What do you know about butterflies?
The role of object conservators in mounting a living butterfly exhibition.

Rachael Perkins Arenstein
National Museum of the American Indian
3401 Bruckner Boulevard
Bronx, NY 10461
ArensteinR@si.edu

In 1997 two institutions, the Milwaukee Public Museum and the Academy of Natural Science in Philadelphia, offered AMNH the opportunity to host their exhibitions of living butterflies. After assessing the costs involved in modifying one of the traveling exhibits for the available gallery space, senior museum officials decided that it would be more cost effective to develop a similar exhibit in-house.

Early in the planning stage members of the exhibit project team met with the head of conservation in the Museum’s Anthropology Department and a preservation consultant who specializes in museum environments. They gave recommendations on construction materials for the vivarium, and ways in which its HVAC and lighting systems could be erected in the museum’s Hall of Oceanic Birds without causing unnecessary damage to the installed permanent collection. Normally, the Objects Conservation Lab has a central role in preparing exhibits initiated by the Museum’s Anthropology Department. This was not an anthropology exhibit, however, and after this initial meeting there was no further role perceived for conservation and the lab had no more involvement in preparations.

• The structure had to form a continuous, sealed shell.
• There could be no right angles in the flight zone to prevent the butterflies from getting caught.
• Specifications for appropriate light, temperature and relative humidity levels had to be maintained.
• The 1,300 square foot structure had to be self-supporting so that it did not touch the walls or ceiling in the gallery and it could only be bolted to the floor at a few points.
• The vivarium’s environmental and lighting systems had to be contained so that they would not affect the hall’s murals and dioramas.
• The conservatory had to be easily disassembled, stored and erected for repeated use each winter.

What was originally conceived as a one-room exhibition space became a complex construction project that had to be completed on a tight budget and even tighter time schedule. Construction planning began in early June with installation to begin about three months later.

In late October, the head of conservation was contacted shortly before the opening date with a request for a conservator to install and monitor dataloggers to provide an independent check that the vivarium’s HVAC system was sustaining suitable conditions. After installing the loggers I was told by the exhibits’ coordinator that they might need additional help as a problem had developed - the first group of butterflies released in the vivarium that morning was already dead.

The butterflies’ adverse reaction in the conservatory was immediate; they went into spasms minutes after their release, becoming disoriented and loosing control of their wings. All were dead within one hour. The project’s entomologists knew from anecdotal evidence that this type of death was consistent with an environmental contaminant. Staff at other institutions with similar exhibits were clear that butterflies are extremely sensitive to their environment and that even trace contaminants can lead to death.

We immediately identified three potential sources of contaminants: the make-up air, materials used in constructing the vivarium and the plants. In conjunction with the entomologists we devised a series of simple tests to try and pinpoint the problem. Butterflies were placed in either closed containers or plastic bags to force them into proximity with specific construction materials and plant specimens. Tests were conducted both in the conservatory and in the Entomology Department room where the pupae were hatched.

While the tests were carried out, the project contractor and I quickly began to research the vivarium’s heating and ventilation systems. In order to maintain the 80° Fahrenheit and 80% relative humidity the HVAC system used only 10% fresh make-up air. The system was designed for 20 air changes per hour, which meant that it was unlikely that an outside contaminant would be able to build to a dangerous level.

Along with the architect we began to investigate materials used in constructing the conservatory. Although primarily built of aluminum, fiberglass and acrylic sheeting, there were several materials such as the recycled rubber floor, gaskets, and HVAC duct sealant that could be off-gassing. These materials had not been mentioned in the early planning meeting. The architect and contractors supplied the names of every material used in the vivarium’s construction and Material Safety Data Sheets were obtained for each product.

It was thought that one of these materials could possibly emit enough noxious gasses that in the recycled air could build to levels dangerous for the butterflies. Basic air quality tests showed trace amounts of many of the chemicals and solvents listed on the MSDS forms and a notably high ethanol count. However, tests with the butterflies showed that this was not the immediate problem.

The most obvious source of the problem was the plants. Foliage contaminated with pesticides was the most serious issue mentioned by staff members responsible for butterfly exhibits at other institutions, and several people had related similar experiences with sudden deaths of their live collections. Unfortunately, research on the possible presence of pesticides was complicated by the fact that the greenhouse that provided the plants used several different growers, making problems difficult to trace.

Fortunately, the testing quickly narrowed down the suspects. While all the butterflies tested in the vivarium died quickly, showing the same symptoms, in the pupae room the greatest problems occurred with leaves off of two Ficus trees: butterflies placed in direct contact with leaves from the trees died immediately. Interestingly, even without direct contact death occurred, albeit more slowly. This was a lucky break, the two trees could be definitively traced back to a single grower.

It was discovered that the grower in question had applied a pesticide called Vydate. With this information I was able to provide concrete assistance to the exhibit team. The combination of the conservator’s background in chemistry, together with our frequent work with a variety of hazardous solvents and chemicals, meant that we were experienced in obtaining information and assessing the risks of materials. Tapping into the extensive pesticide databases on the Internet quickly provided additional information. Vydate, a DuPont product with the active ingredient Oxamyl, is a Class I, restricted-use pesticide, commonly used on ornamental plants. It is applied as a root dip and, as such, becomes systemic. The product literature cites extreme toxicity for birds, fish and bees. We discovered that virtually all growers of ornamentals use pesticides: when the greenhouse promised pesticide-free plants, they assumed that after waiting an appropriate amount of time there would be little residual material of danger – an unfortunate inaccuracy in this particular case.

In project meetings, we realized that the entomologists were having trouble communicating their concerns for the butterflies’ safety in the absence of previous experience and hard data. Exhibition staff were equally frustrated by their inability to proceed without guidance. Conservators often find themselves trouble-shooting in the last stages of an exhibition installation and this proved no different. Our important function in these meetings was acting like a translator, supporting the project’s lepidopterist with additional data, while helping the Exhibition staff decide how to proceed safely.

With the publicized opening date less than a week away decisions needed to be made swiftly. The Ficus trees needed to be pulled from the vivarium, but we worried that this alone might not be sufficient, as butterflies showed milder but similar symptomatic deaths with other plants as well. From my conversations with DuPont scientists, I learned that it was possible for this water-soluble pesticide to become, in effect, airborne in the extremely humid conditions in the vivarium. We decided that we would have to pull out the landscaping entirely, wash down the structure, and start over with new, hopefully uncontaminated foliage.

The new selection of plants was made, based on the pesticide practices of the various growers, and staying away from other plants that had not performed well in the bag tests. Unfortunately, none of these plants had the height and coverage of the Ficus, and it was thought important to have some taller plants on which the butterflies could find shelter and rest. We visited the manufacture site of one of New York’s large suppliers of artificial plants. The materials used in construction, including wood, vinyl coated steel wire, polyester, and hot melt glue, tested well with the butterflies. Two large, artificial Ficus trees were purchased, as well as several smaller, bushy plants to fill out the foliage. These were placed at either end of the conservatory, where they would be less visible.

While the landscapers collected the new plants, we worked with the contractors to correct some of the remaining environmental issues that were thought to be placing additional stress on the butterflies. Extra lighting was added to stimulate flight and, more importantly, feeding. The timing of the lights was also coordinated to approximate a day-night cycle, allowing them to rest. Adjustments were made to the HVAC system when, after downloading the data from our loggers, it was discovered that the sensors quickly moved out of calibration in the demanding hot and humid environment. Issues of lighting and environment are common conservation concerns when installing most exhibits and these problems were easily handled.

Having learned our lesson, we started fresh and tried to proceed systematically with the reinstallation, in the hopes that if there would be continued problems they could be isolated swiftly. The entire interior of the vivarium was washed with a dilute mixture of dishwashing detergent and Chlorox bleach in water. The plants were brought into the Museum in batches and washed by our preparators with the same mixture. As the plants acclimated to their new environment, representative samples were chosen to be bagged with butterflies for further testing. Next, the plants were moved into one side of the vivarium, which was partitioned off with netting, forcing the butterflies into contact with the plants while in the conservatory’s environment. These tests ran smoothly and the reinstallation proceeded swiftly to prepare for the revised opening – two weeks after the initial date.

As of the spring of 2001 the Vivarium in three seasons has attracted over 920,000 visitors. There has not been a reoccurrence of the problems we had during the first installation possibly because of the implementation of three procedures before the second season.

1. The project’s lepidopterist identified which species did well in the conservatory and fine-tuned future ordering to include more of the heartier species.

2. More time was planned for reassembly of the structure, installation of the plants and fine-tuning of the environmental conditions. This has allowed for an earlier initial release of butterflies and the time to trouble-shoot if necessary.

3. Procedures were set up for early identification and testing of any new materials specified for the vivarium before purchase and installation.

Off-gassing from the rubber floor and one of the HVAC duct sealants, although unproven as contaminants, remained under suspicion. Retesting the air quality at the end of the exhibit’s run showed that most of the contaminant levels had decreased slightly, including the high ethanol count, which was never fully explained. To eliminate the flooring as a factor in the vivarium’s second season, alternate materials were investigated. Five months before the start of installation, the architect supplied the Conservation Lab with samples of several materials that were subjected to our standard Oddy Test. This accelerated aging test subjects the sample to high heat and humidity similar to the conditions in the conservatory. The test is evaluated by assessing the state of three types of metal coupons at the end of thirty days. This was a way to weed out materials that seriously off-gassed. The remaining candidates were also tested with butterflies to ensure that they were suitable.

None of the problems we encountered in setting up our conservatory was unusual. Other institutions experienced with living collections cited the same challenges. What was new to us, however, was the fact that with sensitive species there is no margin for error. In summary, we were able to make several recommendations for those contemplating a living exhibit.

• Allow ample time for information gathering, proper planning, installation and testing - to make sure that problems can be corrected early not under pressure.
• Know what you are dealing with. Speak with other institutions or specialists who can provide advice on what challenges to expect.
• Ensure that you have not only the resources but also the expertise on staff - ideally someone trained with living collections as well as the species’ natural history.
• Carefully check all materials used in the environment and test their suitability. If possible, buy the materials in advance to allow them to off-gas safely.
• Use trusted contractors but do not rely on them to understand the needs of your collection.
• Ensure that the environment is appropriate and stable before introducing the collections.
• And finally, do not expect to be in full control. In natural history Museums we have perfected the display of dead and mounted specimens, but there is an element of risk and unpredictability associated with living collections that should not be underestimated.

Exhibit staff often sees conservators as an obstacle to displaying an artifact, while the conservators see themselves as advocates for the preservation of the collection. It was not an unreasonable assumption to assume that, after the initial meeting, object conservators had little to offer in staging this kind of exhibit. However, our chemical literacy, the ability to devise methodical testing and our familiarity with environmental problems proved extremely useful. With our help trouble-shooting on this exhibit, conservators were seen not as impediments, but as participants, acting as a liaison between the scientific and exhibition staff. The quick identification of our problems and the solutions for reinstalling the conservatory was possible because everybody had the same mission – to build an environment in which the butterflies could live safely and be enjoyed by the public.