Zou et al, 1993

a DMI, DM intake; DM, dry matter; NDF, neutral detergent fibre; ADF, acid detergent fibre; HC, hemicellulose; Cell, cellulose; h Proportion of bamboo (leaves, branches and culm) consumed in mixed ingredient diets expressed as percentage of dietary DM.

where DMintake is the mean dry matter intake (g day-1), and DMfaeces is the average quantity of undigested dry matter (g day-1) excreted in faeces (Van Soest, 1994).

Lignin, a polymerised, phenolic compound, is a highly indigestible component of the plant cell wall that has been used as an undigested marker to indirectly calculate food digestibility (Fahey & Jung, 1983; Van Soest, 1994). For example, when using lignin to indirectly estimate apparent digestibility (%) of dietary dry matter (ADDM), the algebraic expression is:

where Lfeed is the concentration (%) of the indigestible marker (i.e. lignin) in the feed, and Lfaeces is the concentration (%) of the indigestible marker in the faeces (Robbins, 1993).

Apparent digestibility (%) of a given nutrient (ADn utr) or component (e.g. protein or hemicellulose) can also be calculated using the indirect marker method:

ADNutr(%) = [1 - (Lfeed x Nutrfaeces)/(Lfaeces x Nutrfeed)] x 100

where Nutrfeed is the concentration (%) of the nutrient or component measured in the feed, and Nutrfaeces is the concentration (%) of the nutrient or component measured in the faeces. Lfeed and Lfaeces are the concentrations (%) of lignin in feed and faeces, respectively (NRC, 1993). Unfortunately, some of the lignin, although not digested, may not be recoverable from faeces. Thus, an error can be introduced into indirect estimates of apparent digestibility.

Long et al. (2004) estimated apparent digestibility of dry matter, hemicellulose and cellulose of bamboo by free-ranging giant pandas in the Qinling Mountains. Daily food intake was predicted from direct observations, an estimate of feeding rate (food mass per time unit) and the percentage of leaf material in foods consumed. Estimates of intake and lignin concentrations in foods consumed were used to estimate lignin intake. Lignin excretion was estimated by determining lignin concentrations in faeces produced by the focal animal. Based on these estimates and using the algebraic relationships described above, Long et al. (2004) estimated dry matter digestibilities to be 12.7 to 17.2% (see Table 6.3). Hemicellulose and cellulose digestibilities were estimated to be 15.2 to 22.4% and 0.5 to 11.0%, respectively.

Schaller et al. (1985) provided another estimate of bamboo dry matter and hemicellulose utilisation in free-ranging pandas in the

Wolong Nature Reserve. Using similar indirect methods of estimating total food intake, distribution of leaves, branches and culm consumed, and collection of faecal boluses, these authors calculated an average dry matter digestibility of 17% (range 12.5-23.3%; see Table 6.3). Calculated hemicellulose digestibility averaged 22% (range 18.2-26.0%; see Table 6.3).

Although a captive animal environment provides the level of control needed to conduct direct digestion trials, investigators have been reluctant or unable to offer such animals single ingredient diets (e.g. only bamboo) for evaluation. Assessment of bamboo digestibility extrapolated from animals fed mixed-ingredient diets is problematic at best. Such measurements assume that there is no interaction from the consumption of supplemental foods (e.g. vegetables, meat or dairy products), and that the digestibility of the basal component (i.e. bamboo portions consumed) is unaltered by the added supplement. Interactions or associative effects among feeds are common and often lead to under- or over-estimates of food digestibility (Van Soest, 1994).

Digestibility of dry matter by captive giant pandas fed mixed-ingredient diets with varying proportions of bamboo range from 38.3 to 67.4% (see Table 6.3), significantly higher than estimates made for animals under field conditions. Digestibility of hemicellulose shows a similar upward trend (28.2 to 56%; see Table 6.3). The differences observed between captive and wild populations are most likely due to variations in diets, levels of activity, methodology or a combination of these three factors. Although estimates made under field conditions include methodological limitations (particularly on intake), it appears that the giant panda has adapted to a diet primarily of bamboo. However, only 25% or less of the dry matter within that bamboo can be digested. Although the captive giant panda appears able to use more dietary dry matter when offered bamboo in the presence of other foods, the physiological impact of these supplements on gastrointestinal function and health has not been critically evaluated.

The ability of giant pandas to digest bamboo is limited by the rapid rate of digesta transit through the gastrointestinal tract. Several measures can be used to assess the time digesta spends in the tract, subject to the processes of mechanical mixing, digestion, microbial fermentation and absorption. When an inert digesta marker is fed, the time elapsing from ingestion to first appearance of the marker in the faeces is termed transit time (TT1). Although this measure is useful, mean retention time (Rgit) is more descriptive of the lag following initial marker excretion (TT1) and total digesta turnover. Mean retention time is determined as follows:

Rgit = [S(Yi x Ti)/SYi] - TTi where Yj is the concentration of the marker at a given time and Tj is the time, since marker ingestion (Blaxter et al, 1956).

Transit time for a giant panda in captivity that consumes a diet of at least 50% bamboo ranges from 6 to 7 hours (Table 6.4) (Dierenfeld et al., 1982; Edwards, 2003). A similar result (mean 7.9 hours) was reported for another captive giant panda, although diet details were not provided (Schaller et al, 1985). Rapid digesta passage in pandas is not conducive to microbial fermentation of digesta, and cellulose digestibility is not significantly above zero (Van Soest, 1994). These observations are supported by low populations of obligate anaerobic bacteria in faeces (Hirayama et al, 1989).

When compared to other herbivores, the physical characteristics of individual faecal boluses produced by the captive giant panda suggest limited mixing of digesta in the gastrointestinal tract. We have measured the extent of mixing using two differently coloured acetate bead markers. Each marker was fed in a single pulse bolus, one at time 0 and the second three hours later. Two animals, whose bamboo intake was greater than 80% of diet dry matter, had no overlap of the two markers in the excreta. A single animal, whose bamboo intake was less than 50% of diet dry matter excreted 78% of the two markers in the same faecal bolus (Edwards, 2003). These results suggest that, unlike other herbivores, the giant panda passes digesta in a single pulse, or wave, through the gastrointestinal tract. This adaptation may have arisen from the consumption of a homogeneous diet, both in type and nutrient content. Implications for captive animal husbandry and feeding are addressed below.


Clearly, bamboo is the predominant food source for the giant panda (Sheldon, 1937; McClure, 1943; Wang & Lu, 1973; Schaller et al, 1985; Carter et al, 1999). Species of bamboo used by free-ranging giant pandas vary with habitat. Pandas in the Qinling Mountains, where nine bamboo species belonging to five genera have been described, consume Bashania fargesii and Fargesia spathacea preferentially (Pan, 1988; Long et al, 2004). These two genera are the dominant native bamboos in the

Table 6.4. Digesta mean (± SEM) transit times (TTJ and mean retention times (Rsit) (how's) of diets fed to the red panda (Ailurus fulgens) versus giant panda


DMI (% BW)h




Marker TTj



Red panda

0 0

Post a comment