Portal Forecasts 1/27/2017

This is a forecast made for the upcoming January 27 - 29 rodent trapping at the long-term Portal research site in southern Arizona. Four models were used to achieve forecasts with 90% prediction intervals. There are two forecast tables. One is for the site as a whole (total means across all plots, all treatment types), named "All". The other is for the 4 long-term Control plots only ("Controls").<br>Two naive models using the defaults in the forecast package in R were used to create forecasts for the total granivore captures (species="Total") over the next 12 trapping periods, the first being tomorrow.<br>The "Forecast" modelforecast(abundances$total,h=12,level=0.95,BoxCox.lambda(0),allow.multiplicative.trend=T)<br>The "AutoArima" modelforecast(auto.arima(abundances$total,lambda = 0),h=12,level=0.95,fan=T)<br>Species-level predictions, as well as another total captures prediction, were made using the tscount package in R.<br>The "NegBinom Time Series" model, by species (species codes are the first letter of the genus and species)<br>model=tsglm(abundances[[s]],model=list(past_obs=1,past_mean=11),distr="nbinom")<br>predict(model,12,level=0.9) <br>The "Poisson Environmental" model, by species (species codes are the first letter of the genus and species), using monthly minimum temperature, maximum tempurature, mean temperature, total precipitation from our on-site weather station, and NDVI from a compiled LANDSAT MODIS estimate. Model selection was based on AIC and performed for each species.<br>model=tsglm(abundances[[s]],model=list(past_obs=1,past_mean=11),distr="poisson",xreg=weather)<br>predict(model,12,level=0.9,newdata=weatherforecast)