In RoyChristian/GeoAviR: Calling MCDS from R.

Startified models with `mcds.wrap()`

R2MCDS can be used with stratified model of distance sampling.

The first step is to transform the survey lines for the project into a spatial object.

library(R2MCDS)
library(rgdal)

### Import and filter the observation data
data(quebec)
df1 <- mcds.filter(quebec,
                   transect.id = "WatchID",
                   distance.field = "Distance",
                   distance.labels = c("A", "B", "C", "D"), 
                   distance.midpoints = c(25, 75, 150, 250),
                   effort.field = "WatchLenKm",
                   lat.field = "LatStart", 
                   long.field = "LongStart",
                   sp.field = "Alpha",
                   date.field = "Date")

### Build a shapefile with transect starts
transect <- data.frame(lat=df1$LatStart,lon=df1$LongStart)
coordinates(transect) <- ~lon + lat
transect<-SpatialPointsDataFrame(transect,data=df1[,"Count",drop=FALSE])
proj4string(transect)<-CRS("+proj=longlat +datum=WGS84 +no_defs +ellps=WGS84 +towgs84=0,0,0")

The second step is to import the shapefile of the survey area and to associate each survey line to a zone in the survey area.

### Import data and transform into a a spatial polygon
data(zonegulf)
zonegulf<-spTransform(zonegulf,CRS(proj4string(transect)))

#overlay transect and Zones
transect_id <- over(transect,zonegulf)
df1$zone <- transect_id$id
df1$zone_area<- transect_id$area
df1 <- df1[!is.na(df1$zone),]

In our case, observations were made during bouts of a few minutes (identified by WatchID) each day. We will therefore use the day as the sample unit and we will sum the length of all the bout made within a day to calculate the total effort for the day.

### Build labels for samples. (zone +date)
df1$SMP_LABEL<-paste(df1$zone,df1$Date,sep="_")
### Aggregate effor by Sample unit
effort <- aggregate(WatchLenKm~SMP_LABEL,data=unique(df1[,c("SMP_LABEL","WatchID","WatchLenKm")]),sum)
names(effort)[2] <- "SMP_EFFORT"
df1 <- merge(df1,effort,sort=FALSE)

We then run the analysis. For this example we will use the observations of Black-legged Kittiwake (Rissa tridactyla). For simplicity, we will only run a detection model with a Hazard rate key function with Cosine adjustment and no explanatory variable.

###Run the analysis for the Gull species only
mod1 <- mcds.wrap(df1,
                  SMP_LABEL="SMP_LABEL",
                  SMP_EFFORT="SMP_EFFORT",
                  DISTANCE="Distance",
                  SIZE="Count",
                  Type="Line",
                  units=list(Distance="Perp",
                             Length_units="Kilometers",
                             Distance_units="Meters",
                             Area_units="Square kilometers"),
                  breaks=c(0,50,100,200,300),
                  estimator=list(c("HA","CO")),
                  detection="All",
                  lsub=list(Alpha=c("BLKI")),
                  split=TRUE,
                  empty=NULL,
                  STR_LABEL="SMP_LABEL",
                  STR_AREA="zone_area",
                  stratum="STR_LABEL",
                  path="C:/temp/distance",
                  pathMCDS="C:/Program Files (x86)/Distance 7",
                  verbose=FALSE)
mod1
summary(mod1)
plot.distanceFit((mod1))