In USFWS/SppDistMonProj: Species Distribution Monitoring Project Data Workflow Tools

# Load packages
library(raster)
library(unmarked)
library(knitr)
library(rmarkdown)

knitr::opts_chunk$set(echo = FALSE, results = "asis", warning = TRUE)

# Get user parameters
s_dir <- params$s_dir
o_dir <- params$o_dir
hab_data <- params$hab_data

o.files <- list.files(o_dir, pattern = "occurrence", full.names = TRUE)
# o.files <- list.files("./data/raw", pattern = "occurrence", full.names = TRUE)
o.years <- as.numeric(sapply(o.files, extract.year))
# o.files <- o.files[o.years <= year]
o.years <- as.numeric(sapply(o.files, extract.year))

s.files <- list.files(s_dir, pattern = "sample_site", full.names = TRUE)
# s.files <- list.files("./data/raw", pattern = "sample_site", full.names = TRUE)
s.years <- as.numeric(sapply(s.files, extract.year))
# s.files <- s.files[s.years <= year]
s.years <- as.numeric(sapply(s.files, extract.year))

##miss.o <- !(s.years %in% o.years)
##miss.s <- !(o.years %in% s.years)
years <- intersect(o.years, s.years)
year <- max(years)
occu.files <- o.files[o.years %in% years]
site.files <- s.files[s.years %in% years]

## Load survey year data
nyears <- length(years)
nlogical <- ifelse(nyears==1, FALSE, TRUE)

# Load spatial data and plot survey map
habcov <- raster(hab_data)
# habcov <- raster("../resources/data/geodata/habcov.asc")
nsites <- length(habcov)

# Load site and occurrence data
sitedfs <- lapply(site.files, read.csv)
occudfs <- lapply(occu.files, read.csv)

## Process and format data
sdmpData <- process.data(occudfs, sitedfs, years, habcov)
class(sdmpData) <- c("sdmpData", "list")

# save(sdmpData,file = paste0("../data/final/species_occurrence_data_", year, ".RData"))
figcap <- paste0("Plot of site-specific naive occurrence summed over ", nyears,
                 " survey years and represented as scaled blue circles. ",
                 "Cell colors represent habitat covariate values.")

\newpage

Project description

The Species Disribution Monitoring Project is an annual refuge Inventory and Monitoring survey began in 2011 to monitor species occurrence and the relationship between occurrence and the distribution of habitat across a National Wildlife Refuge. The survey consists of conducting multiple surveys for a species presence. The survey design as described in the Refuge Species Distribution Monitoring Project Protocol requires the surveys be conducted at a random sample of 20 sites each year. With a given year, each site is surveyed for species presence every day for a consecutive 5-day period, each day (i.e., occasion) constituting an independent sample of occurrence data. Sites are equal-area polygons (i.e., grid cells) as defined by a lattice overlaid on the Refuge boundary (Fig. \@ref(fig:surveymap)) which divides the refuge into r nsites unique sites. Observers are instructed to search each cell to which he/she is assigned with the same intensity to standardize detection by survey effort. Detection is known to be affected by visibility conditions and, therefore, a categorical visibility covariate (i.e., good vs poor) is collected for each survey occasion in each year to be used in data analysis. Habitat quality is a known factor affecting the probability of species occurrence. Because habitat quality across the the refuge does not substantially vary over time (i.e., year to year), a static spatial data set measuring habitat quality at the level of the survey site (Fig. \@ref(fig:habitatmap)) is used in the analysis to account for variation in occurrence due to habitat quality.

\newpage

Overall project summary

The Species Occurrence Project began in 2011 and surveys have been conducted each year up to this year, r max(years). This section provides summary information for the entire duration of the project.

\newpage

Survey area

The survey area corresponds to the entire refuge and is divided into r nsites equal-area polygons, each constituting a single survey site with a unique site ID (Fig. \@ref(fig:surveymap)). Each site is characterized by a measure of habitat quality measured on the the real number scale (Fig. \@ref(fig:habitatmap)).

#print(plot.surveymap(habcov))
verts = as.vector(sp::bbox(habcov))
rasterpts = data.frame(raster::rasterToPoints(habcov))
raster::plot(raster::rasterToPolygons(habcov), axes = TRUE, xlab = "X", ylab= " Y", asp = 1)
text(rasterpts$x, rasterpts$y, label = 1:nrow(rasterpts), cex = 0.75)
polygon(x = verts, y = verts[c(1,3,4,2)], lwd = 2, border = "blue")

#print(plot.habitatmap(habcov))
rasterpts = data.frame(rasterToPoints(habcov))
raster::plot(habcov, xlab = "X", ylab = "Y", asp = 1)
text(rasterpts$x, rasterpts$y, label = 1:nrow(rasterpts), cex = 0.75)

#print(plot.habcov.hist(habcov))
hist(values(habcov), breaks = seq(-3,3,0.2), ylim = c(0, 20),
     main = "", xlab = "Habitat covariate")

\newpage

Survey data

naiveoccus <- sapply(sdmpData$NaiveOccu, mean)
if(nyears==1){
    text1 <- paste0("To date, 1 covariate has been collected for detection and 1 covariate has been collected",
                    " for species occurrence. The naive proportion of sites occupied in ",
                   years, " was ", naiveoccus, ".")
} else {
    text1 <- paste0("To date, 1 covariate has been collected for detection and 1 covariate has been collected",
                    " for species occurrence. The naive proportions of sites occupied each year ranged from ",
                    min(naiveoccus), " to ", max(naiveoccus),
                    "(Fig. \\@ref(fig:naiveoccuplot)).")
}

cat(text1)

if(nyears>1){
    plot(years, naiveoccus, ylim = c(0,1), type = "b", pch = 16, frame.plot = FALSE,
         axes = FALSE, ylab = "Naive proportion of sites occupied", xlab = "Year")
    axis(2)
    axis(1, at = years)
}

\newpage

Individual year summaries

This section contains year-specific summary information ...

\newpage

# Directory path for figures
#figdir.path <- "../products/reports/annual_qc/figure/"
dir.create("figure")
figdir.path <- "./figure/"

dummy <- mapply(annual.summaries, 
                occu = sdmpData$occuDataList, 
                site = sdmpData$siteDataList,
                year = sdmpData$years, 
                unmkd = sdmpData$unmarkedDataList, 
                naiveoccu = sdmpData$NaiveOccu,
                MoreArgs = list(habcov = sdmpData$habcovData, figdir.path = figdir.path),
                SIMPLIFY = FALSE)

\clearpage

Appendix

\clearpage

R session info

sessionInfo()

---

title: "Species Distribution Monitoring Project: Annual data summary and quality control report for r max(years)"

---

USFWS/SppDistMonProj documentation built on May 27, 2020, 9:44 a.m.