Split Column into Other Two Lat Long

You can use strsplit to separate the coordinates, taking advantage of the "look back" option of regular expressions. Because strsplit returns a list, the most practical is to first throw the result to an object and then include each column in the original data.frame.

dados <- read.table(text = c(
 "BIRDS.GENERO         SP1        SP2                       XLOCAL
  Thamnomanes      caesius    glaucus                0°32'S52°12'W 
  Thamnomanes   ardesiacus  obidensis                0°10'S51°50'W
  Thamnomanes   ardesiacus  obidensis               01°40'N51°23'W
  Thamnomanes   ardesiacus  obidensis               01°40'N51°23'W"),
  header = TRUE, stringsAsFactors = FALSE
)

listaC <- strsplit(dados$XLOCAL, "(?<=N|S)", perl = TRUE)
dados$Lat  <- unlist(lapply(listaC, '[', 1))
dados$Long <- unlist(lapply(listaC, '[', 2))

> dados
  BIRDS.GENERO        SP1       SP2         XLOCAL     Lat    Long
1  Thamnomanes    caesius   glaucus  0°32'S52°12'W  0°32'S 52°12'W
2  Thamnomanes ardesiacus obidensis  0°10'S51°50'W  0°10'S 51°50'W
3  Thamnomanes ardesiacus obidensis 01°40'N51°23'W 01°40'N 51°23'W
4  Thamnomanes ardesiacus obidensis 01°40'N51°23'W 01°40'N 51°23'W

I could not get a straightforward solution, I used the separate function of the tidyverse package twice and I made some manipulations on the resulting variables to get the result you expect.

Being dados your bank, follow code:

library(tidyverse)

dados %>% 
  separate("XLOCAL", into = c("Lat1", "Long1"), sep = "N", remove = F, extra = "drop", fill = "right") %>% 
  separate("XLOCAL", into = c("Lat2", "Long2"), sep = "S", remove = F, extra = "drop", fill = "right") %>% 
  mutate( Lat1 = paste0(Lat1, "N"), Lat2 = paste0(Lat2, "S") ) %>% 
  mutate( Long2 = ifelse(is.na(Long2), "", Long2), Long1 = ifelse(is.na(Long1), "", Long1) ) %>% 
  mutate( Lat = ifelse(Long1 == "", Lat2, Lat1), Long = paste0(Long2, Long1) ) %>% 
  select(BIRDS.GENERO:XLOCAL, Lat, Long)

• The first separate is breaking XLOCAL , in two variables ( Lat1 and Long1 ), through the N separator;
• The second separate is breaking XLOCAL , in two variables ( Lat2 and Long2 ), through the S separator;
• The first mutate is including the N and S indicators, respectively, Lat1 and Lat2 , respectively, since the separate function excludes the separators used;
• the second mutate is modifying what is NA for a blank field;
• The third mutate is creating the variables Lat and Long using the variables "auxiliary" Lat1 , Lat2 , Long1 , Long2 ;
• Finally, the select function is selecting the variables of interest