2013 NORTHEASTERN NATURALIST 20(3):441–450 Freeze-Drying to Preserve Birds for Teaching Collections Alexandra V. Shoffner1 and Margaret C. Brittingham1,* Abstract - Collections of bird specimens are an important resource for teaching bird identification, but acquiring suitable specimens can be problematic. Older collections tend to be preserved with a variety of potentially harmful chemicals; additionally, traditional methods for preparing specimens typically require extensive training. Freeze-drying is a method that involves removing water from specimens via sublimation, and may be an acceptable alternative to conventional taxidermy techniques for teaching collections. We freeze-dried 63 birds and 12 bird parts (i.e., talons and wings) of 44 species salvaged from throughout Pennsylvania since January 2008 using a Taxi-Dry Freeze-Dryer (Freeze-dry Specialties, Inc.). To determine the extent of water lost during the freeze-drying process, we measured the masses of birds and parts before and after preservation. Whole birds that were successfully freeze-dried lost 59.4% ± 0.9% (mean ± SE) of their initial mass, and unsuccessfully dried birds lost 46.9% ± 3.5% of their initial mass. Generally, birds with an initial mass >160 g did not lose enough water in the freeze-drying process to be effectively preserved. We conclude that if proper storage and maintenance conditions are met, freeze-drying can be an effective method for preserving small bird specimens for teaching collections. Introduction Collections of avian specimens are helpful, and perhaps even necessary, for teaching bird identification. Birds in the hand are more easily identified than in photographs or in the field (Remsen1995) and as such are a critical resource for teaching identification. Specimens provide hands-on experience allowing the student to see and compare morphology, color, and individual variation (Remsen 1995). Clearly, access to specimens is a priority for those teaching bird identification. However, obtaining or creating high-quality specimens that retain their identifying characteristics (e.g., plumage color and natural posture) can be time-consuming and expensive. Older specimens have typically been preserved with a variety of potentially harmful chemicals, many of which have been phased out in more recent years due to health concerns (e.g., Edolan U and arsenic and mercury compounds; Hower 1979). Preserving new specimens by conventional taxidermy methods can also be problematic because of the skill level required to prepare specimens. Freeze-drying is an alternative preparation method that holds potential for creating specimens for teaching collections. Freeze-drying is a process that dehydrates materials by sublimation, which is the direct transition of ice (solid phase) to water vapor (gas phase) (Meryman 1960). It has been used for preparing specimens used in educational displays and exhibits but is not generally 1Department of Ecosystem Science and Management, The Pennsylvania State University, University Park, PA. *Corresponding author - mxb21@psu.edu. A.V. Shoffner and M.C. Brittingham 2013 Northeastern Naturalist Vol. 20, No. 3 442 recommended for long-term preservation or for specimens that may be used in biochemical research because deterioration and degradation of cells and tissues occurs during and after the freeze-drying process (Florian 1990). Our objective was to evaluate the effectiveness of a commercially available freeze-drying system for freeze-drying bird specimens for use in a University teaching collection. We used freeze-drying to prepare specimens, and discuss the advantages and disadvantages of this method. Methods A basic freeze-drying system consists of a refrigerated specimen chamber, a refrigerated condenser or vapor trap, and a vacuum pump (Fig. 1; Hower 1979). Specimens are kept at low temperatures within a chamber, and a vacuum pump lowers the pressure within the chamber to facilitate sublimation and the removal of water vapor from the chamber. Water that sublimates from specimens moves to the condenser, or vapor trap, that is kept at a lower temperature than the specimen chamber. The optimal equipment and settings for freeze-drying vary slightly with the type of material being preserved. Systems can be purchased from suppliers, though there are also resources available for those who wish to build their own freeze-drying system (Hower 1979, Meryman 1961). We used a Taxi-Dry Freeze-Dryer system purchased from Freeze-dry Specialties, Inc. (Model ARA 1800 R.V.T.; Fig. 1). This system, which is actively marketed for freeze-drying vertebrate specimens, consists of an upright freezer with dimensions 1.65 m H x 0.8 m W x 0.6 m L (65” H x 32” W x 23.5” L) containing an approximately 0.1-m3 (3.5-ft3) chamber with two rows of shelves. A valid Federal Migratory Bird Special Purpose Salvage Permit an d a concurrent state salvage permit are necessary to salvage and possess bird carcasses. Bird carcasses were salvaged (State permit SAL00436, Federal permit MB028785-0) between 2008 and 2012 from two sources: private individuals from throughout Pennsylvania, and from a local wildlife rehabilitation clinic. The majority of birds died from trauma, such as collisions with windows or cars. Bird carcasses that could not be freeze-dried immediately were stored in a chest freezer set to -21 °C . Most birds were stored frozen for less than one year before freeze-drying, although one specimen was successfully freeze-dried after being stored for 13 years. We identified all birds to species level and aged and sexed them by plumage with the aid of Pyle (1997, 2008). Wings and talons were collected from large raptors to be freeze-dried separately. If a bird carcass had been frozen, it was thawed before we measured the wing chord and tail length with calipers, and the mass with a triple beam balance. Birds were configured into desired positions using pins, cotton, and string on Styrofoam and cardboard, and then frozen for at least 24 hours in a chest freezer set to -21 °C before beginning freeze-drying. Once the birds were completely frozen, we loaded them into the chamber (set to -20.5 °C) and turned the vacuum pump on to begin freeze-drying. 443 A.V. Shoffner and M.C. Brittingham 2013 Northeastern Naturalist Vol. 20, No. 3 The manufacturer-recommended freeze-drying schedule for vertebrates was to slowly increase the temperature from -20.5 °C (-5 °F) to 15.5 °C (60 °F) over 10 days. They provided a temperature chart which we modified to a 12-day schedule. The freeze-drying chamber was set to -20.5 °C on the first day, and the temperature was increased incrementally over 12 days to 15.5 °C to induce sublimation in the specimens. Sublimation is the transition from the solid state directly to the vapor state, bypassing an intermediate liquid state; i.e., during Figure 1. Freeze-drying schematic adapted from Hower 1979 (top) and our freeze-drying system (bottom). A = specimen chamber, B = condenser or vapor trap, C = vacuum pump valve, D = vacuum pump, and E = pressure gauge. A.V. Shoffner and M.C. Brittingham 2013 Northeastern Naturalist Vol. 20, No. 3 444 freeze-drying, ice within frozen specimens sublimes to water vapor without transitioning via liquid water (Hower 1970). Temperature and pressure determine the phase (solid, gas, or liquid) a substance takes. At different combinations of low pressure and temperature, sublimation occurs. By increasing the temperature gradually, water is slowly removed from the specimen. After a minimum of 12 days, we turned off the vacuum pump, removed the specimens, and recorded their mass. Rarely, specimens were freeze-dried for longer than 12 days (maximum 24 days) because we had limited access to the feeze-drying system on weekends and holidays. Success of freeze-drying was initially determined subjectively by the feel of the bird. Freeze-dried specimens tend to have a dry, brittle texture (Florian 1990). We initially classified a bird as unsuccessfully freeze-dried based on whether the bird felt dry, thin, and brittle or felt soft and flexible. In addition, birds that were unsuccessfully dried tended to leave oil marks when laid out on cardboard, suggesting they were leaking fat. For each bird and bird part, we calculated the mass before freeze-drying and the percent of mass lost after freeze-drying. For birds that were successfully freeze-dried, we calculated total mass lost and percent mass lost. We used a t-test to determine whether initial mass or percent mass loss differed between birds that were successfully freeze–dried and birds deemed unsuccessfully dried due to fat leakage. We regressed mass lost by initial mass to determine the relationship between the two. If specimens were deemed successfully freeze-dried, we placed them in storage. Because we were unsure of the vulnerability of these specimens to insect damage and their condition over the long-term, we avoided storing them in the cabinets with our collection of birds that had been preserved by traditional taxidermy. Instead, specimens were wrapped in cotton and stored within an airtight plastic container. Each specimen was labeled with a tag providing information on the specimen including species, age, sex when known, date collected, and location where collected. Specimens were periodically examined for evidence of fat leakage and insect damage. Results Between 2008 and 2012, we freeze-dried 63 birds and 12 bird parts (wings and talons), representing 44 species (Table 1). Groups of specimens were freezedried together with runs consisting of 4–14 birds or bird parts. Total mass per run ranged from 121.16 g to 1173 g (mean ± SE = 439 ± 110). Fifty-eight whole birds were successfully dried. Five specimens were classified as unsuccessful due to fat leakage or squishy texture. The 58 successfully freeze-dried specimens consisted of 40 different species ranging in mass from a 2.5-g Archilochus colubris (Ruby-throated Hummingbird) to a 147.8-g Colaptes auratus (Northern Flicker) (Table 1). One Tringa melanoleuca (Greater Yellowlegs) and four Corvus brachyrhynchos (American Crow) ranging in mass from 165 g to 595.5 g were not successfully dried and were discarded . 445 A.V. Shoffner and M.C. Brittingham 2013 Northeastern Naturalist Vol. 20, No. 3 Table 1. Bird specimens that were freeze-dried 2008–2012. Mass measurements of species with multiple representative individuals are listed as ranges. Initial % mass Species Number mass lost Success Whole birds Falco columbarius L. (Merlin) 1 125.0 61.2 Yes Falco sparverius L. (American Kestrel) 2 76.8–104.0 42.3–63.2 All yes Tringa melanoleuca (Gmelin) (Greater Yellowlegs) 1 165.0 41.4 No Chordeiles minor (Forster) (Common Nighthawk) 1 77.0 51.8 Yes Archilochus colubris (L.) (Ruby-throated Hummingbird) 1 2.5 64.0 Yes Picoides pubescens (L.) (Downy Woodpecker) 1 24.3 58.0 Yes Colaptes auratus (L.) (Northern Flicker) 2 112.0–148.0 55.0–66.5 All yes Sayornis phoebe (Latham) (Eastern Phoebe) 1 11.8 66.9 Yes Myiarchus crinitus (L.) (Great Crested Flycatcher) 1 28.2 65.6 Yes Cyanocitta cristata (L.) (Blue Jay) 2 67.0–82.5 62.2–63.7 All yes Corvus brachyrhynchos Brehm (American Crow) 4 298.0–596.0 36.8–56.6 All no Hirundo rustica L. (Barn Swallow) 1 13.8 58.0 Yes Tachycineta bicolor (Vieillot) (Tree Swallow) 2 11.1–13.7 62.0–62.2 All yes Poecile atricapillus (L.) (Black-capped Chickadee) 2 8.7–11.6 57.5–59.5 All yes Baeolophus bicolor (L.) (Tufted Titmouse) 1 19.2 53.1 Yes Certhia americana Bonaparte (Brown Creeper) 1 6.4 43.8 Yes Catharus guttatus (Pallas) (Hermit Thrush) 1 32.3 57.9 Yes Hylocichla mustelina (Gmelin) (Wood Thrush) 1 42.9 62.7 Yes Catharus ustulatus (Nuttall) (Swainson's Thrush) 1 28.9 54.3 Yes Turdus migratorius L. (American Robin) 1 73.8 65.9 Yes Dumetella carolinensis (L.) (Gray Catbird) 3 31.6–37.7 66.8–68.7 All yes Bombycilla cedrorum Vieillot (Cedar Waxwing) 5 21.8–38.6 62.4–66.7 All yes Seiurus aurocapilla (L.) (Ovenbird) 1 17.4 64.4 Yes Mniotilta varia (L.) (Black-and-white Warbler) 1 9.7 60.8 Yes Oreothlypis peregrina (Wilson) (Tennessee Warbler) 1 7.9 67.1 Yes Oreothlypis ruficapilla (Wilson) (Nashville Warbler) 1 8.6 62.8 Yes Geothylpis trichas (L.) (Common Yellowthroat) 4 6.0–9.8 53.8–63.3 All yes Setophaga magnolia (Wilson) (Magnolia Warbler) 1 9.4 59.6 Yes Setophaga castanea (Wilson) (Bay-breasted Warbler) 1 9.7 58.8 Yes Setophaga striata (Forster) (Blackpoll Warbler) 1 17.4 29.9 Yes Pipilo erythrophthalmus (L.) (Eastern Towhee) 1 46.9 56.7 Yes Passerella iliaca (Merrem) (Fox Sparrow) 1 39.6 60.2 Yes Piranga olivacea (Gmelin) (Scarlet Tanager) 3 24.8–34.5 56.4–63.7 All yes Cardianalis cardinalis (L.) (Northern Cardinal) 2 30.8–49.8 59.8–64.9 All yes Pheucticus ludovicianus (L.) (Rose-breasted Grosbeak) 1 41.8 63.9 Yes Passerina cyanea (L.) (Indigo Bunting) 1 10.9 54.1 Yes Icterus galbula (L.) (Baltimore Oriole) 1 32.3 60.6 Yes Carpodacus purpureus (Gmelin) (Purple Finch) 1 29.0 55.9 Yes Carpodacus mexicanus (Muller) (House Finch) 2 17.9–20.3 50.6–65.4 All yes Carduelis pinus (Wilson) (Pine Siskin) 1 13.8 58.7 Yes Coccothraustes vespertinus (Cooper) (Evening Grosbeak) 1 39.0 40.5 Yes Carduelis tristis (L.) (American Goldfinch) 1 11.5 55.6 Yes Total 63 Bird parts Raptor spp. (talons) 7 6.2–26.2 17.7–40.4 Yes Raptor spp. (wings) 5 25.4–78.6 23.4–40.0 Yes Total 12 A.V. Shoffner and M.C. Brittingham 2013 Northeastern Naturalist Vol. 20, No. 3 446 Whole birds that were successfully freeze-dried differed from unsuccessfully freeze-dried birds in both initial mass (t = 4.78, P = 0.009) and percent of mass lost (t = -3.45, P = 0.026). Initial mass of successfully dried birds was 33.7 ± 4.1 g (mean ± SE) as compared to 408.7 ± 78.3 g for unsuccessfully dried birds. For birds that were successfully freeze dried, mass loss varied significantly with initial mass (P < 0.0001, r2 = 96.6%, slope ± SE = 0.579 ± 0.014; Fig. 2). Birds that were successfully freeze-dried lost 59.5% ± 7.2% of their mass, and unsuccessfully dried birds lost only 46.9% ± 7.9%. Birds that were not adequately dried were included in three separate runs all of which also had birds that were successfully dried. Runs where all specimens were successfully dried ranged from a combined mass of 121.2 g to 503.4 g, while those that contained at least one individual that was not completely dried ranged in mass from 567 g to 1173 g. In one run, we had a 165-g Greater Yellowlegs that was not successfully dried while a 124.6-g Falco columbarius (Merlin) was successfully dried. All 12 bird parts (wings and talons) dried successfully. Wings (n = 5) had an initial mass of 51.5 ± 11.2 g and lost 32.5% ± 3.2% of their mass, and talons (n = 7) had an initial mass of 18.5 g ± 3.3 g and lost 30.0 ± 3.2% of their mass. All successfully freeze-dried birds have been free of pest damage and fat leakage since freeze-drying and subsequent storage. Figure 2. Mass loss of 58 successfully freeze-dried whole birds 2008–2012. Mass lost = 0.492 + 0.579 Initial mass (P < 0.001). 447 A.V. Shoffner and M.C. Brittingham 2013 Northeastern Naturalist Vol. 20, No. 3 Discussion There are few studies that present data on the percent of mass lost in birds throughout freeze-drying (e.g., Hower 1979; Meryman 1960, 1961), so it is difficult to pre-determine the percentage loss necessary to be assured that a specimen is completely freeze-dried. In most cases, freeze-drying is deemed complete when the specimen ceases to lose weight with additional days of drying (Meryman 1960). Meryman conducted periodic weighing throughout freeze-drying by maintaining a balance in a “deep-freeze unit” to prevent specimens from thawing while being weighed, and by assuming that breaking the vacuum would have no ill effects on freeze-drying (1960). With our system, we could not easily weigh birds on a daily basis to track change in mass because that would entail losing the vacuum seal and warming the birds and chamber up. However, our results suggest that successfully freeze-dried birds lose almost two-thirds of their mass (59.2%) on average, in contrast to unsuccessfully dried birds, which lost less than half of their mass (46.9%). This is in agreement with Meryman (1961) who reported mass loss of 59% and 62% for a Sturnus vulgaris (European Starling) and a Northern Flicker that he successfully freeze-dried. We were unable to freeze dry birds over 160 grams in our 12-day drying schedule. Although large birds were poorly represented (both in numbers and diversity) in our sample, this suggests that there is an upper limit to the mass of birds that could be freeze-dried by our system. Our results suggest that the mass of the individual bird may be more important than the combined mass of the birds within the run since we had small birds that were successfully freeze dried in runs with a combined mass of over 1000 g. There are many variables that affect the success of freeze-drying a specimen, such as the fat content of the specimen, the temperature and pressure of the freeze-drier, and the length of time a specimen is dried (Cumberland 1999, Hower 1979, Meryman 1960). The difference in our ability to successfully freeze dry the similarly sized Greater Yellowlegs and Merlin suggests there may be interspecific differences perhaps related to fat levels. Shorebirds tend to carry high levels of fat during migration perhaps making them harder to fr eeze dry. Others have successfully freeze-dried birds at least as large as Haliaeetus leucocephalus (L.) (Bald Eagle; Cumberland 1999), and other large animals such as an adult alligator (Hower 1979), suggesting that large animals can be freeze dried under proper conditions. In general, larger animals require a larger freeze-drying chamber, a higher-capacity vapor trap, and a longer drying time; for example, a Strix varia Barton (Barred Owl), (approximately 0.5 kg) required 130 days to complete freeze-drying, and the aforementioned alligator required a 5-m3 (177-ft3) chamber (Hower 1979). Though these lengths of time and chamber sizes are upper limits, it is clear that large specimens quickly become time- and cost-prohibitive to freeze-dry. In addition to freeze-drying, bird specimens can be prepared by conventional taxidermy or air-drying and stuffing study skins, a common museum preservation technique. Each technique has its own advantages A.V. Shoffner and M.C. Brittingham 2013 Northeastern Naturalist Vol. 20, No. 3 448 and disadvantages, and the choice of a technique will depend on the intended purpose of specimens and the resources available to the collector (Cumberland 1999, Meryman 1960). When choosing a preservation technique, factors to consider include the amount of training required to perform the technique, the cost of equipment, the authenticity and lifetime of the resultant specimens, and the need for chemicals for preservation. A major benefit of freeze-drying is that it requires no training outside of learning how to operate the equipment. Unlike conventional taxidermy and drying study skins, freeze-drying does not require cutting animals open or manipulating them internally in any way (Meryman 1960). However, this reduction in preparation time results in losing potentially critical information about the specimen, such as reproductive status and stomach contents, and so other methods may be preferable depending on information needs (Winker 2000). An additional benefit of freeze-drying is that, at least in the preparation stage, no chemicals are used to prepare the specimens. This can be particularly important if the specimens will be handled by students or members of the public who may have sensitivities to some chemicals and preservatives. Freeze-drying also produces authentic-looking specimens: colors are well-preserved, the body is undistorted by drying, and specimens can be easily manipulated in a range of positions before drying (Hower 1970, 1979). Though elaborate displays are feasible, the best positions for use in for teaching collections are those that emphasize the identifying physical characteristics of the bird. A disadvantage of freeze-drying is the upfront cost of equipment. Freezedrying equipment costs thousands of dollars, even for the smallest systems: in 1961, the “simplest possible” freeze-drying system retailed for $2000 (Meryman 1961); our Taxi-Dry freeze-drying system cost approximately $12,000 in 2008. Costs may be reduced by purchasing individual components or building a system from scratch, but this clearly requires more training and time than buying a system. However, freeze-drying may still be the most cost-effective method of producing specimens after considering the minimal training and preparation time required. In addition, cost-sharing a system among individuals working with different taxonomic groups or with nature centers or other organizations interested in developing teaching collections is an additional way to make it affordable. Over the long-term, durability of specimens may be a concern. One potential disadvantage of freeze-dried specimens is their brittleness (Hower 1979, Meryman 1960). Because freeze-drying removes the water and thus the elasticity of living tissue, specimens must be stored, transported, and handled with some care. Reducing the number of protruding parts before freeze-drying can reduce the risks of later damaging specimens. Freeze-dried specimens can have long lifetimes if they are stored properly (i.e., out of the reach of insects, predators, and UV light.) Because freeze-dried specimens generally do not contain chemicals to deter insects, they can be quite vulnerable to insect damage, particularly if they are improperly stored. Incompletely freeze-dried specimens are susceptible to 449 A.V. Shoffner and M.C. Brittingham 2013 Northeastern Naturalist Vol. 20, No. 3 tissue decay, fat leakage, and subsequent insect damage, and even completely freeze-dried specimens are vulnerable to insects, predation, UV damage, and mechanical damage. Predation, though uncommon, is a possibility; a gray squirrel once consumed four freeze-dried bird specimens at the Smithsonian (Meryman 1960), and one of our specimens (a Carduelis tristis [American Goldfinch]) was partially eaten when left unattended overnight, likely by a rat. There is also debate over whether rehydration is possible at high ambient humidity (Cumberland 1999, Meryman 1960). General storage recommendations are to keep specimens enclosed and protected as much as possible—e.g., wrapped in cotton (to prevent mechanical damage) within airtight plastic containers (to prevent insect damage, predation, and UV damage) in a climate-controlled room (to prevent potential rehydration) (Cumberland 1999). These precautions, along with frequent inspection of specimens for any problems, have worked well for our freeze-dried specimens. Additional precautions include storing freezedried specimens separately from specimens preserved by other methods, and re-freeze-drying specimens periodically (such as once a year) to prevent rehydration and insect attraction (Cumberland 1999). Though chemicals are not required to preserve freeze-dried specimens, professional taxidermists use insect-deterring chemicals to protect freeze-dried specimens that are intended for display (Cumberland 1999, Hower 1979). In summary, freeze-drying is most effective for preserving small bird specimens in cases where there is a ready source of specimens, time and training constraints are an issue, and the absence of chemicals, and color preservation are priorities. Freeze-drying is also an effective method for preserving other taxa, such as insects, fishes, reptiles, marine invertebrates, and small mammals (Hower 1979, Meryman 1960). Though larger specimens can be effectively preserved by freeze-drying with different chamber specifications and longer periods of time, we found that a 12-day drying cycle using a 0.1-m3 (3.5-ft3) specimen chamber was not sufficient to completely freeze-dry specimens with initial mass >160 g. Freeze-drying should also not be used if information regarding sex or reproductive status is desirable, because the dissection required for this data eliminates the convenience that is an advantage of freezedrying. In general, freeze-drying is also inappropriate for specimens with high fat content, such as waterfowl, as well as specimens that will be on display (i.e., exposed to sunlight and highly vulnerable to insect damage) such as in a museum setting. Finally, the equipment required for freeze-drying may be costprohibitive for smaller organizations. Despite the limitations of freeze-drying, it may be an acceptable or even preferable alternative to conventional taxidermy for those wishing to produce a teaching collection. Acknowledgments We thank Andrew Weber, Shannon Harding, Sarah Pabian, and members of the Avian Outreach class for assistance with freeze-drying. We also thank Centre Wildlife Care for A.V. Shoffner and M.C. Brittingham 2013 Northeastern Naturalist Vol. 20, No. 3 450 providing bird carcasses. Funds to purchase the freeze-dryer were provided by the School of Forest Resources at The Pennsylvania State University. Literature Cited Cumberland, B. 1999. Using freeze-dried animal specimens in exhibits. Cultural Resource Management 22:23–26. Florian, M. L. 1990. The effects of freezing and freeze-drying on natural history specimens. Collection Forum 6(2):45–52. Hower, R.O. 1970. Advances in freeze-dry preservation of biological specimens. Curator: The Museum Journal 13:135–152. Hower, R.O. 1979. Freeze-Drying Biological Specimens: A Laboratory Manual. Smithsonian Institution Press, Washington, DC. 196 pp. Meryman, H.T. 1960. The preparation of biological museum specimens by freeze-drying. Curator: The Museum Journal 3:5–19. Meryman, H.T. 1961. The preparation of biological museum specimens by freeze-drying: II. Instrumentation. Curator: The Museum Journal 4:153–174. Pyle, P. 1997. Identification Guide to North American Birds, Part I: Columbidae to Ploceidae. Slate Creek Press, Bolinas, CA. 732 pp. Pyle, P. 2008. Identification Guide to North American Birds, Part II: Anatidae to Alcidae. Slate Creek Press, Bolinas, CA. 835 pp. Remsen, J.V. 1995. The importance of continued collecting of bird specimens to ornithology and bird conservation. Bird Conservation International 5:145–180. Winker, K. 2000. Obtaining, preserving, and preparing bird specimens. Journal of Field Ornithology 71:250–297.