In brouwern/DRmencia: Data and code to replicate analyses for Latta, Brouwer, et al. 2018. PeerJ.

knitr::opts_chunk$set(echo = TRUE)

Reconciling the two datasets

SL provided two complementary versions of the data

One was in seperate datasheets for each year

The other was organized to examine site persistance

At first glance the datasets appeared to have different numbers of unique records. This is indeed true but is actualy expected. Differences are almost all due to 2007-2008 records; this makes sense. That was the last year of the study and so site persistance of newly captured birds cannot be determined

Load cleaned and scrubbed LONG data

Data that was contained in seperate files by year

filename <- "mencia_cleaned.csv"
filename <- here("data-raw",
                  "data-raw-mencia",
                  "mencia_all_captures_by_year",
                  filename)

mencia <- read.csv(filename)

Load site persistence df

filename <- "site_persist_cleaned.csv"
filename <- here("data-raw",
                  "data-raw-mencia",
                  "mencia_site_persistence",
                  filename)

sp <- read.csv(file = filename)

Unique records in "mencia" dataframe

length(unique(mencia$band))

Unique records in "site persistance" dataframe

Each row should be unique. There appear to be 3 duplicate bands. This appears to be due to changes between sites within or between years.

Note that the "menica" dataframe from annual records does contain at least some within-year recapture records, perhaps when there was movement between sites.

i.unique.band <- match(unique(sp$band),sp$band)

sp$ID <- with(sp, paste(band,species,site, sep = "."))
i.unique.ID <- match(unique(sp$ID),sp$ID)

length(i.unique.band)
length(i.unique.ID)
dim(sp)

sp$ID[duplicated(sp$ID)]

i.dup <- which(sp$ID %in% sp$ID[duplicated(sp$ID)])

sp[i.dup, 1:12]

  site band.pre band.suf colors species age  sex      date wing tarsus

1347 Cueva 2240 21542 XR-WR AMRE HY F 4-Jan-04 57 17.5 1348 Cueva 2240 21542 XR-WR AMRE HY F 5-Jan-04 NA NA 1513 Cueva 4A 209 BX-YW BCPT HY 9-Nov-03 82 23.0 1517 Cueva 4A 209 BX-YW BCPT HY 6-Jan-04 83 23.9 1649 Cueva 2280 33228 WW-XR CMWA AHY F 18-Feb-04 64 18.4 1662 Cueva 2280 33228 WW-XR CMWA AHY F 22-Feb-04 NA NA

Look at these duplications in other dataframe

dup.band <- sp$band[duplicated(sp$band)]

mencia[which(mencia$band %in% dup.band), 1:12]

Difference in number of unique records

length(unique(mencia$band))-length(unique(sp$band))

Are the species lists the same between years?

There are several spp not in the "site persistance" data BWWA LOST SUTA WHQD

mencia.spp <- unique(mencia$species)
sp.spp <- unique(sp$species)

d1 <- data.frame(df = "mencia",
                 species = mencia.spp)
d2 <- data.frame(df = "sp",
                 species = sp.spp)

merge(d1,d2, by = "species",all = TRUE)

This only accounts for 7 records

spp.x <- c("BWWA","LOST","SUTA","WHQD")

which(mencia$species %in% spp.x)

Compare band lists

There appear to be 313 unique bands in the "mencia" dataframe that are not in the site persistance (sp) dataframe. This matches the differences in dataframe size

mencia.bands <- unique(mencia$band)
sp.bands <- unique(sp$band)

d1b <- data.frame(df = "mencia",
                 band = mencia.bands)
d2b <- data.frame(df = "sp",
                 band = sp.bands)

d3 <- merge(d1b,d2b, by = "band",all = TRUE)

Differences are almost all due to 2007-2008 records; this makes sense . That was the last year of the study and so site persistance of newly captured birds cannot be determined

bands.on.run <- d3[is.na(d3$df.y),"band"]

bands.on.run.i <- which(mencia$band %in% bands.on.run)

summary(mencia[bands.on.run.i, 1:19])

brouwern/DRmencia documentation built on May 6, 2019, 12:24 p.m.