Effects of Landscape Pattern on the Distribution of Feral Honey Bee Colonies in South Texas

Kristen A. Baum1, William L. Rubink2, Robert N. Coulson1, Douglas F. Wunneburger1,3

Knowledge Engineering Laboratory, Department of Entomology, Texas A&M; 2USDA, ARS, Beneficial Insects Research Unit, Weslaco, Texas; 3GeoInformatics Studio, Department of Architecture, Texas A&M



Honey bees (Apis mellifera L.) play an important role in many ecosystems, pollinating a wide variety of native, agricultural, and exotic plants. In the United States, managed honey bee colonies declined about 25 percent from 1995 to 1996. Few studies have documented population trends for feral colonies, but Loper (1997) reported an 82 percent decline in spring populations from 1992 to 1997. More data on the population dynamics of feral honey bees are needed to address important issues related to pollination and the spread of Africanized honey bees.

Our goal was to study the spatial and temporal distribution of feral honey bees on the Welder Wildlife Refuge in San Patricio County, Texas. Specific objectives included developing a spatial database of the Welder Wildlife Refuge, identifying habitat features associated with feral colony locations, and evaluating the spatial and temporal dynamics of the feral populations.



The decline of managed and feral colonies has been attributed to parasitic mites, pesticide use, and the arrival of Africanized honey bees. Tracheal mites (Acarapis woodi Rennie), first found in the United States in 1984, live exclusively in the prothoracic tracheae of adult bees, forming feeding lesions on the tracheal wall and disrupting respiration. Varroa mites (Varroa jacobsoni Oudemans) arrived in the United States in 1992. Female mites attach to adult bees between abdominal segments or body regions, where they feed on hemolymph. Female Varroa enter brood cells shortly before capping, where they feed on larval hemolymph and lay eggs. The developing mites feed on the honey bee larvae, causing malformations and decreasing the overall vigor of the bees.

Africanized honey bees arrived in Texas in 1990, with the first record in San Patricio County reported in 1992 (see Figure 1). Due to their accentuated defensive behavior and increased rates of swarming and absconding, Africanized honey bees are incompatible with current beekeeping practices and may cause problems in areas of livestock and human habitation.


Figure 1. The spread of Africanized honey bees in Texas.




Study sites were located on the Welder Wildlife Refuge, a 7800-acre ranch in San Patricio County, 35 miles north of Corpus Christi, Texas. The refuge is actively managed for cattle ranching and supports a wide array of wildlife research. Feral colonies have been monitored on the Welder Wildlife Refuge since 1992 by Dr. William Rubink. Data collected included cavity status (active or inactive), substrate (tree species), and entrance orientation. Samples of bees were collected from each colony during surveys and examined for tracheal and Varroa mites, and for levels of Africanization.


A base map was generated from color infrared aerial photographs acquired at 5500 feet and 11500 feet above ground level during October 1999. Color positive transparencies were scanned at 600dpi. Each image was geographically registered using Erdas Imagine®. Harvester ant (Pogonomyrmex sp.) mounds served as ground control points (see Figure 2). Coordinates were recorded using a Trimble GPS PathfinderTM receiver and Asset SurveyorTM data logger. A mosaic of the registered images was created with Intergraph® IRASC. We also recorded the spatial coordinates for each colony site. When cavities were located in areas of dense canopy cover, we used an Advantage® Laser Rangefinder to calculate the offset for each colony location.


Figure 2. Locating and recording ground control points (Pogonomyrmex mounds).



Results and Discussion

Ninety-five colony locations were found on the Welder Refuge. Each cavity contained a honey bee colony found to be active during at least one of the surveys (see Figure 3).


Figure 3. A colony located in a living live oak (Quercus virginiana). Cavity locations are shown in green.


Eighty-eight percent of the cavities used by feral colonies were located in live oak mottes (see Figure 4). Other cavity tree species included anacua (Ehretia anacua), cedar elm (Ulmus crassifolia), sugar hackberry (Celtis laevigata), and red mulberry (Morus rubra ). Most of the cavities faced north, south, or west (see Figure 5).


Figure 4. Tree species of cavities.


Figure 5. Orientation of cavity entrances.


Honey bees select for certain cavity attributes. Cavity volume and shape constrain colony size and influence microhabitat conditions within the cavity. Other factors include orientation in relation to the sun and wind, as well as the surrounding environment. Proximity to target resources, such as pollen, nectar, and water sources, also contribute to the suitability of a cavity. A future goal of this study is to define the functional heterogeneity of feral honey bee habitat. Functional heterogeneity refers to how an animal perceives and responds to variations in landscape structure. Functional heterogeneity can be evaluated by identifying target resources for feral honey bees, developing habitat suitability maps, and applying a weighted connectivity index (Coulson et al. 1999) to evaluate the spatial connectivity of target resources.


Figure 6. An automated mosaic of part of the Welder Wildlife Refuge showing locations of the feral honey bee colonies in green.


The percent of active cavities remained above 80 percent until 1995 (Figure 7). From 1995 to 1996, the numbers decreased from 70 percent to 30 percent. From 1996 to 1999, the number of active colonies increased from 35 to 57 percent.


Figure 7. The number of active and inactive cavities for each year.


Future Directions

The observed population trends correspond to the arrival of Varroa mites and Africanized honey bees. Varroa mites were first recorded on the Welder Refuge in 1995. By 1996, the feral populations had declined dramatically. Africanization began to increase in 1995, with all the colonies Africanized by 1998 (unpublished data, William Rubink). These data suggest populations of European feral honey bees were decimated by Varroa mites in 1995, and then replaced by Africanized honey bees in the following years. Colonies will continue to be monitored throughout the duration of this study.  

Pollination ecology and honey bee population dynamics encompass a broad range of spatial and temporal scales. We suggest a great deal can be learned by examining a natural population of honey bees in a landscape ecological context. Such an approach utilizes novel analytical tools to address issues related to pollination, feral honey bee distribution, the spread of Africanized honey bees, and the role of certain disease factors, such as parasitic mites.



We examined the distribution and abundance of feral honey bees on the Welder Wildlife Refuge. Colonies were located in live oak mottes. Entrance orientation was variable. The decline in the feral population corresponded to the arrival of Varroa mites, while the recovery can be attributed to the arrival of Africanized honey bees. Further investigation using a landscape ecological approach provides promise for predicting the role of Africanized bees in more northern latitudes.


Literature Cited

Coulson, R. N., B. A. McFadden, P. E. Pulley, C. N. Lovelady, J. W. Fitzgerald, and S. B. Jack. 1998. Heterogeneity of forest landscapes and the distribution and abundance of the southern pine beetle. Forest Ecology and Management 114: 471-85.

Loper, G. M. 1997. Over-winter losses of feral honey bee colonies in southern Arizona, 1992-1997. American Bee Journal December: 823.



Funding for this project was provided by the Welder Wildlife Foundation, the USDA, ARS, Beneficial Insects Research Unit in Weslaco, Texas, and the Texas Legislative Initiative: Protection and Management of Honey Bees - Pollinators of Agricultural Crops, Orchards, and Natural Landscapes.