Fig 1.
Germination success of fig seeds on filter paper.
Mean germination success for Ficus polita, F. grevei and F. lutea for faecal (dark shading) and for unprocessed seeds removed from ripe fruits (light shading) sown on filter paper. There were 15 replicates of 20 seeds/ treatment. The error bars are standard deviations.
Fig 2.
Germination success of fig seeds in semi-natural conditions.
Mean germination success of Ficus polita faecal, unprocessed and from rejecta pellet seeds under three different light intensities. Each treatment was replicated 15 times with 20 seeds/replicate (2,700 seeds tested). The bars represent standard deviations.
Fig 3.
Gut retention times of Madagascan flying foxes.
Seed retention time in four colour coded individual bats. The vertical axis represents number of fig seeds present in the faeces, the maximum of 25 refers to a score of ‘25 and over’. The horizontal axis represents time since the last feeding of a bat (seed ingestion). Measurements were made every 30 mins for up to 6 h after food was presented (after this time droppings were produced rarely). The bats produced 63 droppings in total.
Fig 4.
Travel distances of Madagascan flying foxes.
Distances travelled by 11 Pteropus rufus bats from fig trees for up to 3 h after feeding (the time in which most droppings containing more than 25 seeds were produced in our GRT study–see Fig 3). The position of the bat was considered every 12 min as a straight line along its flight path in relation to the feeding tree it used. The bars represent average recorded distance from the feeding sites (n = 71) with the standard deviation. The circles indicate the maximum distance recorded.
Fig 5.
Seed shadow plot from one feeding event.
Sample trajectory with one single feeding event (12 hours: 18.00h-06.00h). The spatio-temporal data (GPS fixes in blue) have been plotted at the same resolutions as recorded. The start and end points of the trajectory have been marked (the roost at 0,0 km) as well as the first and subsequent defecation event horizons. Notice that the two circles are not concentric. Concentric circles would appear only if the individual would have returned at time T2 at the same location where it was estimated to be at time T1.
Fig 6.
Seed shadow plot with multiple feeding events.
Sample trajectory with several different feeding events and locations from one bat covering nearly a week of feeding. The start and end of the trajectory are at the roost (0,0) with two defecation event horizons shown for each feeding event. The light green and blue colour indicates a single defecation event while the darker colours indicate more than one overlapping circle due to several defecation events. The small blue dots are GPS fixes.
Fig 7.
Seed shadow plots from all bats combined.
The combined data of all 11 bats (45 nights of feeding) are plotted relative to the average position (a) and overlaid on a Google Earth map of the area (b) with major sisal plantation (red) and forest fragments (green) indicated (c). The rest of the habitat is composed of agricultural land, clearings, spiny forest or small groups of trees. All feeding locations are marked and the two defecation event horizons plotted corresponding to each feeding event. Light green and blue indicates a single defecation event (30% and 90% defecation probability respectively) while the darker colours indicate accumulations of several defaecation events. In (b) and (c) map data: Google, DigitalGlobe.