Partial replacement of an artificial nectar diet with native browse for feather-tail gliders (Acrobates pygmaeus) in captivity
Abstract
Captive-bred feather-tail gliders (Acrobates pygmaeus) housed at Taronga Zoo have had a long history of eye cholesterol plaques that may be associated with a largely sugar-based diet such as artificial nectar. The gliders also have prolonged periods of reduced activity when they are not visible in exhibits. This may be due to the ad libitum supply of an energy rich feed and reduced need to forage. This study examined behavioral and physiological changes associated with supplementing the high sugar-based diet with two species of native browse. The experiment was conducted over two consecutive periods of 3 weeks and consisted of two treatment groups: one group was offered the artificial nectar only, while the other group was offered the artificial nectar supplemented with a variety of native flowers. Live weight was recorded weekly. There was no change (P > 0.10) in artificial nectar intake with the supplementation of native browse in the diet. Blood metabolites (cholesterol, triglycerides, glucose) tested for the two groups had no differences (P > 0.10) between treatments. Upon examination, there were no signs of tooth decay or cholesterol plaques in all animals throughout the experiment. Feed intake and behavior were recorded via sensor cameras. There was an increase (P < 0.05) in the daily foraging activity of gliders supplemented with native flowers compared to gliders fed the artificial nectar alone. In conclusion, supplementing to provide a more native diet to A. pygmaeus enhanced their natural foraging behavior, suggesting that it may result in long-term improvements in their health. Zoo Biol. 32:394–399, 2013. © 2013 Wiley Periodicals Inc.
INTRODUCTION
Weighing only 10–14 g the feather-tail glider (Acrobates pygmaeus) is the smallest of the Australian marsupials belonging to the family Burramyidae [Ward, 1990]. It is restricted to mainland forests and woodland ranging widely along the eastern coast from Northern Queensland to Southern Australia [Ward, 1990].
Due to its' small size and the difficulties associated with regularly trapping this arboreal marsupial, little is understood about the ecological requirements of A. pygmaeus in comparison to other species [Goldingay and Kavanagh, 1995]. Furthermore, evidence of the nutritional importance of various dietary components is similarly scarce. In its natural habitat, the diet of A. pygmaeus is similar to that of the yellow-bellied glider (Petaurus australis) and squirrel glider (Petaurus norfolcensis) [Smith, Nagy, Fleming, and Green, 1982], with the diet consisting of nectar and pollen from a variety of native plant species; particularly banksias and eucalypts, manna (a clear carbohydrate-rich substance found on leaves of Eucalyptus sp.), honeydew (a sugar-rich secretion produced by insects) [Lindenmayer, 1997] and arthropods such as lerp insects, soft-bodied termites, or white ants [Goldingay and Kavanagh, 1995; Jackowiak and Godynicki, 2007]. Loose shedding bark, blossoms, and foliage provide the most important foraging substrates [Goldingay and Kavanagh, 1995; Turner, 1984; Huang, Ward, and Lee, 1987] and the use of a variety of floral species for nectar and pollen suggests that this small marsupial may serve as an important pollinator of many plant species in forest habitats [Quin, Goldingay, Churchill, and Engel, 1996]. The special arrangement of lingual papillae on the elongated tongue of A. pygmaeus and its small, agile body make this glider strongly adapted to its specialized nectar-based diet [Jackowiak and Godynicki, 2007].
Observational studies coupled with fecal and stomach content analysis have previously been used to examine the diet and foraging behavior of A. pygmaeus in the wild [Goldingay and Kavanagh, 1995; Smith, 1980; Turner, 1984; Huang et al., 1987]. However, the nutritional requirements of this species in captivity have not been explored and a thus complete understanding of A. pygmaeus' nutritional requirements is lacking. Observations have revealed that these small arboreal marsupials depend predominantly on nectar and pollen to satisfy their dietary needs [Jennings, 2004]. Nectar is an essential energy provider and pollen is a rich source of protein containing trace vitamins and minerals that are vital for mammalian nutrition [Jennings, 2004]. Seeds and insects also provide a rich source of protein and polyunsaturated fatty acids [Geiser, Stahl, and Learmonth, 1992].
Patterns of activity have been previously described for free-living A. pygmaeus [Fleming, 1985; Geiser et al., 1992]. Differences in daily activity between gliders bred in captivity and field individuals [Geiser and Ferguson, 2001] may be partially explained by diet and the different foraging costs associated with the availability of food items [Geiser and Ferguson, 2001]. The aim of this study was to investigate the influence of diet on the activity of A. pygmaeus in captivity.
Acrobates pygmaeus housed at Taronga Zoo, Sydney have experienced a long history of eye cholesterol plaques that may be associated with a sugar-based diet of artificial nectar. The ad libitum supply of an energy rich feed and the reduced need to forage may also explain prolonged periods of reduced activity exhibited by gliders not visible in exhibits. This study examined several characteristics of nutrition affecting A. pygmaeus in captivity. Aiming to determine the extent to which native browse can improve the health and visibility of gliders in the exhibit and whether there are any behavioral or physiological impacts associated with the partial replacement of their artificial nectar diet with native plants. It was postulated that A. pygmaeus would be more visible in exhibits (e.g. increase activity) when native browse partially replaced their artificial nectar diet due to increased foraging activity and that the addition of native browse to the diet would improve their health by lowering the overall sugar content of the diet. The objectives of this study were to determine if native species of browse given fed to A. pygmaeus would change feeding behavior, to determine the percentage of the artificial nectar diet that can be replaced by native browse, and to determine the physiological impacts of browse addition to the diet.
METHODS
Experimental Design
The experiment included two groups of four feather-tail gliders (all males of approximately 10–14 g in weight). The experiment consisted of two consecutive 3-week periods, with the first week of each experimental period consisting of an adaptation period, whereby the gliders were adapted to their new diets, followed by 2 weeks of observations. Observation occurred 3 days per week via direct visual and sensor-trigger video recording. In the first experimental period, four gliders were given the artificial nectar diet (Diet A), while the other four gliders received the artificial nectar diet supplemented with a variety of native flowers (Diet B). Native flowers included crimson bottlebrush (Callistemon citrinus) and grevillea cultivars, including Moonlight (Grevillea banksii × G. whiteana) and Robyn Gordon (G. banksii × G. bipinnatifida). The amount of artificial nectar (e.g. 20 ml) offered was held constant between the two groups of animals and the amount eaten in each group was measured daily during the second and third weeks in each period. In the second experimental period, the diets of the two groups were exchanged. That is the gliders that were given the Diet A in the first experimental period were switched to Diet B, while the animals given the Diet B were switched to Diet A during the second experimental period. Each glider was monitored closely for any behavioral and physiological changes throughout the study period.
Diets
The Diet A consisted of an artificial nectar mix combining 875 g water (35.5%), 1,200 g honey (48.6%), 210 g hard-boiled egg whites (8.5%), 150 g original bran cereal (6.1%), and 32 g Sustagen (Nestlé Australia Ltd Notting Hill, Victoria, Australia) (1.3%). The Diet B (e.g. artificial nectar) supplemented with two native flowers (∼150 g) per animal. Both treatment groups were given one chunk of cantaloupe and one chunk of sweet potato for behavioral enrichment. The artificial nectar availability was not reduced for animals on Diet B so that feed preferences could be determined and to ensure that an adequate amount of palatable feed was available. All other dietary components remained constant during the study period. The native flowers provided were either fresh (i.e. collected that day) or refrigerated for a maximum of 3 days. Feeding times did not vary significantly from the routine feeding regime.
Feed Composition
Weekly samples of artificial nectar mix, C. citrinus and Grevillea cultivar flowers, cantaloupe and sweet potato were collected for chemical analysis. Dry matter (DM) content of the flower, cantaloupe and sweet potato samples was determined via oven drying at 55°C to constant weight. Artificial nectar mix was dried at 100°C for 7 days due to its high viscosity. Duplicate samples of artificial nectar mix, C. citrinus and Grevillea cultivar flowers, cantaloupe and sweet potato were analyzed for neutral detergent fiber (NDF) using Van Soest, Robertson, and Lewis [1991] procedures modified for an Ankom 200/220 Fiber Analyzer (Ankom® Technol. Corp, Fairport, NY, USA), with both heat-stable α-amylase and sodium sulphite included in the analyses. Concentrations of NDF were expressed inclusive of residual ash. Ash content was determined after ignition for 2 hr at 600°C in a muffle furnace using method 942.05 [AOAC, 2006]. Samples were re-ground using a ball grinder (Mixer Mill MM2000, Retsch, Haan, Germany) for determination of nitrogen (N) by combustion using method 990.03 [AOAC, 2006]. Crude protein (CP) was calculated as N × 6.25. Ether extract (EE) content was determined by extraction with diethyl ether [method 920.39; AOAC, 2006] using an Ankom XT10 Extraction System (Ankom® Technol. Corp.). For water-soluble carbohydrate (WSC) analysis, approximately 0.25 g of ground feed sample was extracted with a 10 ml of (0.2%) benzoic acid–water solution on a reciprocal shaker for 1 hr, then filtered. Filtered extract was analyzed in a flow injection analyzer (Fluidquip Australia, Blaxland, NSW) using the alkaline ferricyanide decoloration method: (1) extracted carbohydrates were hydrolyzed to invert sugars by 1 N HCl at 90°C; (2) inverted sugars were then dialyzed into an alkaline stream of potassium ferricyanide at 90°C; (3) then read the absorbance at 420 nm. The yellow colored ferricyanide in the presence of inverted sugars converts to colorless ferrocyanide. The decreased color is directly proportionate to the amount of WSC in the solution.
Feed Intake
The two treatment groups were provided with three nectar feeders each containing 20 ml of artificial nectar mix daily. During the last 2 weeks of each experimental period the amount of nectar left in the feeders was measured daily using a syringe and the average daily nectar intake per animal (mg/animal day) was then calculated. During the last week of each experimental period, the cantaloupe and sweet potato chunks were weighed daily using digital scales (before and after provision to each treatment group) and the average daily fruit intake per animal (mg/animal day) was then calculated. Intake of natural flower exudate per animal could not be directly measured due to the particular feeding behavior of this exudivorous species, although video and direct visual observations were used to record feed preferences.
Feeding Behavior
Behavioral observation was conducted 3 days/week during the last 2 weeks of each experimental period, with sensor-trigger video recording predominantly used. Direct visual observations were also carried out 1 day/week during the 2-week observational period to support video data. The activity of gliders was recorded for each treatment group and represented the number of animals seen out at any one time during each interval of video recording (one interval lasted 60 sec). The number of video intervals recorded for each group was determined by the number of times an animal would trigger the sensor to start recording. The number of times (i.e. frequency) gliders visited nectar feeders was also recorded for each treatment group and, for the group offered Diet B, the frequency of visits to C. citrinus and Grevillea flowers was recorded. Behavioral traits observed from period 1 and 2 were analyzed to explore any diet effect and other feed preferences. A total of four different measures (artificial nectar, C. citrinus, Grevillea spp., foraging activity) were available for analysis.
Blood Parameters
In each period, gliders were, anaesthetized with isoflurane in oxygen administered via a face mask, placed in dorsal recumbency and had ≤0.1 ml of blood collected from the ventral tail vein with a heparinized 30 U insulin syringe. Blood was alternatively collected from the jugular vein if blood could not be obtained from the tail vein. The harvested blood was analyzed immediately via Reflotron using 28.5–31.5 µl of undiluted blood per reagent strip to measure cholesterol, triglycerides, and glucose concentrations (in order of priority).
Live Weight
While anaesthetized the gliders were also weighed, and their teeth were examined to assess dental status and their eyes checked for cholesterol plaques. Animals were weighed in bags using digital scales prior to the start of the project and once per week during the experiment. The same scales were used each time to ensure consistency. Weekly variation in live weight (g) was recorded for each individual to confirm that an adequate body weight was maintained within a 10% critical limit (1–1.4 g).
Statistical Analysis
Regression analyses were used to test the effect of the flower diet on the three blood parameters (cholesterol, triglyceride, glucose), body weight, feeding behavior (intake/activity) and total food intake (nectar, sweat potato, cantaloupe). To eliminate period effects and therefore allow the interpretation of the diet effect across both periods, the data were analyzed using Restricted or Residual Maximum Likelihood analysis in Genstat (13th edition), giving adjusted means. Different models were used to account for the trait distribution (or variability between gliders) where physiological measures, body weight and fruit/nectar intakes were approximately normally distributed and the behavioral measures were counts. Also, for physiological measures and body weight, mixed models were used to account for the gliders as a random effect, while other traits were analyzed assuming repeated models. Parameters were set to account for unbalanced observations within the experiments. Diet (n = 2) and period (n = 2) were included as fixed effects in the model and the Wald statistic was used to estimate the effect of the fixed factors. Differences among means were tested using the least squares mean linear hypothesis test with significance declared if P < 0.05.
RESULTS
Feed Composition
The chemical composition of the feeds is shown in Table 1. C. citrinus and Grevillea spp. flowers were higher in CP content (11.1% and 15.7% higher, respectively) compared to the nectar mix. Artificial nectar mix and cantaloupe were considerably higher in water-soluble carbohydrates (WSC) (above 24%); EE content in the DM was 37.1% higher in C. citrinus compared to Grevillea spp. and nectar mix. Comparative analyses of the nutritional quality (i.e. nectar content) of fresh and stored flowers revealed that the concentration of WSC in the DM were slightly higher in the fresh flowers compared to those that had been stored in a refrigerator for three consecutive days (15.0% and 13.1%, respectively). Furthermore, flowers collected in the morning contained slightly more WSC in the DM compared to those that were collected in the afternoon (14.7% and 13.4%, respectively).
| DM-% as fed | % in the DM | |||||
|---|---|---|---|---|---|---|
| CP | WSC | NDF | Ash | Fat | ||
| Artificial nectar | 44.5 | 4.8 | 24.3 | 14.9 | 0.7 | 2.2 |
| Callistemon citrinus | 24.0 | 5.4 | 16.4 | 38.3 | 4.6 | 3.5 |
| Grevillea | 24.3 | 5.7 | 16.6 | 36.1 | 2.9 | 2.2 |
| Cantaloupe | 10.7 | 7.4 | 25.1 | 12.2 | 13.2 | – |
| Sweet potato | 36.2 | 0.3 | 20.8 | 18.0 | 2.5 | – |
- DM, dry matter; CP, crude protein; NDF, neutral detergent fiber; WSC, water-soluble carbohydrates.
Feed Intake
For experimental periods 1 and 2, intake of nectar, sweet potato and cantaloupe was similar (P ≥ 0.532) between control and flower diets (Table 2). However, in both treatment groups, intake of sweet potato and cantaloupe decreased (47.2% decrease P = 0.014, and 49.4% decrease P = 0.005, respectively) in period 2 compared with period 1 (Table 3). There was no period effect for artificial nectar intake (P = 0.160).
| Treatment | Foraging activity (%) | Nectar intake (mg/animal day) | Cantaloupe intake (mg/animal day) | Sweet potato intake (mg/animal day) |
|---|---|---|---|---|
| Diet A | 65.6b | 30.5 | 4.81 | 4.21 |
| Diet B | 75.4a | 30.3 | 4.95 | 4.39 |
| SEM | 2.69 | 1.35 | 0.93 | 0.72 |
| P-value | <0.01 | 0.90 | 0.53 | 0.86 |
- Diet A, gliders offered artificial nectar only; Diet B, gliders offered artificial nectar supplemented with flowers; SEM, standard error of the means. a and b within a column, mean values are significantly different (P < 0.05).
| Period | Cantaloupe intake (mg/animal day) | Sweet potato intake (mg/animal day) | Callistemon citrinus intake (no. visits/day) | Grevillea spp. intake (no. visits/day) |
|---|---|---|---|---|
| 1 | 4.65a | 5.67a | 35.8a | 21.6b |
| 2 | 3.79b | 2.93b | 19.6b | 29.3a |
| SEM | 0.72 | 0.82 | 1.46 | 1.42 |
| P-value | <0.05 | 0.01 | 0.02 | 0.03 |
- SEM, standard error of the means. a and b, mean values are significantly different (P < 0.05).
Feeding Behavior
The frequency of visits made to artificial nectar was similar (P = 0.087) between diets. This result was related to the insignificant change to nectar intake when there was supplementation of native browse (Table 2). However, there was a 13% increase (P < 0.001) in the daily foraging activity of gliders supplemented with the flower diet (Diet B) compared to the Diet A (Table 2). Furthermore, within group offered Diet B there were intake differences of C. citrinus and Grevillea spp. between different days of each experimental period emphasizing changes in flower preference. For period 1, there was a greater (P < 0.001) intake of C. citrinus compared to Grevillea spp. (mean camera triggers 35.8 and 21.6, respectively). In period 2, Grevillea spp. were visited more frequently (P < 0.001) compared to C. citrinus (mean camera triggers 29.3 and 19.6, respectively; Table 3).
Blood Parameters and Live Weight
Blood cholesterol, triglycerides, glucose, and body weight did not change when flowers were included in the diet (P ≥ 0.121; Table 4). There was no correlation between glucose and body weight in general (r = 0.44; P < 0.05).
| Treatment | Cholesterol (mmol/L) | Triglycerides (mmol/L) | Glucose (mmol/L) | LW (g) |
|---|---|---|---|---|
| Diet A | 3.39 | 1.74 | 7.70 | 12.1 |
| Diet B | 3.42 | 1.55 | 6.69 | 11.7 |
| SEM | 0.12 | 0.05 | 0.49 | 0.42 |
| P-value | 0.36 | 0.99 | 0.12 | 0.69 |
- LW, live weight; SEM, standard error of the means; Diet A, gliders offered artificial nectar only; Diet B, gliders offered artificial nectar supplemented with flowers.
DISCUSSION
Compositional analysis of feed components offered to gliders in this study indicate that native floral species, C. citrinus and Grevillea spp., are important sources of energy containing relatively large amounts of readily digestible carbohydrates and good sources of protein for growth and development. They also contain important trace minerals important for normal body functioning, physiology and development [Smith and Russell, 1982; Ward, 1990; Quin et al., 1996].
Throughout both experimental periods an increase in daily foraging activity was observed for gliders supplemented with native flowers. The lesser foraging activity observed in the gliders given the artificial nectar diet alone (e.g. Diet A) may be explained by different foraging costs associated with the availability of food items as suggested by previous investigations on the effect of diet on physiology and behavior between captive-bred and wild individuals [Geiser and Ferguson, 2001; Geiser et al., 1992]. In the wild, resources are either well dispersed (e.g. arthropods) or highly clumped (e.g. sap, honeydew, manna, nectar) and therefore gliders need to devote a large amount of time to foraging [Goldingay, 1989; Goldingay, Quin, and Churchill, 2001]. Smith et al. [1982] remarked that the cost of foraging places a significantly greater energy demand on these marsupials than any other life process. A diet rich in high-energy soluble carbohydrates [Holland, Bennett, and van der Ree, 2007] accommodates this demand as it can be rapidly digested, thus allowing gliders to feed for extended periods [Goldingay, 1989]. In captivity, however, the feeding of ad libitum artificial nectar, combined with a reduced need to forage, has led the increase in glider activity and natural foraging behavior observed within this study is likely attributed to the addition of native browse to their diet. This has positive long-term implications for the health and visibility of A. pygmaeus in captivity.
Results also suggest that there was no effect on the artificial nectar intake of Acrobates pygmeaus as a result of native browse supplementation. Nectar intake was related to the amount of times gliders visited the nectar feeders and, for both experimental periods, neither the artificial nectar intake (nor visits to a nectar feeder) changed with the addition of flowers to the diet. Thus, no health benefits could be determined on the basis of the blood parameters tested (cholesterol, triglycerides, glucose). Bodyweight was similarly unaffected by native browse supplementation. No correlation between glucose and body weight was observed, with the fluctuations observed considered typical for an animal with such a high metabolic rate. Ward [1990] acknowledged that adult body weights fluctuate in response to feed abundance, with blood glucose level (mmol/L) the most obvious indicator of this.
Natural pollen and nectar preferences have not yet been documented for A. pygmaeus. Evidence suggests that dietary requirement is a strong factor influencing feed choice [Jennings, 2004] as many arboreal marsupials feed selectively on resources that are nutritionally important in their diet [Jennings, 2004]. It has been suggested that the nitrogen content of certain plant species favored by arboreal marsupials may be higher than that of other plants [Fægri and van der Pijl, 1979], with these plants thought to produce greater quantities of high quality nectar [Turner, 1982; Wooller, Russell, and Renfree, 1984]. Consequently the quantity and quality of nectar from different floral species can vary significantly, with some native browse species considered to be of higher nutritional value than others. Acrobates pygmaeus often overcome dietary deficiencies by accessing multiple floral species [Somerville, 2000]. In the wild, the climate of a region largely dictates flowering frequency and floral species diversity, thus nectar and pollen may be abundant at different times of the year [Somerville, 2000]. In theory, one would expect the nectar content of flowers harvested in this study to be affected by a variety of factors including collection time, and storage length and temperature. Analysis of WSC content in the DM suggested that nectar content was slightly greater for flowers collected in the morning compared to flowers gathered in the afternoon, which may be attributed to the overnight accumulation of nectar in flowers and its removal during the day by nectar-feeding animals. Furthermore, freshly collected flowers had a higher nectar content than flowers that had been stored at 4°C for three consecutive days.
CONCLUSION
Understanding an animal's diet is fundamental to understanding its ecology and hence its management in captivity. This study has shown that the natural foraging behavior of captive A. pygmaeus can be enhanced by supplementing its diet with a natural feed (i.e. native browse). Furthermore providing such supplementation may improve the health and visibility of animals in the exhibit and thus may be beneficial in the long term. No difference in artificial nectar intake was detected between control and flower treatments. As a consequence no significant difference was detected in blood analyses between treatments. No signs of tooth decay or ocular cholesterol plaques were observed in any animal. Evidently future research is drastically needed in specific regard to A. pygmaeus focused digestibility trials, combining fecal and stomach analysis with behavioral observations. This will allow for a more complete understanding of the importance of the various dietary components required by A. pygmaeus and shed light on the implications of their supplementation in captivity.
ACKNOWLEDGMENT
This project was approved by Taronga Zoo Animal Ethics (AEC protocol number 4e/03/11—partial replacement of nectar diet with native browse for feather-tail and squirrel gliders).
