Study Area
Where did the study take place?
Bilsa Biological Station in Ecuador
What is Oenocarpus?
A long-lived palm, large distribution, "hyper-dominant" in the Amazon
What was the timing of frugivore observations?
February 2009-June 2016 (!)
What habitat characteristics were measured around each tree?
% canopy cover, canopy height, # of trees in the genus Cecropia, # of large trees
What analysis was used to test for habitat associations?
Multiple generalized mixed models "the linear predictor contains random effects in addition to fixed effects"
Where were data collected, specifically?
130ha area at 435-615m elevation
What is the fruiting pattern of Oenocarpus?
How were frugivores observed?
Observations made of focal fruiting trees for 3 hours at a distance of 20m in early or late afternoon
How was canopy cover measured?
Estimated with a spherical densiometer
How were fruit removal rates calculated?
The total number of seeds removed by a specie during an individual 3-hour observation period averaged across 7 years of observation
What are the biological characteristics of the site?
High endemism, high biodiversity, high rates of deforestation
How long are fruits on the palm?
4 weeks
What data were recorded?
Bird ID, # of individuals, length of visit, # of intact seeds removed
How was canopy height measured?
Digital range finder of tallest tree in a 10m radius from focal tree
What is a Poisson error structure?
"The assumed probability distribution of the response"
What are the seasons at the site?
July-December: Cool and dry
January-July: Warm and wet
Why is it assumed that current Oenocarpus are "remnants"?
They are slow-growing and take 80 years to mature
Which seed predators were not documented and why?
Nocturnal (observations only made in daylight) and ground foragers (ideas?)
How were large trees counted?
Trees with DBH>50cm within 20m of focal tree
What is the problem with multicollinearity?
If several independent variables in a model are correlated it may result in less reliable statistical inferences
What are the forest types at the site?
Primary old-growth, 19-30 yo secondary forest, selectively logged forest
What is the distribution of Oenocarpus across the different forest types?
Similar densities across forest types
How was percent removal calculated?
Why were the habitat characteristics that were measured used?
Taken from the literature as indicators of forest structure
How did Mike figure out all these stats?
He got help!