Supplementary materials
Taï chimpanzees anticipate revisiting high-valued fruit trees from further distances
Animal Cognition
Simone D Ban1,2,3, Christophe Boesch1, Karline R L Janmaat1
1Max Planck Institute for Evolutionary Anthropology, Deutscher Platz 6, 04103, Leipzig, Germany.
2Centre Suisse de Recherches Scientifiques, 01 BP 1303 Abidjan 01, Côte d’Ivoire.
3UFR Biosciences, Université Félix Houphouët Boigny, 22 BP 1106 Abidjan 22, Côte d’Ivoire.
Author E-mail address:
Simone Dagui Ban ()
* Corresponding author: Simone Dagui Ban
Max Planck Institute for Evolutionary Anthropology,
Deutscher Platz 6,
04103, Leipzig, Germany.
Tel.: +49-(0) 341-3550-258,
E-mail address: ,
Supplementary materials for estimation of the maximum detection distances
When measuring visual detection distance, both crown and trunk visual distances were measured but only the highest was used to estimate the visual detection distance of a revisited tree. To determine themaximum detection distance we compared this visual detection distance with the olfactory detection distance and selected the largest distance (see below FigureS1).When assistants were unable to measure the visual and or the olfactory distances due to vegetation growth or time limits (when the number of visited trees were high), we used the highest value measured for the respective species.
FigureS1: Drawing the measurement of maximum detection distance (MDD)
Supplementary materials for estimation of fruit amount and tree crown size
To assess the amount of fruit that was available at the previous feeding visits we measured the proportion of the branches that werecarrying fruit (the fruit production class, see main manuscript) in combination with the size of the crown. To estimate the crown size we measured: 1) the diameter of the trunk at breast height (DBH; measured at 1.2 m above ground; Leighton and Leighton 1982; Chapman et al. 1992) for trees, 2) the crown radius in all four wind directions (from the trunk to the end of the crown) for figs and 3) the circumferences of the roots that reached the forest floor(Janmaat et al. 2013a).For strangler figs, theestimated crown size was defined as the surface of the crown as if it was perpendicularly projected on the forest floor(Janmaat et al. 2013a). We assumed the crown of each fig tree to be an ellipse and therefore used the following formula to calculate the ellipse surface: (PI*r N*r E)/4+(PI()*r E*r S)/4+(PI()*r S*r W)/4+(PI()*r W*r N)/4, (r is the radius of the tree crown measured in north (N), south (S), east (E) and western (W) direction). We z-transformed the estimated crown size of tree, figs and lianas to make them comparable. We obtained thisstandardized estimated crown size by subtracting the mean crown size of the species concerned and dividing the result by the standard deviation of these same crown sizes. We onlyincluded only plants species that have at least 5 individuals eaten(Janmaat et al. 2013a).
Supplementary materials for calculation of relative cumulative energy balance at the change point
To estimate the energy balance of the females at the moment they significantly change their travel direction towards the revisited trees, we calculated the “relative cumulative energy balance” using the methods described in N’Guessan et al.(2009). We started our calculations of each relative energy balance at value zero at the moment when the female left her nest in the morning. We added energy intake and distracted energy expenditure until the moment the female reached the last changepoint after which she revisited a feeding tree. As a slight improvement of theirmethodology we calculated energy expenditure including each female’s real velocity values (measured by Garmin 60Csx GPS during travel) and the climbing and nut cracking times of the respective female on the respective day.
The velocity (V) was calculated using the following formula:
V= D (m) / t (s), where D wasthe distance recorded as travelling between the different activities points and tis the time of travelling. (Weran a program in R to calculate distances by using the following formula: D
(x2= Longitude 2, x1= Longitude 1, y2= Latitude 2, y1= Latitude 1).
In addition, we used individual tree and female specific fruit intake rates whenever vegetation cover had allowed to do these measurements. If no intake rate could be recorded we used the average value for the species or if that was not available the value calculated by Antoine N’Guessan (unpublished data). For most food items fed on we used the average species-specific energy values measured by N’Guessan et al.(2009). For food items for which such information was missing we used energy values known for food items of similar type (e.g. fruit or young leaves) or of similar size. We realize that our calculations resulted in rough estimations of energy balance, for example the energy expenditure was based on a captive chimpanzee walking a treadmill. However, since we used the same calculations for each female and only included changepointsthat were preceded by complete follows at which we did not loose the females, our measures provide a valuable estimation of the relative energy balance.
Supplementary materials for the result of previous fruit production class on out-of-sight approach distances
FigureS2: The previous fruit production class did not affect the out-of-sight approach distance.
Y-values,shown on a square root scale, represent out-of-sight approach distancesbetween last change point until their entry within the maximum detection field (Figure1). X-values represent the previous fruit production classes of the revisited trees. Class zero means that the trees were depleted during the previous visit. The circles represent out-of-sight approach distancesfor the respective fruit classes. The oblique line represents the out-of-sight approach distancespredicted by the model. Boxes represent average out-of-sight approach distancesper fruit class. Bars in straight and dash lines represent respectively the median and mean values of the out-of-sight approach distances, upper and lower boundaries of boxes represent the upper and lower quartiles.
Supplementary materials for details on tree species revisited, out-of-sight approachdistances and odiferous and non-odiferous fruit tree species
TableS2: Tree species revisited with mean out-of-sight approach distancesand odiferousand non-odiferous fruit tree speciesTree species / Number of trees / Number of revisits / Mean out-
of-sight approach distances / Obvious smell
Chrysophyllum taiensis / 2 / 2 / 904.22 / n
Cordia platythyrsa / 1 / 1 / 299.54 / n
Dialium aubrevillei / 7 / 9 / 891.02 / n
Drypetes aubrevillei / 2 / 6 / 517.38 / n
Duboscia macrocarpa / 3 / 4 / 507.88 / y
Ficus elasticoides / 6 / 9 / 566.9 / y
Ficus saussureana / 2 / 3 / 449.49 / y
Ficus sp / 1 / 1 / 710.86 / y
Ficus kamerunensis / 3 / 3 / 336 / y
Ficus sansibarica / 6 / 8 / 452.69 / y
Ficus polita / 1 / 1 / 297.89 / y
Ficus mangeferoides / 1 / 1 / 455.4 / y
Ficus umbellata / 1 / 1 / 419.3 / y
Garcinia kola / 1 / 2 / 562.34 / y
Grewia malacocarpa / 1 / 1 / 2153.6 / n
Magnistipula butayei / 3 / 4 / 185.12 / y
Irvingia grandifolia / 6 / 17 / 447.65 / y
Klainedoxa gabonensis / 35 / 48 / 464.1 / y
Musanga cecropioides / 3 / 3 / 439.46 / y
Nauclea diderrichii / 6 / 7 / 709.68 / y
Duguetia staudtii / 1 / 1 / 648.03 / y
Panda oleosa / 1 / 1 / 796.47 / n
Parinari excelsa / 4 / 4 / 577.08 / y
Sacoglottis gabonensis / 16 / 25 / 419.85 / y
Scottelia klaineana / 9 / 10 / 744.5 / n
Scytopetalum tieghemii / 6 / 7 / 535.41 / n
Zanha golungensis / 1 / 1 / 626.59 / y
1