Coping with feast and famine: Integrated behavioral and metabolically flexible responses of wild orangutans to ecologically driven dietary variation

Diet and nutrition are critical factors influencing energetics and health. Laboratory studies show that organisms adjust to changes in nutrient intake through flexible metabolic responses. While the physiological effects of nutrient balance in humans have been studied, data from closely related species living in nature are lacking. We integrate macronutrient regulation and metabolic flexibility to elucidate how wild Bornean orangutans are buffered against natural fluctuations in nutrient intakes. We found that orangutans regulate protein and regularly switch between exogenous and endogenous nutritional substrates as preferred food availability declines. When total caloric, lipid, and carbohydrate intakes decline, orangutans drew on fat and endogenous amino acids for energy. This strategy is beneficial only in the context of alternating periods of fruit scarcity and abundance. Our findings provide a direct analog for the current global obesity pandemic, which has arisen in parallel with transitions in human diets towards energy-dense, protein-dilute foods.

Methods
Behavioral observations and urine sampling of wild orangutans (Pongo pygmaeus wurmbii) were collected between 2003 – 2018 and 2004-2017 (Fig. 1A), respectively, in the Tuanan Research Area located in the Mawas Conservation Area, Central Kalimantan, Indonesia  (see 34, 37 for a detailed site description). Tuanan is a peat swamp forest with a peat depth of 1-3 meters in most areas (69). Annual rainfall is between 1309.8 - 4176.0 mm with an average of 2602.4 mm, and minimum and maximum temperatures range between 20.5 °C to 32.0 °C (average 23 °C to 28.5 °C).  For this study, we observed a total of 26 adult females, 48 adult flanged males, 25 adult unflanged males, 20 independent immatures, 15 weaned immatures who still traveled with their mothers (SI Appendix Table S8). All field research was approved by the Institutional Animal Care and Use Committees of Rutgers, the State University of New Jersey.

The fruit availability index (FAI) was determined from monthly monitored phenology plots spread across the study area comprising 1,522 and 3,103 tagged trees (see SI Appendix Methods). We calculated FAI as the percentage of trees in the plots with fruit each month (34, 37). High and low fruit period categories were determined by calculating the overall median from 2003-2018, which was used as the cut-off point to assign low and high fruit periods to each month.