Levels of tannins in fruit diet of grey-cheeked mangabeys ( Lophocebus ugandae , Groves ) in Lake Victoria Basin forest reserves

Tannins are known to have anti-digestive properties among primates and undoubtedly influence food selection among grey-cheeked mangabeys (Lophocebus ugandae) living in Mabira and Lwamunda forest reserves. During the 24 months study, we observed the feeding behaviour of three groups of mangabeys in forests with different levels of degradation and management histories. Results indicated that mangabeys selected fruits parts based on tannin levels. Chemical analysis indicated that fruit parts with tannin levels exceeeding 2.6g/100 catechin equivalent were invariably rejected. These were unripe fruit pulp and mature seed. Accepted parts included endosperms of immature fruit and ripe pulp belonging to 12 priority tree species. This selection criteria applied to all types of fruit regardless of forest habitat. The level of rejection of fruit parts with high levels of tannins was more frequent in the severely degraded Lwamunda forest reserve than in Mabira. Consumption of endosperms of unripe fruit appears to be related to fruit scarcity.


INTRODUCTION
Grey-cheeked mangabeys living in Mabira and Lwamunda forest reserves are faced with spatio-temporal food scarcity.This is due to anthropogenic activities which have resulted in a reduction of potential food tree density.To meet their daily dietary requirements, mangabeys of the study area are compelled to consume food items known to have high levels of secondary compounds and specifically tannins.These are watersoluble and complex polyphenolic secondary metabolites with anti-nutritional properties (Kraus et al., 2003).The plants manufacture them for protection against herbivory (Kraus et al., 2003;Coley, 1986) and potential frugivores (Reynolds, 2005).Tannins are sometimes called antifeedants and have anti-digestive properties which they achieve by binding to proteins in the diet of animals and making them inedible.Some tannins (condensed or hydrolyzable) react with proline-rich proteins secreted in saliva to form a complex that is usually insoluble and unabsorbable (Lua and Bennick, 1998) or more difficult for animals to digest (Norconk and Veres, 2011;Robbins et al., 1987); hence protein digestibility is highly compromised.Too much tannin in the gut leaves an animal unable to digest its food properly and in the case of some species it can be fatal (Wrangham et al., 1998) or destructive (Erikasson and Ehrlén, 1998).
Cercopithecine monkeys have developed effective internal mechanisms for detoxification of secondary compounds (Reynolds, 2005) although it may be too expensive, sometimes as costly as reproduction in some mammals (Sorensen et al., 2005).External mitigation measures include: avoidance and regulation (Sorensen et al., 2004).Biotransformation entails elimination of PSMs from the body by excretion or chemical change or a combination of these processes (Schwedhelm et al., 2003;Wiggins et al., 2003).In nature, secondary compounds or metabolites appear to play a major role either in restricting the palatability of the plants in which they occur or in causing animals to avoid the plants altogether.Frugivorious primates avoid consumption of some toxic fruits (Herrera, 1982).The consumption of unripe fruits is a regulated trade-off mechanism (Schaefer et al., 2003) although it exposes mangabeys to plant secondary metabolites (PSMs), with a wide range of physiological effects: from direct toxicity to digestion impairment (Dearing et al., 2005).The regulation of PSM intake ensures that low potency toxins do not cause damage to the primate.One of the strategies used by primates to tregulate PSM is the development of salivary proline-rich proteins (PRPs) as a defence against tannins by forming complexes with them and thereby preventing their interaction with other biological compounds and absorption from the intestinal canal.The consequences of binding tannins with salivary proteins are twofold.Tannins are inactivated as toxins, enabling the animal to eat tannin-rich browse and permeability glycoprotein and cytochrome P4503A regulated absorption of PSMs by gut cells (Sorensen et al., 2004).Another mechanism used by mangabeys to regulate PSMs intake is by trade-offs between macro-nutrients and PSMs (Schaefer et al., 2003).Pre-ingestive mechanisms where taste and trigeminal stimulation are applied can also be used to regulate intake of PSMs.Foods can be repellent to animals by irritating trigeminal nerves in the mouth and causing a burning sensation.If it is too intense, food consumption is reduced.Animals use trigeminal feedback to regulate their intakes of these irritant PSMs for example, Jakubas et al. (1993) showed that coniferyl benzoate (CB) was a trigeminal irritant in birds and that ruffed grouse could regulate their intake of CB in aspen.Compounds that are intensely bitter like phenolic glycoside may also be effective deterrents at high concentrations and lead to thresholds in the ingestion of the PSM which can constrain feeding rates of mammalian herbivores (Dearing et al., 2005).PSMs can be eliminated from the body by biotransformation of lipophilic PSMs (Boyle et al., 1999) prior to excretion through urine.Terrestrial animals have evolved a powerful suite of biotransformation enzymes to convert lipophilic PSMs into more polar metabolites readily excreted in urine or bile.These enzymes are considered to have evolved in part as a response to dietary plant toxins (Gonzalez andGelboin, 1994, Gonzalez andNebert, 1990).
Unripe fruits are usually well protected chemically with secondary metabolites similar to those found in leaves.Most unripe fruit have mechanical protection as they are hard to crack and may contain sclerified cells or calcium oxalate crystals which are known to irritate humans and other mammals (Gargiullo and Stiles, 1993).The concentrations of some secondary compounds may change during fruit ripening (Barnea et al., 1993) but while they last, they may limit the quantity of fruits consumed and length of a single foraging bout thereby prompting frugivores to forage elsewhere.As the chemically and mechanically well defended fruits ripen, the chemical defence reduces but the qualitative nature of phenolics does not change in a ripe fruit to compensate the over-all quantitative loss (Gargiullo and Stiles, 1993).
According to Mack (2000), tannins are also found in ripe fruits of broad and diverse array of plants.Generally, most food resources with high PSMs are not only toxic but are also poor in quality and low in nutrients (Coley, 1986).Consequently, some frugivores may switch foraging patches because of the defensive role of secondary compounds.This natural selection has favoured animals that avoid tannin-rich foods (Reynolds, 2005).During times of scarcity mangabeys seem to either develop tolerance to higher levels of tannins (Chapman et al., 2012;Doran-Sheehy et al., 2006;Wrangham et al., 1998) or engage in crop-raiding (Ogango, 2006).Raided food crops are selected on account of high digestibility, high energy content and low concentration in secondary compounds (Rode et al., 2006).
We hypothesized that plant secondary compounds exemplified by tannins, determine fruit parts fed on by mangabeys in Mabira and Lwamunda forest researves.We tested this hypothesis by analysing the tannin levels of fruit parts eaten by three social groups of magabeys living in three forests with varying levels of degradation.

Study area
We studied grey-checked mangabeys in Mabira (306 km 2 , between 0° 24`and 0° 35`North latitudes and 32° 52` and 33° 07`East longitudes) and Lwamunda (67 km 2 , between 0 o 15′ and 0º 23′ North latitudes and 32 o 30′ and 32º 43′ East longitudes) Forest Reserves (in Uganda) from June 2005 to March 2007.Researchers have conducted studies in Lwamunda forest since 1965 (Gombya-Ssembajjwe, 1999;Baldwin et al., 1976;Chalmers, 1968;Spuhler and Jorde, 1975;Waser, 1975), so the animals were fairly habituated.Both forests being within the Lake Victoria Basin, they have bimodal rainfall patterns with 1200-1500 mm per year and typically distributed between two distinct rainy seasons: April-May and October-November.Mabira is located on undulating hills and valleys at an elevation range of 1070-1340 m above sea level while Lwamunda is a riverine forest reserve situated at 1200 m of altitude and surrounded by isolated flat-topped hills with steep slopes.Mabira consisted of fairly undisturbed (M1) and regenerating (M2) The influence of tannins on fruit parts consumed by mangabeys was investigated by examining a total of 2076 samples comprising of 786 whole fruit, 888 pulp, 261 endosperm and 141 seed at different stages of development.The samples were collected from the top 10 priority fruit tree species consumed by mangabeys.Three (3) groups of habituated mangabeys one in Lwamunda and two in Mabira were systematically observed for five consecutive days at the beginning of each month which constituted 1200 h during the 24 month study period.However, the number of hours varied on daily basis depending on our ability to follow them.A systematic 5 min-scan sampling was conducted (Fashing, 2001;Waser, 1975) to collect data on fruit species and development stage of fruit being eaten by mangabeys.When regular systematic sampling was obstructed by mid-storey tree canopies especially in Mabira forest reserves, systematic sampling was supplemented with ad libitum observations.The parts of fruit/seed and the remains consumed by mangabeys were noted.Fruit remains under trees where mangabeys had been feeding were also carefully examined for identification of the exact parts the mangabeys were feeding on and mechanisms of fruit processing prior to consumption.Top 10 priority fruit tree species (in 10 replicates) utilized by mangabeys in each study area were identified (Table 1) marked and their location noted by GPS.
The availability of fruit on the marked fruit trees was recorded on a monthly basis using a visual abundance score (VAS: 0-4), where 0 denoted total absence of fruit, 1 denoted 1-25% of fruit cover, 2 denoted 26-50%, 3 denoted 51-75% and 4 represented above 76% fruit availability (Chapman et al., 1992).The level of food availability was calculated from the percentage of marked trees with edible ripe fruits within the range of each group.The age of fruit maturity and level of ripeness (Saini et al., 2009) were determined by first establishing the time when the tree fruited and subtracted from the time of the visit during phenological data collection.However, the stages of fruit ripeness was recorded on a weekly basis.Furthermore, the weight and size of the fruit were measured to verify the age of the fruit.The actual fruit ripening was visually assessed by colour changes.Immature green fruits were designated stages I-III of development with age variation of ≤ 12 weeks, yellowish colours were normally stages III-V with age variation of ≤ 36 weeks and red or maroon represented stages V-VII with age >36 weeks of development.

Analysis of tannins
According to Conklin-Brittain et al. (1998), tannins are one of the major classes of secondary compounds or antefeedants that occur in higher concentrations in the diets of cercopithecines.Other researchers (Rode et al., 2006;Krief et al., 2005;Chapman and Chapman, 2002;Cipollini and Levey, 1997;Herrera, 1982) have used a whole spectrum of secondary compounds to study nutritional quality of non-human diets but owing to lack of critical laboratory equipment it was not possible in this study.Instead, one secondary compound and specifically using vanillin assay (Broadhurst and Jones, 1978) was used to determine tannins levels.
Essentially, 7 test tubes with varying concentrations of catechin standard were prepared as follows: (0; 0.05; 0.10; 0.15; 0.20; 0.25 and 0.30 mg/ml (in 100% methanol).1 ml of each replicate of sample extract was added to test tubes, in duplicate, whilst in a water bath at 30°C. 5 ml freshly prepared vanillin reagent (1:1 ratio of 1% vanillin/methanol and 8% concentrated HCl/methanol) was added at 1 min intervals.The mixture was left in the water bath at 30°C for 20 min.The absorbance was read at 500 nm against a blank.The blank is equivalent to 1 ml vanillin reagent 5 ml of 4% concentrated HCl/methanol was added at 1 min intervals and then kept at 30°C for 20 min.Small differences in the original weight are corrected by multiplying absorbance by 200/(sample weight in milligrams).Tannin levels were used as basis for establishing one of the contributory factors influencing fruit choice or rejection of certain parts of some 10 priority fruits.The assumption was that mangabeys in both forest reserves avoided or traded-off fruit parts of the priority fruits with tannin levels above acceptable threshhold.

Data analysis
Due to different micro-environments obtaining in different forests or study sites, different stages of fruit development and different fruit parts consumed or rejected, it deemed necessary to compare level of tannins in the same fruit type occuring in all the three sites.Combined effects of forest type, tree species, fruit part and ripening stage on the level of tannin in fruits were explored using mixed model factorial ANOVA.Forest type had three groups, tree species were 12, edible fruit parts were five, while ripening stages were nine.In the case of significant interactions, data were split and independent main effects of each parameter explored using oneway analysis of variance while adjusting for the unique effect of other independent variables.

RESULTS
There were statistically significant main effects of forest type (F(2,9.94)=7.099,P<0.001)), tree species (F(11, 53.514)=38.207,P<0.001), edible fruit part (F(4,277.074)=197.818,P<0.001) and ripening stage (F(8,61.325)=43.783,P<0.001).The respective Eta squared for forest type, tree species, fruit part and ripening stage were 0.007, 0.182, 0.295 and 0.139; indicating that the order of importance of the independent variables in explaining fruit tannin levels is fruit part, followed by tree species and ripening stage.Forest type which explained fruit tannin levels to be only 0.7% was dropped from further analysis.The results also show significant two and three level interactions of tree species, edible fruit part and ripening stage (P <0.001).After adjusting for the effect of edible fruit part and ripening stage, the concentration of tannin in fruit species was significantly different (mean ±SD, P<0.001) and in the order: Celtis mildbraedii (1.17±0.81)<Myrianthus Adjusting for both fruit species and ripening stage, the concentration of tannin in the different edible parts of the fruits was significantly different (mean ± SD, P<0.001) and in the order: pulp (4.63±3.59)>whole (3.47±1.75)>seed (2.49±1.31)>endosperm (1.31±0.52)>aril (0.57±0.06).As expected, tannin level significantly decreased (mean ± SD, P<0.001) with increasing ripening, after adjusting for the unique effects of fruit species and edible fruit part (Table 2).
It was observed that mangabeys utilized specific fruit parts at certain stages of development.The fruit development stage at which mangabeys consumed the priority fruit part varied with the type of fruit (Table 3).In most cases, fruit maturity corresponded with level of fruit ripeness.
It appears the level of tannin determined the stage at which different parts of the fruit were eaten by mangabeys.It varied from 0.7 g/100 g in pulp of M. arboreus to C. mildbraedii pulp that registered 2.6 g/100 g.The weighted mean for maximum allowable limit for consumption of the priority fruit part was around 2.6 g/100 g as indicated in Table 4.
However, during periods of fruit scarcity, mangabeys consumed different fruit parts which tended to vary with the type of fruit (Table 1).Whereas mangabeys consumed young seeds and fairly ripe pulp of Celtis species, they consumed whole fruit of Ficus species from weeks 20-32 after fruiting, young endosperms and ripe pulp of C. indicum, M. arboreus, M. eminii and P. microcapa and the ripe pulp of R. edulis.Mangabeys concentrated on the consumption of B. unijugata seed up to 18 weeks after fruiting then switched to aril after fruit Generally, the level of tannins in whole fruit declined with maturity (Figure 1a, b and c).The rate of decline in tannin content as indicated by R 2 value varied with fruit type and habitat.In the fairly undisturbed Mabira (Figure 1a), regenerating Mabira (Figure 1b) and Lwamunda forest reserves (Figure 1c), the rate of decline varied from R 2 = 0.7835 for M.eminii to R 2 = 1 for C. indicum.Consumption of whole fruit was limited to fig species and yellow fruits (R. edulis).The level of tannins in figs varied from 4.7 g/100 g in the three months old whole fruit to 0.5 g/100 g at week 28 which was the preferred stage for consumption.Yellow fruit was ingested whole, the fruit skin spat out and the seed passed out in faeces.Consumption of whole ripe M.eminii and C. indicum fruit was observed more often in Lwamunda than in Mabira and the mode of their consumption was similar to yellow fruit.
According to our field observations, pulp was the most consumed fruit part regardless of the fruit type and     indicated by the number of fruits whose pulp was regularly retrieved as food (Figure 2a, b and c).
Mangabeys started consuming pulps from most fruits from 10 weeks when the tannin content was around 2.6 g/100 g up to weeks 28 or 32 when tannin levels were less than 1 g/100 g.Generally, tannin content decreased with pulp maturity (Figure 2a, b and c) and rate of decline (R 2 ) varied with type of fruit regardless of habitat.Mangabeys started consuming endosperms between weeks 3 and 10 after fruiting as observed in Table 2.When the tannin content exceeded 2.6 g/100 g and the pericarp hardened, they reverted to consumption of pulp whose tannin levels were declining to lower than 1 g/100 g.This was from weeks 20 to 32 depending on type of fruit.In severely degraded Lwamunda forest reserve where fruit resources were perpetually at low ebb, endosperm consumption was more prevalent than in both types of Mabira forest reserves.The level of tannins in endosperms seemed to increase with the fruit development regardless of fruit type and habitat (Figure 3a, b and c).
As expected, since the level of tannins increased in the endosperms with maturity, it would inevitably increase in the seeds (Figure 4) regardless of the fruit type.Consumption of the seeds was predominantly observed in Lwamunda forest reserve.This was presumably due to lack of ripe fruit and generally prevailing food scarcity.

DISCUSSION
Avoidance of fruit parts with levels of tannins exceeding 2.6 g/100 g threshold is undoubtedly a selection criteria among mangabeys inhabiting Mabira and Lwamunda forest reserves.Mangabeys tended to select fruit parts for consumption that did not exceed the threshold regardless of the fruit type and habitat.This was in agreement with other authors (Danish et al., 2006;Ganzhorn, 1986;Glander, 1982) who noted that presence of undesirable secondary compounds played a crucial role in the selection of foods among primates.Other authors have observed other factors that influence food selection among primates.These include evolutionary disposition of fruit and taste (Chalmers, 1968), energetics (Chivers, 1986) and food availability (Olupot et al., 1997).According to Ganzhorn (1986) apart from the concentration of tannins and alkaloids, selection of a food item was also dependent on availability of other foods and seasonality.We observed that food availability in degraded Lwamunda forest reserve was more compromised than in Mabira (M1) that was fairly undisturbed.It is therefore probable that mangabeys in Lwamunda have developed trade-off mechanisms to cope with habitat changes.The trade-off could entail consumption of food items with high levels of tannins but with high protein or lipids as a mitigation strategy to meet their daily dietary requirements.As such, mangabeys inhabiting severely degraded Lwamunda forest reserve were forced to consume fruits with higher levels of tannins as a trade-off.This food selection criterion was more evident in Lwamunda than in Mabira (M1) which was indicative of food scarcity.It appears the Mangabey  food choice was primarily determined by quantity and availability of food items as opposed to its quality attributes.
Mangabeys may have also avoided consumption of fruit parts with unacceptable levels of tannins as a strategy to conserve energy for other physiological activities since detoxification of foods with high levels of tannins and other secondary compounds is not cost-   effective (Nabihamba, 1996).In the absence of preferred ripe fruit pulp with < 1 g/100 g of tannins, mangabeys would rather spend comparatively less energy to access fruit endosperm with low nutritive quality per unit volume (Wrangham et al., 1998) than consume unripe fruit that would require detoxification.However, not all endosperms of fruits are nutritively poor; Verderane et al. (2013) reported that endosperms of palm nuts of catulé and piaçava species were rich in fats (60%) and carbohydrates (30%) and White (2011) noted that endosperms contained unavailable amino acids required by the seedling.Nonetheless, consumption of foods with high levels of condensed tannins compromises the nutritive quality of food resources and affect rates of digestibility and nutrient extraction (Doran-Sheehy et al., 2006).From a plant"s perspective, synthesis of tannin molecules is energetically costly owing to their complex chemical structure (Coley, 1986;Sagers and Coley, 1995), so unless the external protection pressures are high, plants invest in other less costly ecosystem activities.Since the production of secondary compounds including tannins is a deterrent mechanism employed by the plant to fend off folivorous and frugivorous pests (Howe and Jander, 2008;Bennett and Wallsgrove, 1994) prior to the intended use, complete avoidance of fruit parts with high levels of tannins is not likely to be a realistic strategy by mangabeys in Lwamunda or Mabira (M2) forest reserves especially during periods of food scarcity.
As such, primates have deviced strategies to overcome the severe impacts of secondary compounds in their diet.They produce salivary proline-rich proteins that bind tannins into complexes thereby ensuring that absorption of vital nutrients from the intestinal canal takes place (Sorensen et al., 2004).Alternatively, they regulate the intake of secondary compounds by a trade-off strategy with macro-nutrients (Schaefer et al., 2003).A vivid example in the present study was the regular consumption of seeds in Lwamunda forest reserve despite their relatively high tannin content as they reached maturity.However, at times mangabeys reverted to consumption of pulp.It is not clear whether because the seed coat was too tough or the tannin levels were too high.Considering that Mangabey molars are quadrate, relatively long and bilophodont (Flannery, 2000), cracking hard fruit or tough pericarps, nuts and seeds (Marshall and Wrangham, 2007), should not hinder the consumption of endosperm.So the reversion to pulp consumption was actually due to unacceptable levels of tannins.
Indeed, according to other researchers (Chapman et al., 2012;Wrangham et al., 1998), mangabeys have developed tolerance to higher levels of tannins.They advanced that development of tolerance is likened to human children who if they encounter tannins in the diet regularly, their sensitivity wears off and they develop an improved ability to secrete protein-rich mucins from mucosal surfaces of the mouth which go into their saliva and serve to neutralize tannins (Reynolds, 2005).
The foraging pattern was evident during periods of food scarcity or seasonal variability of food resources especially among mangabeys inhabiting the severely degraded Lwamunda forest reserve.Mangabeys were compelled to consume endosperms retrieved from unripe fruit through a labourous and time consuming process of biting and spitting the bitter pulp until they gained access to endosperm with low levels (< 0.6/100 g) of tannins.Unripe fruit pulp is known to have high levels of tannins (Nathiya and Dorcus, 2012;Doran-Sheehy et al., 2006;Wrangham et al., 1998;Garber, 1987) and mangabeys were observed spitting out raw pulp when processing fruit to access endosperm.
A similar behaviour was reported among the Chimpanzees of Budongo forest reserve (in Uganda) that devised a tannin-avoidance mechanism by rolling fig seeds into oral boli "wadges" and spitting them out because fig seeds contain higher levels of condensed tannins than pulp (Reynolds et al., 1998).
The Mangabey selection of a fruit part for consumption also depended on general food availability.When food was abundant, mangabeys concentrated on consumption of ripe fruit with low tannin levels.However, during periods of food scarcity, mangabey diet consisted of unripe fruit and raided crops (Ogango, 2006).Figs were invariably consumed whole while Canarium, Rheedia, Pseudospondias, Maesopsis and Myrianthus were manipulated to access fruit endosperm.The food selection may also be species-specific (Wieczkowski, 2003).Some primate species like Chimpanzees (Pan troglodytes) prefer ripe fruit while mangabeys concentrate on ripe fruit when it is available but otherwise their diet is essentially omnivorous.Hence, during periods of food scarcity, they utilize seed, unripe fruit and other forestry products.
The nutritional quality of some endosperms is low (Wrangham et al., 1998) and food was extremely scarce in Mabira (M2) and Lwamunda forest reserves because we observed that during July to December 2006, there was literally no fruiting tree except a few trees of F. sur with 2 months old fruits which were inedible.It was therefore not possible that those 2-3 trees with immature fruits could sustain a group of 10 mangabeys notwithstanding their competitors.
To sustain the dietary requirements of mangabeys and other frugivores, there ought to have been other food resources which these forest fauna relied on for survival.Ogango (2006) observed that during periods of food scarcity in Lwamunda which were more frequent than in other habitats, mangabeys engage in crop-raiding.Fungo et al. (2013) reported the highest crop-raiding in the maize-cassava mixtures in the villages that were close to the more disturbed zones of Central Mabira forest.According to Rode et al. (2006), raided food crops are selected on account of their high rate of digestibility, high energy and low secondary compounds.

Conclusion
Mangabeys are capable of detecting differences in tannin concentrations in fruit for consumption regardless of type or part or habitat as illustrated by the determined threshold.In incidences when consumption of seed becomes unavoidable, mangabeys seem to consider the quality of macro-nutrients contained therein and employ trade-off mechanism.In other words, because the seed is rich in protein, lipids and carbohydrates, the energy acquired through its consumption is more than the energy used for detoxication of the ingested tannins.Although Mabira and Lwamunda forest reserves are within the Lake Victoria Basin, the micro-environments are totally different and therefore the influence of tannin levels on selection of fruit parts eaten by mangabeys varies according to food availability and seasonality.

Figure 1a .
Figure 1a.Tannin variation in whole fruit at different stages of development for priority fruit trees commonly utilized as food in the fairly undisturbed portion of Mabira forest reserve.

Figure 1b .
Figure 1b.Tannin variation in whole fruit at different development stages of priority fruit trees commonly utilized as food in the regenerating portion of Mabira forest reserve.

Figure 1c .
Figure 1c.Tannin variation in whole fruit at different development stages of priority fruit trees commonly utilized as food in the severely degraded Lwamunda forest reserve.

Figure 2a .
Figure 2a.Tannin variation in fruit pulp at different stages of development for priority fruit trees commonly utilized as food in the fairly undisturbed portion of Mabira forest reserve.

Figure 2b .
Figure 2b.Tannin variation in fruit pulp at different development stages of priority fruit trees commonly utilized as food in the regenerating portion of Mabira forest reserve.

Figure 2c .
Figure 2c.Tannin variation in fruit pulp at different development stages of priority fruit trees commonly utilized as food in the severely degraded Lwamunda forest reserve.

Figure 3a .
Figure 3a.Tannin variation in fruit endosperm at different stages of development for priority fruit trees commonly utilized as food in the fairly undisturbed portion of Mabira forest reserve.

Figure 3b .
Figure 3b.Tannin variation in fruit endosperm at different development stages of a single fruit tree commonly utilized as food in the regenerating portion of Mabira forest reserve.

Figure 3c .
Figure 3c.Tannin variation in fruit endosperm at different development stages of priority fruit trees commonly utilized as food in the severely degraded Lwamunda forest reserve.

Figure 4 .
Figure 4. Tannin variation in seed at different development stages of a few priority fruit trees commonly utilized as food in the severely degraded Lwamunda forest reserve.

Table 1 .
Top 10 priority fruit tree species utilized by mangabeys in Mabira and Lwamunda forest reserves.
portions but in the same location.

Table 2 .
Decline of tannins in fruit parts from immaturity to maturity and final ripening stage.

Table 3 .
Fruit parts frequently consumed by mangabeys and development stages for priority trees utilized as food by mangabeys.

Table 4 .
Tannin levels and development stage at which mangabeys start eating priority fruit part.