Whales of the deep: analyzing movement and diving of humpback whales to understand oceanic breeding congregations in New Caledonia

Humpback whales’ (Megaptera novaeangliae) habitat use in low-latitude breeding grounds is well documented from decades of coastal research. Yet, the use of pelagic habitats during the breeding season and migration has only recently been given attention.

In New Caledonia, an archipelago located in the Pacific South West, several seamounts and banks play an important role for the local humpback whale population. Yet, the reason why whales would aggregate and move between these offshore waters remains unknown. The relative abundance of maternal females in these unsheltered waters is also puzzling, in comparison to the shallow coastal waters usually occupied by these groups.

Using the newest satellite tracking technology, this project aims at understanding the environmental and social drivers of humpback whale oceanic habitat use during the breeding season. Dive depth will be related to environmental context in order to shed light on the role played by offshore seamounts for humpback whales of the Pacific South West.

This study case will provide a unique opportunity to understand the spatio-temporal scale of the humpback whale floating lek systems and explore the drivers of habitat selection during the breeding season.

bandeau logos MARACAS 06 final

This work is part of the WHERE project supported by the New Caledonian Government, the French Ministère de la transition écologique et solidaire, the World Wildlife Fund.

The « Whales of the deep » project is specifically supported by the Society of Marine Mammalogy (Louis M. Herman research scolarship).


November 2018 update

[by Solène Derville]






# Dataframe of ARGOS whale positions (in a EPSG4326 reference system: lat long)
## Warning in readChar(con, 5L, useBytes = TRUE): cannot open compressed file
## 'crawl_argos_df.RData', probable reason 'No such file or directory'
## Error in readChar(con, 5L, useBytes = TRUE): cannot open the connection
# Raster of bathymetry converted to a dataframe format
## Warning in readChar(con, 5L, useBytes = TRUE): cannot open compressed
## file 'bathyNOAA_southpacific_df.RData', probable reason 'No such file or
## directory'
## Error in readChar(con, 5L, useBytes = TRUE): cannot open the connection
# version of the bathymetry limited to -2000 m to highlight shallow areas/ seamounts etc
bathyNOAA_df_shallow <- bathyNOAA_df
## Error in eval(expr, envir, enclos): object 'bathyNOAA_df' not found
bathyNOAA_df_shallow[bathyNOAA_df_shallow$z < (-2000),]$z <- -2000
## Error in bathyNOAA_df_shallow[bathyNOAA_df_shallow$z < (-2000), ]$z <- -2000: object 'bathyNOAA_df_shallow' not found
# Shapefile of reef polygons converted to a dataframe format

# Images to be inserted as .png inside the map
img <- readPNG("whale-silhouette-png.png", TRUE)
bandeau <- readPNG("bandeau horizontal MARACAS 2018.png", TRUE)

# create discrete colour scale for whale tracks
colour_whales <- colorRampPalette(brewer.pal(11, "Spectral")[c(1:5,7:11)])(12)
colour_whales <- c(colour_whales, colorRampPalette(brewer.pal(11, "PiYG")[c(1, 2, 4)])(3))
colour_whales <- c(colour_whales, colorRampPalette(brewer.pal(11, "PuOr")[c(1:3)])(3))
colour_whales <- sample(colour_whales)

# produce each image in a loop
# locations are separated by 6 hours intervals and the longest track has 256 locations
for (i in seq(1,256,1)){
  crawl <- ddply(crawl_df, ~id, function(d) {
    d <- d[1:i, ]
    d <- d[!is.na(d$id), ] # remove NA rows for dataframe that are shorter than i
    d$index <- c(1:nrow(d))/nrow(d)
  crawl_last <- ddply(crawl, ~id, function(d) { d <- d[nrow(d), ]}) # select only the last position
  # produce map
  g <- ggplot(crawl, aes(x = lon360, y = lat)) +
    geom_tile(data = bathyNOAA_df, aes(x, y, fill = z)) +
    scale_fill_distiller(palette = "Blues", direction = -1, guide = F) +
    stat_contour(data = bathyNOAA_df_shallow, aes(x, y, z=z), binwidth = 500, color = "grey60", size = 0.2) +
    geom_polygon(data = reef_shallow_df, aes(long, lat, group = group),
                 fill = "darkgrey", col = "transparent") +
    geom_map(data = world_map, map=world_map, aes(x = long, y = lat, map_id = id), fill = "black") +
    geom_path(aes(lon360, lat, group = id, col = id), size = 1.5) +
    scale_color_manual(values = colour_whales, guide = F) +
    geom_polygon(data = land_df, aes(long, lat, group = group), fill = "black", col = "black") +
    geom_text(aes(x = 165, y = -19.5, label = "New Caledonia"), size = 7) +
    geom_text(aes(x = 151.1, y = -22, label = "Australia"), size = 7) +
    xlab("longitude") +
    ylab("latitude") +
    theme_void() +
    scale_x_continuous(lim=c(150, 170)) +
    scale_y_continuous(lim=c(-28, -18)) +
    coord_fixed(expand = F) +
    xmin = 150.5, xmax = 156,
    ymin = -19.8, ymax = -18.2)
  # add the colour whale silhouettes over the last location of each individual track
  for(j in 1:nrow(crawl_last)){
    if (crawl_last$lon[j] < 170 & crawl_last$lon[j] > 150 & crawl_last$lat[j] < -18 & crawl_last$lat[j] > -28){
      g <- g + annotation_custom(
        xmin = crawl_last$lon360[j]-0.21, xmax = crawl_last$lon360[j]+0.21,
        ymin = crawl_last$lat[j]-0.21, ymax = crawl_last$lat[j]+0.21
      img_col <- readPNG(paste("./whale_silhouettes_18/whale-silhouette-png-", j, ".png", sep =""), TRUE)
      g <- g + annotation_custom(
        xmin = crawl_last$lon360[j]-0.18, xmax = crawl_last$lon360[j]+0.18,
        ymin = crawl_last$lat[j]-0.18, ymax = crawl_last$lat[j]+0.18
  # save the plot as .jpg in a separate folder
  ggsave(plot = g, file = paste("./anim_soutenance_coralsea/image_", i, ".jpg", sep = ""),
         width = 400, height = 200, dpi = 300, units = "mm")
## Error in ddply(crawl_df, ~id, function(d) {: object 'crawl_df' not found


October 2018 update

[by Solène Derville]

The whale season is almost behind us and it is time to analyze all the data we acquired this winter. As mentioned in the previous post (see below), the photo-identification of individuals is one of the routine monitoring method that we use in the field. Indeed, humpback whales can be individually identified based on the trailing edge and black-and-white patterns on the underside of their fluke. Every year, we photograph as many individuals as possible in the field and compare these pictures to the catalog of the New Caledonia humpback whale population. To date, this catalog includes more than 1500 unique individuals. Given the small size of the New Caledonia population, individuals are relatively frequently resighted across and among breeding seasons. As a result, cataloged whales have “sighting histories”, which allow us to retrace their movements and life events through time.

And just by chance, fieldwork sometimes surprises us with unexpected coincidences… The history of “Zealandia”, a whale tagged in July over the Antigonia seamount, is a case in point. Zealandia was one of the first whales observed when we arrived over the seamount on the early morning of July 17th, 2018. It was later resighted in a competitive group and tagged that same day. Although the whale did not show its fluke during the focal follow, its white flanks had been photographed on both sides and were quite distinctive… now we usually do not compare the photographs of dorsal fins and flanks because it would be too time consuming. Genetic analysis of biopsy skin samples allows the identification of individuals in such case. But we were so eager to know who this whale was that we made an exception and scrolled through the thousands of dorsal photographs we had in the catalog. And Eureka! We found the whale’s identity, which will later be confirmed through genetic analysis. Referenced under the code “HNC524”, this individual turned out to be quite exceptional…


Photographs of Zealandia’s right and left flanks (source: Opération Cétacés)

Zealandia is a female humpback whale observed for the first time in 2001 on a small bank located south of the New Caledonia mainland. She was then resighted in 2008, 2009, 2010… with a calf on each occasion. This may not seem incredible to someone who is unfamiliar with the biology of humpback whales… but let me explain why I am making such a fuss about this particular one. Humpback whales are known for having a birth interval of about 2-3 years.

Zealandia (HNC524) caudal fluke, as referenced in the New Caledonia humpback whale catalog (source: Opération Cétacés)

Females breed in the winter in their tropical breeding ground, migrate to their feeding ground, come back on the next year to give birth and migrate one more time accompanied by their calf. Calving in consecutive years is therefore though to be very rare… Yet, recent evidence from population dynamics modeling in New Caledonia suggests otherwise. Among the 452 females resighting histories analyzed by Chero et al. (In review), Zealandia is the only one known to have given birth three years in a row.

So how does this link back to our satellite tracking project? Well, individual movements are best interpreted if we know as much as possible about the whale’s past. As it happened, after she was tagged, Zealandia visited three major breeding grounds located in the southeastern New Caledonian waters: Antigonia, Orne bank and the South Lagoon. Why did she purposefully cover these important distances to visit several spots, both coastal and offshore? Although at the time she was tagged she was not with a calf, she could have given birth later in the season. Progesterone analyses to be conducted on the blubber sample collected in the biopsy will tell us whether she was pregnant at the time of tagging. If she was not, this roaming behavior could be interpreted as a search for mating opportunities. Given how early she started her migration south (around August 2nd), this hypothesis is quite likely.

Satellite tracking of humpback whale « Zealandia » in July-August 2018 (source: IRD/Opération Cétacés/WWF/gouv.nc; map background: Wildlife Computers/Google Satellite)

Finally, Zealandia was tracked during part of her journey toward the Southern Ocean feeding grounds. Interestingly, she spent about a week around Norfolk Island, located a few hundred kilometers northwest of New Zealand. This observation confirms the importance of this island as a stop-over or perhaps a resting site during the humpback whale migration.

In the end, we managed to bring to light a small part of a humpback whale’s mysterious life by combining multiple research approaches. Understanding the movements and behaviors of these elusive creatures is a challenge for scientists, but it feels very rewarding to occasionally catch a glimpse of an whale’s life. So wait for the next update to hear another whale’s story!


Chero, G., Pradel, R., Gimenez, O., Bonneville, C., Derville, S., Garrigue, C. (In Review) High breeding capacity highlights recovery potential of humpback whale populations in the Southern hemisphere. Animal Conservation.

August 2018 update

[by Solène Derville]

It is now been more than two months since the first humpback whales showed up in New Caledonia for their annual breeding season. In July, our team composed of members of the French Institute of Research for Development, the Opération Cétacés NGO and the World Wildlife Fund, left the Nouméa harbour aboard the Amborella. This year again, the New Caledonian government supported our project by providing this large vessel, which allows us to survey in the open ocean.

The goal for this cruise was clear: deploy 6 satellite tags on adult humpback whales at the Antigonia seamount (see previous blogpost). Tags are deployed on adult whales only, at the front of the dorsal fin, using a modified pneumatic line-thrower (called “ARTS”, see video below). While being tagged, whales are also photographed and biopsied in order to obtain the sex (through molecular sexing from the skin sample) and the identity (through genotyping and photo-identification) of the animal. These informations are essential to reconstruct the “capture history” of the tagged whales, that is where and when it has been observed in the past.

2017-07-25-c 004 M.Oremus
Claire, Rémi and I are preparing the tags for the next day. We are not letting ourselves get distracted by the beatiful scenery of the Isle of Pines in the background!

Despite rough weather on the first day (20 knot wind and big swell!) we managed to deploy two tags. The others followed on the next day at sea. Tagging 12 m long whales is one of the toughest job I have ever witnessed in the field. A great level of precision and communication among crew members is essential for this task to succeed. So you can imagine our joy when we managed to deploy 4 tags in a single day!

This short video presents the method for tag deployment from a small semi-inflatable boat using a modified pneumatic line-thrower (in 2017 for whale « Sial » and 2018 for whale « Gondwana »). The crossbow is used simultaneously to collect a biopsy sample (a small piece of skin and blubber).

One unexpected challenge that follows tagging is… finding names! For this year’s batch we went for the “geology theme”, with a special South Pacific touch… Whales were nicknamed: Pangea, Gondwana, Zealandia, Oceania, Nazca, and Tethys. Hopefully lucky names that will make these tags last for long!

Once the tags deployed we were free to use our remaining time at sea to visit other areas. We therefore decided to visit the Ellet seamount near Walpole Island (see previous blogpost), a previously unsurveyed seamount that had been used by a whale tagged in 2016. Surely enough, whales were found there! Photographies and skin samples collected there will help us understand the connexion between this new seamount and neighboring pelagic breeding grounds.

2015-10-04-RL- 075
Photo of a tagged humpback whale. The top of the tag is visible, as well as the rocket that detached on impact and the arrow that contains a tiny skin sample!