Fig 1.
Annual cycle of sunrise and nautical twilight at Tiputini.
Sunrise, the time when the edge of the sun first appears above the horizon, and nautical twilight, the time when the sun is 12º below the horizon, at the Tiputini Biodiversity Station (76°08’ W, 0°38’ S). Day 1 is January 1. Values were downloaded from the website of the US Naval Observatory (http://www.usno.navy.mil/USNO/astronomical-applications); they vary only 1–2 min from year to year. Curves are averages for years 2006–2013.
Fig 2.
Weather patterns at the Tiputini Biodiversity Station.
A. Monthly variation in mean daily temperatures, mean daily minimum temperatures, and mean daily minimum temperatures for the subset of days when behavioral data were collected. Temperature was recorded for 77% of 4364 days between June 2005 and June 2017. B. Mean monthly rainfall recorded for 86% of 2871 days between June 2008 and April 2016. Months with data for fewer than 15 days were excluded from the averages. Error bars indicate standard errors.
Fig 3.
Ripe fruit abundance at Tiputini.
Monthly abundance of ripe fleshy fruit as estimated by total basal area of trees bearing ripe fruit (m2) along a 9 km phenology transect. Values can be converted to basal area per hectare by dividing by 5.25 ha, which is the total sampling area of the transect.
Table 1.
Statistical models, factors and interactions, and tests performed.
Fig 4.
Time of first movement from the sleeping tree (hours relative to sunrise).
Boxplots: median and interquartile range (IQR); whiskers indicate extreme values unless there are data points more than 1.5*IQR below the lower quartile or more than 1.5*IQR above the upper quartile, in which case whiskers stop at the first data points within those boundaries and open circles indicate more extreme points. n = number of observations.
Table 2.
Time of first movement from the sleeping tree (hours relative to sunrise).
Table 3.
Times when canopy features became visible to human observers.
Fig 5.
Time to begin first feeding bouts (hours relative to sunrise).
Boxplots indicate the median and the interquartile range (IQR); whiskers indicate the extreme values unless there are data points more than 1.5*IQR below the lower quartile or more than 1.5*IQR above the upper quartile, in which case the whiskers stop at the first data points within those boundaries and open circles indicate more extreme points. n = number of observations.
Table 4.
Time of first feeding bouts (hours relative to sunrise).
Table 5.
Feeding delays and distance traveled.
Table 6.
Analyses of associations between timing of first feeding bouts and ecological factors.
Fig 6.
Relationships between times of first feeding bouts and ecological factors.
Panels A, C, E, G: Time of first feeding bout as a function of minimum temperature on the day of the observation. Panels B, D, F, H: Time of first feeding bout as a function of ripe fruit abundance in the biweekly period containing the observation day. Least squares regression lines are plotted, based on the displayed points.
Fig 7.
Times of first feeding bouts of all taxa relative to sunrise, grouped by quadrimester.
Median values for times of first feeding bouts and numbers of observations are in Table 7.
Table 7.
Median times of first feeding bouts (hours relative to sunrise).
Table 8.
Relationships of first movements and feeding times to biotic and abiotic factors.