Exploring the Success and Setbacks of the White Abalone Captive Breeding Program

Author
Affiliation

Leela Dixit

UCSB Master of Environmental Data Science

Published

March 5, 2026

Image of the underside of a white abalone.

Background

White abalone are a species of marine snail, one of 7 native species to the California coast. They are a culturally, ecologically, and economically important species to many communities. They range from Point Conception in California to Baja California in Mexico, historically supporting many indigenous communities for sustenance, and their shells used in art, rituals, and ceremonies.

Abalone, including white abalone, are a common seafood delicacy. So much so that by the 1970s, humans fished 99% of them, crashing their wild populations and fishery. Abalone need to be very close in order to reproduce, but because of the extreme drop in population they could no longer reproduce in the wild. Despite the fishery closure in 1996, white abalone could not rebound on their own, and they were declared the first marine invertebrate to be endangered in 2001.

Because of the success of captive breeding other abalone species, a recovery plan was developed with the goal of captive breeding from the remaining few wild white abalone. Thus begins the White Abalone Captive Breeding Program (WACBP)! The program has found great success captively breeding and outplanting white abalone, slowly working towards recovering this endangered species. However, they still face challenges including disease, larval survival, and genetic diversity.

In this infographic, I decided to explore the data taken for the last decade from the WACBP to shed light on the successes this program has had, and where there may still important be work to do! My infographic will address the following questions:

When thinking about the reproductive patterns in each stage of a white abalone’s life cycle:

  • How do male and female adult white abalone differ in their reproductive output? Does it change based on their age?

  • Have captive populations of juvenile white abalone increased over time? How many of those juveniles have been sent for outplanting in the wild?

  • Do different genetic lines of white abalone spawn more? How do offspring of wild white abalone spawn?

This infographic is made of 3 plots to answer these questions, so I will break down each one before revealing the final graphic!

Data Sources

Data have been collected in this program since its inception, consisting of counts, vitals (lengths, weights, and sex information), and spawning frequency and amounts! Data for this project are not publicly available.

Heatmap

Having witnessed many white abalone spawns myself, there was always the looming question “Who is going to spawn today?”. Spawning is still fairly unpredictable for white abalone, but over time the program has started spawning younger animals instead of just the older broodstock. I was curious to see if the younger animals really do spawn more, and if perhaps males spawn earlier than females!

I wanted to focus on comparing spawning output between sex’s to each other, so I decided a heatmap would show density of spawning along a single scale and along a time series. In ggplot, I produced a (almost) finalized heatmap with annotations and color scheme. I went with a very minimal theme to reduce visual noise and make the colors of the heatmap pop!

While it looks alright, it can be better! I moved into Affinity Designer to polish it off and look more professional with improving the typeface and annotations.

Histogram time series

Next, I wanted to highlight the impressive amount of offpsring (juveniles) that the WACBP has produced in the last decade! Truly an impressive feat, creating thousands of an endangered species that would not have existed otherwise! Similarly to the heatmap, in ggplot I produced a (mostly) finished plot! Again, I went with a minimal theme, and I decided to remove the y-axis for simplicity, opting to add annotations in Affinity later.

Once again, I moved this plot into Affinity Designer to polish the look and add more informative annotations.

Sankey Diagram

Finally, the Sankey! This plot is my favorite (and most difficult to make) piece of this infographic. A sankey diagram represents the flow and amount of one state to another. In the diagram for this infographic, I am representing the amount of populations each white abalone population has produced. Particularly, I wanted to highlight the amount of wild white abalone genetics have been passed on to future generations.

In ggplot, I developed a clean data frame based on the captive white abalone pedigree information. Broodstock in one column, the offspring they’ve produced in the next (F1), and the offspring the first offspring have produced in the final column (F2). I produced a sankey for males and females separately, and into Affinity we go!

Once in Affinity, I darkened the flows to be more visible, added some much needed annotations to indicate the flow of wild genetics, and most importantly flipped the male plot, just to look cool!

Putting it all together

Now that the individual plots are finished, its time to put it all together. In Affinity, I can easily move these plots around on my infographic to look clear and clean, along with adding some extra text and small pictures for context and main takeaways.

Color and Typography

I wanted my infographic to have a relatively simple color scheme, as my plots were already quite visually engaging! I went back and forth between a dark vs light background, and ultimately decided on a muted beige. I chose a few shades of blue that would be my main colors for the plots. Any differences in values are represented with different shades, and are still color blind friendly.

For typeface and font, I went with my favorite Montserrat for most annotations, and tried out a new font Zolando for titles! I thought they complimented each other well, both being modern sans serif and evenly spaced, with Zolando a little more loud and wide which, to me, took my attention in a way I would want a title to.

General Design, contextualizing data, and centering the primary message

Because my plots are relatively complex, it was a challenge to organize all three in a clean way to not overwhelm the reader. I decided to play on the ‘life cycle’ theme and build my infographic around a circular life cycle in the middle, indicating what each plot pertains to.

Given that the Sankey diagram is extremely visually complex, I had to make that plot the largest and build around it. I put it at the bottom, which then meant the histogram in the middle and heatmap at the top. Each plot’s main takeaway is in the title and quite large, so a reader can quickly skim the infographic and get the main messages.

Along with the life cycle theme, I did not direct the reader to follow any certain path. All life stages are interesting and important to the program’s success, and as such the reader may choose where to start. Background information, like the program’s history, white abalone’s native range and information on abalone importance to indigenous communities is at the top, as I think it provides necessary context and is the most important information a reader should look at first.

To center my primary message (Successes and setbacks of the WACBP), it is present in the subtitle and also in each plot. Each data visualization has the successes in the title, along with setbacks pointed out in annotations. I decided to not include setbacks in the title to maintain a positive outlook in the infographic (important in restoration work and also everyday life!).

The reveal!

Feel free to explore all of the code used to create this infographic below!

Code 
### load libraries ###
library(tidyverse)
library(here)
library(janitor)
library(showtext)
library(ggsankey)
library(patchwork)

### read in data ###
# counts
counts <- read_csv(here("data", "Count_All_DataRaw.csv")) %>% 
  clean_names() %>% 
  remove_empty(c("rows", "cols"))
# spawning
spawn <- read_csv(here("data", "Spawning_All_DataRaw.csv")) %>% 
  clean_names() %>% 
  remove_empty(c("rows", "cols"))
#pedigree
pedigree <- read_csv(here("data", "White_Abalone_Pedigree_Data_Metadata(Pedigree).csv")) %>% 
  clean_names() %>% 
  remove_empty(c("rows", "cols"))
# transfer
transfer <- read_csv(here("data", "Transfer_All_DataRaw.csv")) %>% 
  clean_names() %>% 
  remove_empty(c("rows", "cols"))

### wrangle data ###
# counts
counts <- counts %>% 
  mutate(pop_id = case_when(
    pop_id == "16-03-02-Acjachemen,\n16-03-12-Tongva,\n16-03-02-Miwok,\n16-03-02-Pomo" ~ "16-03-02-NativeTribes",
    pop_id == "17-03-01-Merida, 17-03-01-Tiana" ~ "17-03-01-Tiana/Merida",
    pop_id == "17-03-01-Tiana, 17-03-01-Merida" ~ "17-03-01-Tiana/Merida",
    pop_id == "16-03-02-Acjachemen, 16-03-02-Tongua, 16-03-02-Miwok, 16-03-02-Pomo" ~ "16-03-02-NativeTribes",
    pop_id == "22-05-10-Bubbles, 22-05-10-Blossom" ~ "22-05-10-Bubbles/Blossom",
    pop_id == "19-04-16-Spongebob, 19-04-16-Patrick" ~ "19-04-16-Spongebob/Patrick",
    pop_id == "19-04-16-Squidward, 18-04-19-Ursula" ~ "19-04-16-Squidward/Ursula",
    pop_id == "21-04-21-Moderna,\n21-04-21-Pfizer,\n21-04-29-JandJ" ~ "21-04-21-Vaccines",
    pop_id == "17-03-01-Tiana,\n17-03-01-Merida" ~ "17-03-01-Tiana/Merida",
    pop_id == "16-03-02-Acjachemen, 16-03-02-Chumash, 16-03-02-Tongva, 16-03-02-Miwok, 16-03-02-Pomo" ~ "16-03-02-NativeTribes",
    pop_id == "22-03-01-Mardi, 22-03-09-WALLE" ~ "22-03-01-Mardi/WALLE",
    pop_id == "22-03-01-Mardi, 22-03-09-Walle" ~ "22-03-01-Mardi/WALLE",
    pop_id == "17-03-01-Belle, 16-03-02-Acjachemen, 16-03-02-Tongva, 16-03-02-Miwok, 16-03-02-Pomo, 16-03-02-Chumash,19-04-16-Squidward, 18-04-19-Ursula, 19-04-16-Spongebob, 19-04-16-Patrick, 17-03-01-Tiana, 17-03-01-Merida, 20-03-10-Smaug" ~ "20-03-10-Mixed",
    pop_id == "16-03-02-Acjachemen, 16-03-02-Tongva, 16-03-02-Miwok, 16-03-02-Pomo, 16-03-02-Chumash" ~ "16-03-02-NativeTribes",
    pop_id == "17-03-01-Belle,16-03-02-Acjachemen, 16-03-02-Tongva, 16-03-02-Miwok, 16-03-02-Pomo, 16-03-02-Chumash" ~ "16-03-02-NativeTribes/Belle",
    pop_id == "25-01-07-Hammerhead, 25-01-07-Nurse, 25-01-07-Epaulette" ~ "25-01-07-Sharks",
    TRUE ~ pop_id
  ))

counts_clean <- counts %>% 
  filter(rack_type != "BST") %>% 
  mutate(year = year(date)) %>% 
#counts_clean <- counts_clean %>% 
  filter(facility == "BML") %>% 
  group_by(date) %>% 
  summarise(count = sum(count, na.rm = TRUE)) %>% 
  ungroup() %>% 
  mutate(type = "counts")

# spawning
spawn <- spawn %>% 
  mutate(pop_id = case_when(
    pop_id == "14-04-23-Beaker, 14-05-13-Fozzie" ~ "14-05-13-Fozzie/Beaker",
    pop_id == "17-03-01-Tiana, 17-03-01-Merida" ~ "17-03-01-Tiana/Merida",
    pop_id == "19-04-16-Squidward, 18-04-19-Ursula" ~ "19-04-16-Squidward/Ursula",
    pop_id == "16-03-02-Acjachemen, 16-03-02-Tongva, 16-03-02-Miwok, 16-03-02-Pomo, 16-03-02-Chumash" ~ "16-03-02-NativeTribes",
    pop_id == "19-04-16-Spongebob, 19-04-16-Patrick" ~ "19-04-16-Spongebob/Patrick",                                                    
    pop_id == "14-04-30-Beaker,14-05-13-Fozzie" ~ "14-05-13-Fozzie/Beaker",
    pop_id == "16-03-02-Acjachemen, 16-03-02-Tongva, 16-03-02-Miwok, 16-03-02-Pomo, 16-03-02-Chumash, 16-03-02-Kumeyaay, 16-03-02-Salinan" ~ "16-03-02-NativeTribes",
    pop_id == "16-03-02-Acjachemen, 16-03-12-Tongva, 16-03-02-Miwok, 16-03-02-Pomo, 16-03-02-Chumash" ~ "16-03-02-NativeTribes",
    pop_id == "17-03-01-Belle, 16-03-02-Acjachemen, 16-03-02-Tongva, 16-03-02-Miwok, 16-03-02-Pomo, 16-03-02-Chumash" ~ "16-03-02-NativeTribes/Belle", 
    pop_id == "19-04-16-Squidward/18-04-19-Ursula" ~ "19-04-16-Squidward/Ursula",
    pop_id == "17-03-01-Tiana/17-03-01-Merida" ~ "17-03-01-Tiana/Merida",
    pop_id == "16-03-02-Acjachemen/16-03-02-Tongva/16-03-02-Miwok/16-03-02-Pomo/16-03-02-Chumash" ~ "16-03-02-NativeTribes",
    pop_id == "19-04-16-Spongebob/19-04-16-Patrick" ~ "19-04-16-Spongebob/Patrick",
    pop_id == "07-05-19-Jeffrey" ~ "17-05-19-Jeffrey",
    TRUE ~ pop_id
  ))

spawn_clean <- spawn %>% 
  # remove unknown popID since no date
  filter(pop_id != "NA",
         pop_id != "UNKNOWN") %>% 
  mutate(born = str_sub(pop_id, start = 1, end = 8), # pull out years from pop ID
         born = as.Date(born, format = "%y-%m-%d"),
         # fix date for swedish fish pop since only year
         born = case_when(pop_id == "03-F1-SwedishChef" ~ as.Date("2003-01-01"),
                          TRUE ~ born),
         # calculate age at spawn date
         age = floor(time_length(interval(born, date), "years")),
         egg_count = as.numeric(egg_count)) %>% 
  # add 5 years to wilds, since we do not know their age
  mutate(age = case_when(pop_id == "04-12-05-Redwood" ~ age + 5,
                         pop_id == "17-01-31-Loblolly" ~ age + 5,
                         pop_id == "17-06-09-Sugar" ~ age + 5,
                         pop_id == "17-05-19-Jeffrey" ~ age + 5,
                         pop_id == "17-03-15-Ponderosa" ~ age + 5,
                         pop_id == "17-04-12-Torrey" ~ age + 5,
                         pop_id == "19-03-18-Ash" ~ age + 5,
                         pop_id == "17-11-13-Sequoia" ~ age + 5,
                         TRUE ~ age))

# transfer
transfer_clean <- transfer %>% 
  filter(life_stage == "settled") %>% 
  # transform any variable with outplant or growout into only 'outplant'
  mutate(transport_purpose = case_when(
    transport_purpose == "Outplant" ~ "outplant",
    transport_purpose == "Outplant, Growout" ~ "outplant",
    transport_purpose == "growout, outplant" ~ "outplant",
    transport_purpose == "Growout" ~ "outplant",
    transport_purpose == "grow-out" ~ "outplant",
    transport_purpose == "growout" ~ "outplant",
    TRUE ~ transport_purpose))

### initialize color theme ###
pal <- c("background" = "#E9E5C7",
         "dark" = "#1F363D",
         "light" = "#40798C",
         "lightsteel" = "#CFDEE7")

### initialize fonts ###
font_add_google(name = "Zalando Sans Expanded", family = "zalando")
font_add_google(name = "Montserrat", family = "montserrat")

### Plot 1: heatmap ###
# wrangle data 
spawn_clean <- spawn_clean %>% 
  mutate(sex = ifelse(!is.na(egg_count), "F", NA)) %>% 
  mutate(sex = ifelse(!is.na(time_initial_spawn) & is.na(egg_count), "M", sex))

spawn_female <- spawn_clean %>% 
  filter(!is.na(sex),
         sex == "F") %>%
  group_by(age, sex) %>% 
  summarise(egg_count = sum(egg_count)) %>% 
  ungroup() %>% 
  mutate(ratio = egg_count / max(egg_count))

spawn_male <- spawn_clean %>% 
  filter(!is.na(sex),
         sex == "M") %>%
  group_by(age, sex) %>% 
  summarise(n = n()) %>% 
  ungroup() %>% 
  mutate(ratio = n / max(n)) 

spawn_plot <- full_join(spawn_female, spawn_male)

# plot
showtext_auto(enable = TRUE)

annotation_heatmap1 <- glue::glue(
  "No abalone have spawned 
  at these ages in captivity")

heatmap <- ggplot(spawn_plot, aes(x = age, y = sex, fill = ratio)) +
  geom_tile() + 
  coord_fixed() +
  annotate(geom = "text",
           x = 16, y = 1.5,
           label = annotation_heatmap1,
           color = pal['dark'],
           hjust = "center",
           family = "montserrat") +
  scale_fill_gradient(low = pal['steel'], high = pal["dark"]) +
  theme_classic(base_size = 15) +
  labs(title = "White abalone reproductive output by age and sex",
       subtitle = "Peak spawning age is around 5 years old") +
  theme(legend.direction = "horizontal",
        legend.position = "bottom",
        panel.background = element_rect(fill = pal['background']),
        plot.background = element_rect(fill = pal['background']),
        plot.title = ggtext::element_markdown(family = "zalando",
                                              color = pal['dark']),
        legend.box.background = element_blank()) +
  guides(fill = guide_colorbar(title = "Reproductive\nOutput", 
                               barwidth = 15, barheight = 1))

showtext_auto(enable = FALSE)

### Plot 2: histogram ###
# wrangle data
transfer_outplant <- transfer_clean %>% 
  filter(#transport_purpose == "outplant",
         origin_facility == "BML") %>% 
  group_by(date) %>% 
  summarise(count = sum(as.numeric(numberofanimals), na.rm = TRUE)) %>%
  ungroup() %>% 
  mutate(type = "transfer")

counts_outplant <- full_join(transfer_outplant, counts_clean) %>% 
  mutate(year = year(date)) %>% 
  group_by(year, type) %>% 
  summarise(count = round(sum(count))) %>% 
  ungroup() %>%
  filter(year != 2025)

# plot
years <- c("2016", "2017", "2018", "2019", "2020", "2021", "2022", "2023", "2024")
breaks <- c(2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024)

title <- glue::glue(
  "Juvenile
  <span style='color:#40798C;'>White Abalone</span>
  and
  <span style='color:#1F363D;'>outplanting</span>
  counts"
)

annotation_histo1 <- glue::glue(
  "COVID-19 Shutdown
  decreased production")

annotation_histo2 <- glue::glue(
  "Lagged effect from
  COVID-19 Shutdown")

showtext_auto(enable = TRUE)

histogram <- ggplot(counts_outplant %>% filter(type == "counts"), 
       aes(x = year, y = count)) +
  geom_col(width = .85, fill = pal["light"]) +
  geom_smooth(data = counts_outplant %>% filter(type == "counts"),
              se = FALSE, color = pal["light"]) +
  geom_col(data = counts_outplant %>% filter(type == "transfer"),
           aes(x = year, y = count),
           fill = pal['dark'], width = .5) +
  geom_smooth(data = counts_outplant %>% filter(type == "transfer"),
              se = FALSE, color = pal['dark']) +
  annotate(geom = "text",
           x = 2020, y = 20000,
           label = annotation_histo1,
           color = pal['dark'],
           hjust = "center",
           family = "montserrat",
           size = 8) +
  annotate(geom = "text",
           x = 2023.5, y = 10000,
           label = annotation_histo2,
           color = pal['dark'],
           hjust = "center",
           family = "montserrat",
           size = 8) +
  labs(title = title,
       ) +
  theme(base_size = 10,
        panel.background = element_rect(fill = pal['background']),
        plot.background = element_rect(fill = pal['background']),
        panel.grid.major = element_blank(),
        panel.grid.minor = element_blank(),
        plot.title = ggtext::element_markdown(family = "zalando",
                                              color = "#040D10",
                                              size = rel(3)),
        axis.title.y = element_blank(),
        axis.ticks.y = element_blank(),
        axis.text.y = element_blank(),
        axis.text.x = element_text(size = rel(2), 
                                   family = "montserrat"),
        axis.title.x = element_text(size = rel(2.5),
                                    family = "montserrat")) +
  scale_x_continuous(breaks = breaks,
                     labels = years)

showtext_auto(enable = FALSE)

# Plot 3: Sankey ###
# wrangle data for female 
pedigree_f <- pedigree %>% 
  filter(!is.na(mother_pop_id)) %>% 
  #filter(if_any(c(mother_pop_id), ~str_detect(., test))) %>% 
  select(pop_id, mother_pop_id) %>%
  separate_longer_delim(mother_pop_id, delim = ", ") %>% 
  separate_longer_delim(mother_pop_id, delim = "/") %>% 
  separate_longer_delim(mother_pop_id, delim = ";") %>% 
  separate_longer_delim(mother_pop_id, delim = " OR ") %>% 
  mutate(mother_pop_id = case_when(
    mother_pop_id == "01-04-23-Scooter (GRN_310)" ~ "01-04-23-Scooter",
    mother_pop_id == "13-04-04-Rowlf (ORN_032)" ~ "13-04-04-Rowlf",
    mother_pop_id == "Parental PopIDs possible: 16-03-02-Acjachemen" ~ "16-03-02-Acjachemen",
    mother_pop_id == "SEE NOTES. N=6 females:\nN=2 18-04-09-Ursula (NYL_667" ~ "18-04-09-Ursula",
    mother_pop_id == "\nN=1 19-04-16-Patrick (NYL_681)" ~ "19-04-16-Patrick",
    mother_pop_id == "\nN=1 16-03-02-Acjachemen" ~ "16-03-02-Acjachemen",
    mother_pop_id == "16-03-02-Chumash (LAV_063)" ~ "16-03-02-Chumash",
    mother_pop_id == "01-04-23-Scooter (GRN_364)" ~ "01-04-23-Scooter",
    mother_pop_id == "13-04-04-Rowlf (ORN_017)" ~ "13-04-04-Rowlf",
    mother_pop_id == "\nN=1 19-04-16-Spongebob (NGN_624)" ~ "19-04-16-Spongebob",
    mother_pop_id == "17-03-01-Tiana (LAV_006)" ~ "17-03-01-Tiana",
    mother_pop_id == "\nN=1 17-03-01-Merida" ~ "17-03-01-Merida",
    mother_pop_id == "20-03-10-Smaug\n" ~ "20-03-10-Smaug",
    TRUE ~ mother_pop_id
  )) %>% 
  filter(mother_pop_id != "NYL_677)",
         mother_pop_id != "Unknown PopID BST Tank 5")

# pull out broodstock and join direct descendents
pedigree_f_brood <- tibble(unique(pedigree_f$mother_pop_id)) %>% 
  rename(mother_pop_id = "unique(pedigree_f$mother_pop_id)") %>% 
  left_join(pedigree_f) %>% 
  rename(f1 = mother_pop_id,
         mother_pop_id = pop_id) 

# pull out f1 generation popid's
pedigree_f_1 <- pedigree_f_brood %>% 
  filter(!f1 %in% mother_pop_id)

# pull out f2 generation popid's
pedigree_f_2 <- pedigree_f_brood %>% 
  filter(f1 %in% mother_pop_id) %>% 
  rename(mother_pop_id = f1,
         f3 = mother_pop_id)

# join f1 and f2 generation to broodstock and make long for sankey
pedigree_f_long <- pedigree_f_1 %>% 
  left_join(pedigree_f_2) %>% 
  rename(f2 = mother_pop_id) %>% 
  filter(!is.na(f3)) %>% 
  make_long(f1, f2, f3)

# plot female
showtext_auto(enable = TRUE)

ped_female <- ggplot(pedigree_f_long, aes(x = x, 
               next_x = next_x, 
               node = node, 
               next_node = next_node, 
               fill = factor(node), 
               label = node)) +
  geom_sankey(flow.alpha = 0.7,
              show.legend = FALSE) +
  scale_fill_manual(values = c('17-03-15-Ponderosa' = "#1F363D",
                               '17-01-31-Loblolly'  ="#1F363D",
                               '17-03-01-Tiana' = "#1F363D",
                               "19-04-16-Plankton" = "#1F363D",
                               "19-04-16-Spongebob" = "#1F363D",
                               "19-04-16-Patrick" = "#1F363D",
                               "21-04-21-BioNTech" = "#1F363D",
                               "19-04-16-Squidward" = "#1F363D",
                               "20-03-10-Smaug" = "#1F363D",
                               "22-02-10-Ciri" = "#1F363D",
                               "22-05-10-Buttercup" = "#1F363D",
                               "24-02-13-Cormorant" = "#1F363D",
                               "23-02-08-Eevee" = "#1F363D",
                               "23-02-08-Polywag" = "#1F363D",
                               "23-04-05-Fiona" = "#1F363D",
                               "23-04-05-Farquaad" = "#1F363D",
                               "23-05-17-Funktopus" = "#1F363D",
                               "23-11-18-Artichoke" = "#1F363D",
                               "24-02-13-Goldfinch" = "#1F363D",
                               "24-02-13-Scrubjay" = "#1F363D",
                               "25-01-07-Nurse" = "#1F363D",
                               "25-03-11-Merry" = "#1F363D",
                               "25-03-11-Samwise" = "#1F363D",
                               "25-03-11-Rosie" = "#1F363D",
                               "25-05-15-Cauliflower" = "#1F363D",
                               "24-02-13-Peregrine" = "#1F363D",
                               "24-02-13-Seagull" = "#1F363D",
                               "25-01-07-Sevengill" = "#1F363D",
                               "25-01-07-Whale" = "#1F363D",
                               "25-03-11-Bilbo" = "#1F363D")) +
  labs(title = "Female genetic lines") +
  theme_minimal(base_size = 15) +
  theme(panel.background = element_rect(fill = pal['background']),
        plot.background = element_rect(fill = pal['background']),
        axis.text.y = element_blank(),
        panel.grid.major = element_blank(),
        panel.grid.minor = element_blank(),
        plot.title = ggtext::element_markdown(family = "zalando",
                                              color = pal['dark']))

showtext_auto(enable = FALSE)

# wrangle data for male
pedigree_m <- pedigree %>% 
  filter(!is.na(father_pop_id)) %>% 
  select(pop_id, father_pop_id) %>% 
  separate_longer_delim(father_pop_id, delim = ", ") %>% 
  separate_longer_delim(father_pop_id, delim = "/") %>% 
  separate_longer_delim(father_pop_id, delim = ";") %>% 
  separate_longer_delim(father_pop_id, delim = " OR ") %>% 
  separate_longer_delim(father_pop_id, delim = ",") %>% 
  separate_longer_delim(father_pop_id, delim = "\n") %>%
  mutate(father_pop_id = case_when(
    father_pop_id == "Parental PopIDs possible: 16-03-02-Acjachemen" ~ "16-03-02-Acjachemen",
    father_pop_id == "13-04-04-Rowlf and" ~ "13-04-04-Rowlf",
    father_pop_id == "or 13-03-02-Janice" ~ "13-03-02-Janice",
    father_pop_id == "17-03-01-Merida (WHT_159)" ~ "17-03-01-Merida",
    father_pop_id == "17-03-01-Belle (WHT(TAN?)_222)" ~ "17-03-01-Belle",
    father_pop_id == "Jabalone Expt: 18-04-19-Ursula (NYL_672)" ~ "18-04-19-Ursula",
    father_pop_id == "18-04-19-Ursula (NYL_669)" ~ "18-04-19-Ursula",
    father_pop_id == "17-03-01-Belle (NGN_151)" ~ "17-03-01-Belle",
    father_pop_id == "01-04-23-Scooter (GRN_371" ~ "01-04-23-Scooter",
    father_pop_id == "13-03-02-Janice (ORN_001)" ~ "13-03-02-Janice",
    father_pop_id == "01-04-23-Scooter(GRN_305" ~ "01-04-23-Scooter",
    father_pop_id == "\n17-03-01-Belle (n=1) " ~ "17-03-01-Belle",
    father_pop_id == "16-03-02-Salinan (n=1)" ~ "16-03-02-Salinan",
    father_pop_id == "17-03-01-Belle (n=1) " ~ "17-03-01-Belle",
    TRUE ~ father_pop_id
  )) %>% 
  filter(father_pop_id != "",
         father_pop_id != "GRN_371)",
         father_pop_id != "GRN_305)")

# pull out broodstock popid's
pedigree_m_brood <- tibble(unique(pedigree_m$father_pop_id)) %>% 
  rename(father_pop_id = "unique(pedigree_m$father_pop_id)") %>% 
  left_join(pedigree_m) %>% 
  rename(f1 = father_pop_id,
         father_pop_id = pop_id) 

# pull out f1 generation popid's
pedigree_m_1 <- pedigree_m_brood %>% 
  filter(!f1 %in% father_pop_id)

# pull out f2 generation popid's
pedigree_m_2 <- pedigree_m_brood %>% 
  filter(f1 %in% father_pop_id) %>% 
  rename(father_pop_id = f1,
         f3 = father_pop_id)

# join f1 and f2 generation to broodstock and make long for sankey
pedigree_m_long <- pedigree_m_1 %>% 
  left_join(pedigree_m_2) %>% 
  rename(f2 = father_pop_id) %>% 
  filter(!is.na(f3)) %>% 
  make_long(f1, f2, f3)

# plot male
showtext_auto(enable = TRUE)

ped_male <- ggplot(pedigree_m_long, aes(x = x, 
               next_x = next_x, 
               node = node, 
               next_node = next_node, 
               fill = factor(node), 
               label = node)) +
  geom_sankey(flow.alpha = 0.7,
              show.legend = FALSE) +
  scale_fill_manual(values = c('17-11-13-Sequoia' = "#1F363D",
                               "19-04-16-Patrick" = "#1F363D",
                               "23-11-18-Artichoke" = "#1F363D",
                               '24-02-13-Goldfinch' = "#1F363D",
                               "24-02-13-Osprey" = "#1F363D",
                               "24-02-13-Scrubjay" = "#1F363D",
                               "24-02-13-Seagull" = "#1F363D",
                               "24-04-23-Nemo" = "#1F363D",
                               "25-01-07-Epaulette" = "#1F363D",
                               "2025-01-07-Rocky" = "#1F363D",
                               "25-05-15-Cauliflower" = "#1F363D"
                               )) +
  labs(title = "Male genetic lines") +
  theme_minimal(base_size = 15) +
  theme(axis.text.y = element_blank(),
        panel.background = element_rect(fill = pal['background']),
        plot.background = element_rect(fill = pal['background']),
        panel.grid.major = element_blank(),
        panel.grid.minor = element_blank(),
        plot.title = ggtext::element_markdown(family = "zalando",
                                              color = pal['dark']))

showtext_auto(enable = FALSE)

# join male and female plots together
sankey <- (ped_female | ped_male)