the Texas Office of the State Climatologist for a survey period of at least five years from each of the nine locations.

In order to use the temperature data, it was necessary to convert the mean daily temperatures into heat units (Higley et al., 1986), which were then correlated to the mean peak swarm dates. The accumulated heat unit model was started on Dec. 21 (the winter solstice) of each year, and up to, and through, the peak swarming dates for that year (Judd and Gardiner, 1997). We also set a base temperature of 4°C (39°F) (Veeranna and Basalingappa 1989) at which time positive heat units began to accumulate. Stated another way, heat units are like a bank account that fluctuates through time with an overall balance figured at the end of the reporting period. If the temperature is above 4°C, heat units were accumulated; if it was colder than this temperature, heat units were not added to the total.

Because termites are poikilothermic (cold-blooded), their physiological development is based on the temperature of their physical environment and when heat units are not being accumulated (below 4°C), their development is slowed or stopped. As the temperatures rises, the rate of the physiological and morphological completion of the swarmer caste is increased, and then when fully developed, the adult reproductives are genetically pre-determined to then go through the swarming process, if the proper environmental conditions are met.

Heat and Precipitation. The next phase of this work was to correlate the heat unit and precipitation (rain or snow) data with the estimated peak swarming dates from each of the nine areas of the state. What we learned was that the average earliest date for subterranean termite swarming was Feb. 23 in Corpus Christi, while the peak swarm in Amarillo was on May 6. In general, the swarm dates are later in the spring for the northern latitudes as compared to the southern. This is closely correlated with the Hopkin’s Law of Bioclimatics (Hopkins 1938) which correlated the phenological development of plants and animal to the “advance of spring,” which was predicted based on the relationship of elevation, latitude and longitude. In Texas, spring moves from south to north at the rate of approximately 25 miles per day.

There was a gradient in swarm dates within the nine locations, with a predicted date of swarming on March 15 for Reticulitermes flavipes (eastern subterranean termites) for the middle of the state. It was further determined that the minimum number of heat units accumulated for swarming was 602, with this accumulation being reached first in Corpus Christi and last in Amarillo. The next correlation attempted was that of amount of precipitation (rain and snow) to swarming. There was NOT a strong correlation in this regard, because of the large variations in annual precipitation amounts across the state. But, there was a good correlation to when it rained, and when termite swarming started.

Based on an in-depth analysis of 54 unique swarming events, after 602 heat units were accumulated, 24 percent of the swarms occurred on the day of rain, 48 percent within one day, 83 percent within two days and 91 percent within three days of a rainfall event. We all understand that there are changes in weather patterns through the years, but climatologists recognize that these variations follow cycles, which are somewhat predictable. With the use of huge weather data sets and predictive computer models, forecasting is very reliable for a specific area of the country. In general, weather patterns for a specific date and location are about the same, within about seven days. In this regard, there is an 80 percent chance of the weather being the same tomorrow as it is today; it is that 20 percent chance of change that makes life more interesting.

What it means. The information on termite swarming dates was summarized and presented to Paul Hardy, senior technical director for Rollins, and Orkin’s technical team. Hardy then replied, “What you are saying is that the subterranean termites will swarm within a few days of when they swarmed last year, if it rains.” Yep, that sums it all up very well! I guess that Paul’s 50 years of experience with termites really is important! (Authors’ note: We will really miss Paul as he prepares for his retirement later this year.)

While there are fluctuations in swarming dates (Scheffrahn et al., 1988), and the numbers of swarmers that emerge in any given year, termite colonies will swarm eventually, as they mature. The best way to anticipate the peak dates of the swarm for this year are for PMPs to check their customer call sheets and termite treatment records from the last few years, and then watch the weather for the predictions of first seasonal rains. It is these rains that trigger the swarming behavior due in part to the rise in humidity, which increases the survival of the alates. If dry weather occurs during the normal swarming season, the swarmers will hold in the tubes as long as they can before they begin to leave in small groups, return to the colony or die. In dry years, the end results are “poor” swarms wherein the alates emerge from the colonies over several days, rather than the dramatic swarms that keep our telephones ringing.

Another observation that our research group has looked at is that frequently, the first subterranean swarmers are called in by customers who received a termite treatment the past year. Sometimes these have been referred to as “panic swarms,” which has a double meaning. The first “panic” is on the part of the customer, who was educated about termites as a result of your sales contacts, and because they expected control of the termite populations. The other “panic” behavior occurs in many plant and animal populations when put under stress, due to unfavorable environmental conditions or sudden population reductions, which will then show a “biological panic” and put disproportionate resources into reproduction in the form of early seeds (plants), or in the case of termites, untimely swarmers.


The authors are with the Department of Entomology at Texas A&M University, College Station, Texas. Email Gold at and Furman at


Furman, B.D. and R.E. Gold. 2002. Prediction of spring subterranean termite swarms in Texas with relations to temperature and precipitation. In Proceedings, 4th Intern. Conf. on Urban Pests 4:303-318.

Gold R.E. and S.C. Jones. 2000. Handbook of household and structural insect pests. Entomological Society of America Handbook Series.

Henderson, G. 1996. Alate production, flight phenology and sex-ratio in Coptotermes formosanus Shiraki, an introduced subterranean termite in New Orleans, Louisiana. Sociobiology 28: 319-326.

Higley, L.G., L.P. Pedigo and K.R. Ostlie. 1986. DEGDAY: A program of calculating degree days and assumptions behind the degree day approach. Environ. Entomol. 15:999-1016.

Hopkins, A.D. 1938. Bioclimatics-A science of life and climate relations. U.S. Department of Agriculture Misc. Pub. No.280 (January 1938).

Howell, H.N., Jr., J.W. Austin, and R.E. Gold. 2009. Swarming dates and distribution of Zootermopsis laticepts Banks (Isoptera: Termposidae) alates in El Paso, Texas. J. Agric. Urban Entomol. 26(1): 11-21.

Howell, H.N., Jr., P.J. Hamman, and T. A. Granovsky. 1987. The geographical distribution of the termite genera Reticulitermes, Coptotermes and Incisitermes in Texas. The Southwestern Entomologist 12(2): 119-125.

Judd, G.J.R. and M.G.T. Gardiner. 1997. Forecasting phenology of Orthosia Hibiscui Guenee (Lepidoptera: Nocatuidae) in British Columbia using sex-attractant traps and degree day models. Can. Entomol. 129: 815-825.

Miller, E.M. 1969. Caste differentiation in the lower termites, pgs. 283-303. In K.Krishna and F.M. Weesner [eds.], Biology of Termites. Academic Press, New York, N.Y.

Scheffrahn, R.H., J.R. Mangold and N.Y. Su. 1988. A survey of structure-infesting termites of peninsular Florida. Fla. Entomol. 71: 615-630.

Thorne, B.L. and B.T. Forschler. 1998. NPCA research report on subterranean termites. National Pest Control Association, Dunn Loring, Va.

Veeranna, G. and S. Basalingappa. 1989. Nesting pattern of the termites Odontotermes obesus Rambur and Odontotermes wallonensis Wasmann (Isoptera: Termitidae). Insect Sci. Appl. 10: 169-180.


For more information on Texas A&M’s work with termites, visit PMPs’ questions, comments and suggestions are always welcomed. — Gold and Furman

Top Photo: David Cappaert,

Roger E. Gold and Barry D. Furman Archive