3 Nutritional contribution and economic profiling

Code

library(ggplot2)


catch_groups_name <-
  timor.nutrients::catch_groups %>%
  dplyr::select(
    grouped_taxa = interagency_code,
    catch_name
  )

tot_catch <-
  timor.nutrients::region_stats %>%
  dplyr::mutate(year = lubridate::year(date_bin_start)) %>%
  dplyr::group_by(grouped_taxa, year) %>%
  dplyr::summarise(catch = sum(catch)) %>%
  dplyr::group_by(grouped_taxa) %>%
  dplyr::summarise(catch = mean(catch)) %>%
  dplyr::left_join(catch_groups_name) %>%
  dplyr::select(-grouped_taxa) %>%
  dplyr::select(catch_name, catch) %>%
  dplyr::mutate(catch = catch / 1000)

base_plot <-
  timor.nutrients::nutrients_table %>%
  dplyr::left_join(catch_groups_name) %>%
  dplyr::select(catch_name, grouped_taxa, Selenium_mu:Vitamin_A_mu) %>%
  rename_nutrients_mu(hyphen = FALSE) %>%
  tidyr::pivot_longer(-c(catch_name, grouped_taxa), names_to = "nutrient", values_to = "concentration") %>%
  dplyr::filter(!nutrient == "selenium" & !catch_name %in% "Other") %>%
  dplyr::left_join(RDI_tab) %>%
  dplyr::mutate(
    nutrient = stringr::str_to_title(nutrient),
    nutrient = dplyr::case_when(
      nutrient == "Omega3" ~ "Omega-3",
      nutrient == "Vitamina" ~ "Vitamin-A",
      TRUE ~ nutrient
    )
  ) %>%
  dplyr::mutate(rdi = (concentration * 100) / conv_factor) %>%
  dplyr::group_by(catch_name) %>%
  dplyr::mutate(tot = sum(rdi, na.rm = T)) %>%
  dplyr::arrange(-tot) %>%
  dplyr::left_join(tot_catch) %>%
  tidyr::replace_na(list(catch = 0)) %>%
  dplyr::arrange(dplyr::desc(nutrient)) %>%
  dplyr::ungroup() %>%
  dplyr::mutate(
    fish_group = dplyr::case_when(
      grouped_taxa %in% c("COZ", "CUX") ~ "Molluscs",
      grouped_taxa %in% c("PEZ") ~ "Shrimps",
      grouped_taxa %in% c("MZZ") ~ "Other",
      grouped_taxa %in% c("SLV", "CRA") ~ "Crustaceans",
      grouped_taxa %in% c("OCZ", "IAX") ~ "Cephalopods",
      grouped_taxa %in% c("SKH", "SRX") ~ "Sharks and rays",
      grouped_taxa %in% c("SNA", "GPX", "PWT", "GRX", "MUI", "BGX") ~ "Large demersals",
      grouped_taxa %in% c("CGX", "TUN", "BEN", "LWX", "BAR", "SFA", "CBA", "DOX", "ECN", "DOS") ~ "Large pelagics",
      grouped_taxa %in% c("YDX", "SPI", "EMP", "SUR", "TRI", "MOJ", "WRA", "BWH", "LGE", "MOB", "MHL", "GOX", "THO", "IHX", "APO", "IHX", "PUX", "DRZ", "DSF") ~ "Small demersals",
      grouped_taxa %in% c("RAX", "SDX", "CJX", "CLP", "GZP", "FLY", "KYX", "MOO", "CLP", "MUL", "MIL", "THF") ~ "Small pelagics",
      TRUE ~ NA_character_
    ),
    fish_group = ifelse(fish_group %in% c("Shrimps", "Molluscs", "Cephalopods", "Crustaceans"), "Marine invertebrates", fish_group),
  )

tot_catch_plot <-
  base_plot %>%
  dplyr::group_by(catch_name) %>% # Added .drop false to ensure that all factor levels remain in plot
  dplyr::summarise(
    catch_t = dplyr::first(catch),
    tot_t = dplyr::first(tot)
  )
# base_plot %>% View()

base_plot <-
  dplyr::left_join(base_plot, tot_catch_plot, by = "catch_name") %>%
  dplyr::filter(catch_t > 0) %>%
  dplyr::mutate(
    VarColor = dplyr::case_when(
      fish_group == "Small pelagics" ~ "#bc4749",
      fish_group == "Small demersals" ~ "#e9c46a",
      fish_group == "Large pelagics" ~ "#2a9d8f",
      fish_group == "Large demersals" ~ "#007ea7",
      fish_group == "Sharks and rays" ~ "#264653",
      fish_group == "Marine invertebrates" ~ "#898989",
      TRUE ~ "#000000"
    ),
    ColoredVar = glue::glue("<span style='color:{VarColor}'>{catch_name}</span>"),
    fish_group = factor(fish_group, levels = c(
      "Small pelagics", "Large pelagics", "Small demersals", "Large demersals",
      "Sharks and rays", "Marine invertebrates"
    ))
  )

tons_label <- cowplot::draw_label("Mean annual\ncatch (tons)", x = 0.941, y = 0.98, size = 9)

long_plot <-
ggplot2::ggplot() +
  ggpubr::theme_pubr() +
  ggchicklet::geom_chicklet(base_plot,
                            mapping = ggplot2::aes(
                              y = rdi, x = reorder(ColoredVar, tot),
                              fill = nutrient,
                              color = nutrient
                            ),
                            position = ggplot2::position_stack(reverse = FALSE),
                            alpha = 0.8,
                            width = 1
  ) +
  geom_text(base_plot,
            mapping = aes(y = rdi, x = reorder(ColoredVar, tot), label = round(rdi, 2) * 100),
            position = position_stack(0.5),
            color = "white",
            size = 3
  ) +
  geom_text(
    data = base_plot,
    mapping = aes(
      # Set y to a fixed position above the bars
      y = 1.35,  # Adjust this value if needed
      x = reorder(ColoredVar, tot_t),
      label = ifelse(catch_t < 1, "<1", scales::comma(round(catch_t, 0))),
      alpha = sqrt(catch_t),
    ),
    size = 3,
    hjust = 0,        # Left-align the text
    nudge_y = 0.05
  ) +
  facet_grid(fish_group ~ ., scales = "free", space = "free") +
  coord_flip(expand = FALSE, ylim = c(0, 1.5)) +
  ggplot2::scale_fill_manual(values = timor.nutrients::palettes$nutrients_palette) +
  ggplot2::scale_color_manual(values = timor.nutrients::palettes$nutrients_palette) +
  ggplot2::scale_y_continuous(labels = scales::percent, n.breaks = 10) +
  ggplot2::labs(x = "", y = "", fill = "Nutrient", subtitle = "Groups nutrient density") +
  ggplot2::theme(
    legend.position = "bottom",
    panel.grid.major.y = element_line(linetype = "dashed"),
    strip.background = ggplot2::element_blank(),
    strip.text = ggplot2::element_blank(),
    axis.text.y = ggtext::element_markdown(size = 9),
    panel.spacing = unit(0.1, "lines"),
    plot.subtitle = element_text(face = "bold")
  ) +
  guides(
    alpha = "none",
    color = "none"
  )

long_legend <- ggpubr::get_legend(long_plot)
long_plot <- long_plot +
  theme(legend.position = "none")

long_plot <- cowplot::plot_grid(long_plot) + tons_label


# groups_palette <- c("#2a9d8f", "#264653", "#007ea7", "#e9c46a", "#898989", "#bc4749")
groups_palette <- c("#bc4749", "#2a9d8f", "#e9c46a", "#007ea7", "#264653", "#898989")


groups_plot <-
  timor.nutrients::region_stats %>%
  dplyr::mutate(fish_group = dplyr::case_when(
    grouped_taxa %in% c("COZ") ~ "Molluscs",
    grouped_taxa %in% c("PEZ") ~ "Shrimps",
    grouped_taxa %in% c("MZZ") ~ "Other",
    grouped_taxa %in% c("SLV", "CRA") ~ "Crustaceans",
    grouped_taxa %in% c("OCZ", "IAX") ~ "Cephalopods",
    grouped_taxa %in% c("SKH", "SRX") ~ "Sharks and rays",
    grouped_taxa %in% c("SNA", "GPX", "PWT", "GRX", "MUI", "BGX") ~ "Large demersals",
    grouped_taxa %in% c("CGX", "TUN", "BEN", "LWX", "BAR", "SFA", "CBA", "DOX", "ECN", "DOS") ~ "Large pelagics",
    grouped_taxa %in% c("YDX", "SPI", "EMP", "SUR", "TRI", "MOJ", "WRA", "BWH", "LGE", "MOB", "MHL", "GOX", "THO", "IHX", "APO", "IHX", "PUX", "DRZ") ~ "Small demersals",
    grouped_taxa %in% c("RAX", "SDX", "CJX", "CLP", "GZP", "FLY", "KYX", "MOO", "CLP", "MUL", "DSF", "MIL", "THF") ~ "Small pelagics",
    TRUE ~ NA_character_
  )) %>%
  dplyr::select(-grouped_taxa) %>%
  dplyr::group_by(fish_group) %>%
  dplyr::summarise(dplyr::across(dplyr::where(is.numeric), ~ sum(.x, na.rm = T))) %>%
  na.omit() %>%
  tidyr::pivot_longer(-c(fish_group, catch), names_to = "nutrient") %>%
  dplyr::group_by(nutrient, fish_group) %>%
  dplyr::summarise(value = sum(value)) %>%
  dplyr::arrange(-value, .by_group = TRUE) %>%
  dplyr::select(fish_group, nutrient, value) %>%
  dplyr::ungroup() %>%
  dplyr::mutate(
    nutrient = as.factor(nutrient),
    fish_group = ifelse(fish_group %in% c("Shrimps", "Molluscs", "Cephalopods", "Crustaceans"), "Marine invertebrates", fish_group),
    fish_group = factor(fish_group, levels = c(
      "Shrimps", "Molluscs", "Cephalopods", "Crustaceans",
      "Marine invertebrates", "Sharks and rays", "Large demersals",
      "Small demersals", "Large pelagics", "Small pelagics"
    ))
  ) %>%
  dplyr::filter(!nutrient == "selenium") %>%
  dplyr::mutate(
    nutrient = stringr::str_to_title(nutrient),
    nutrient = dplyr::case_when(
      nutrient == "Omega3" ~ "Omega-3",
      nutrient == "Vitamina" ~ "Vitamin-A",
      TRUE ~ nutrient
    ),
    fish_label = dplyr::case_when(
      fish_group == "Small pelagics" ~ "SP",
      fish_group == "Large pelagics" ~ "LP",
      fish_group == "Small demersals" ~ "SD",
      fish_group == "Large demersals" ~ "LD",
      fish_group == "Sharks and rays" ~ "SR",
      fish_group == "Marine invertebrates" ~ "MI"
    ),
    fish_label = factor(fish_label, levels = rev(c(
      "SP", "LP", "SD", "LD", "SR", "MI"
    )))
  ) %>%
  ggplot(aes(fish_label, value / 1000, fill = fish_label)) +
  ggpubr::theme_pubr() +
  ggchicklet::geom_chicklet(
    alpha = 0.8,
    radius = grid::unit(5, "pt")
  ) +
  facet_wrap(. ~ nutrient, scales = "free", ncol = 2) +
  labs(y = "Tons (2018-2023)", x = "Functional\ngroup", fill = "Fish group", subtitle = "Nutrient yield") +
  theme(
    legend.position = "",
    panel.grid = element_blank(),
    strip.background = ggplot2::element_blank(), # This removes the background from the facet titles
    axis.text.y = element_text(size = 7),
    axis.ticks.y = ggplot2::element_blank(),
    panel.spacing.x = unit(0.1, "lines"),
    panel.spacing.y = unit(0, "lines"),
    plot.subtitle = element_text(face = "bold")
  ) +
  scale_fill_manual(
    values = groups_palette,
    breaks = c(
      "SP", "LP", "SD", "LD",
      "SR", "MI"
    )
  ) +
  scale_y_continuous(n.breaks = 3) +
  coord_flip(expand = F)

###

groups_palette <- c("#bc4749", "#2a9d8f", "#e9c46a", "#007ea7", "#264653", "#898989")

top_groups <-
  timor.nutrients::region_stats %>%
  dplyr::select(grouped_taxa, catch) %>%
  dplyr::mutate(
    tot_catch = sum(catch, na.rm = T),
    fish_group = dplyr::case_when(
      grouped_taxa %in% c("COZ") ~ "Molluscs",
      grouped_taxa %in% c("PEZ") ~ "Shrimps",
      grouped_taxa %in% c("MZZ") ~ "Other",
      grouped_taxa %in% c("SLV", "CRA") ~ "Crustaceans",
      grouped_taxa %in% c("OCZ", "IAX") ~ "Cephalopods",
      grouped_taxa %in% c("SKH", "SRX") ~ "Sharks and rays",
      grouped_taxa %in% c("SNA", "GPX", "PWT", "GRX", "MUI", "BGX") ~ "Large demersals",
      grouped_taxa %in% c("CGX", "TUN", "BEN", "LWX", "BAR", "SFA", "CBA", "DOX", "ECN", "DOS") ~ "Large pelagics",
      grouped_taxa %in% c("YDX", "SPI", "EMP", "SUR", "TRI", "MOJ", "WRA", "BWH", "LGE", "MOB", "MHL", "GOX", "THO", "IHX", "APO", "IHX", "PUX", "DRZ") ~ "Small demersals",
      grouped_taxa %in% c("RAX", "SDX", "CJX", "CLP", "GZP", "FLY", "KYX", "MOO", "CLP", "MUL", "DSF", "MIL", "THF") ~ "Small pelagics",
      TRUE ~ NA_character_
    )
  ) %>%
  dplyr::mutate(fish_group = ifelse(fish_group %in% c("Shrimps", "Molluscs", "Cephalopods", "Crustaceans"), "Marine invertebrates", fish_group)) %>%
  dplyr::group_by(fish_group, grouped_taxa) %>%
  dplyr::summarise(
    tot_catch = dplyr::first(tot_catch),
    catch = sum(catch, na.rm = T),
    catch_percent = catch / tot_catch * 100
  ) %>%
  dplyr::group_by(fish_group) %>%
  dplyr::slice_max(catch_percent, n = 1) %>%
  dplyr::arrange(-catch_percent, .by_group = T) %>%
  na.omit() %>%
  dplyr::rename(interagency_code = "grouped_taxa") %>%
  dplyr::left_join(timor.nutrients::catch_groups, by = "interagency_code") %>%
  dplyr::select(fish_group, catch_name, interagency_code, catch_percent) %>%
  dplyr::ungroup()

revenue_dat <-
  timor.nutrients::catch_data %>%
  tidyr::unnest(landing_catch) %>%
  tidyr::unnest(length_frequency) %>%
  dplyr::filter(!is.na(.data$weight), !is.na(landing_value)) %>%
  dplyr::group_by(landing_id) %>%
  dplyr::mutate(n = dplyr::n()) %>%
  dplyr::filter(n == 1) %>%
  dplyr::select(-n) %>%
  dplyr::ungroup() %>%
  dplyr::mutate(tot_obs = length(unique(landing_id))) %>%
  dplyr::group_by(catch_taxon) %>%
  dplyr::filter(catch_taxon %in% top_groups$interagency_code) %>%
  dplyr::group_by(.data$landing_id, .data$catch_taxon) %>%
  dplyr::summarise(
    landing_value = dplyr::first(landing_value),
    dplyr::across(
      c(.data$weight:.data$Vitamin_A_mu),
      ~ sum(.x)
    )
  )

models <-
  revenue_dat %>%
  dplyr::ungroup() %>%
  dplyr::select(catch_taxon, landing_value, weight) %>%
  # dplyr::group_by(catch_taxon) %>%
  # dplyr::mutate(model = broom::augment(stats::lm(formula = log(landing_value + 1) ~ log(weight / 1000 + 1)))) %>%
  # dplyr::ungroup() %>%
  # dplyr::mutate(cooksd = .data$model$`.cooksd`) %>%
  # dplyr::select(-.data$model) %>%
  # dplyr::mutate(out = dplyr::case_when(.data$cooksd > (5 * mean(.data$cooksd)) ~ 1, TRUE ~ 0)) %>%
  # dplyr::filter(out == 0) %>%
  # dplyr::select(-out, -cooksd) %>%
  dplyr::select(catch_taxon, landing_value, weight) %>%
  split(.$catch_taxon) %>%
  purrr::map(~ lm(log(landing_value) ~ log(weight), data = .x)) %>%
  purrr::map(~ predict(.x, data.frame(weight = log(1000)))) %>%
  dplyr::bind_rows(.id = "catch_taxon") %>%
  dplyr::rename(price = `1`) %>%
  dplyr::mutate(price = round(exp(price) * 10, 2))


revenue_dat <-
  revenue_dat %>%
  rename_nutrients_mu() %>%
  tidyr::pivot_longer(-c(landing_id:weight), names_to = "nutrient") %>%
  dplyr::filter(!nutrient == "selenium") %>%
  dplyr::left_join(RDI_tab) %>%
  dplyr::ungroup() %>%
  dplyr::mutate(
    value = value / weight * 100,
    inds = value / conv_factor * 100
  ) %>%
  dplyr::group_by(landing_id, catch_taxon) %>%
  dplyr::mutate(inds_100 = sum(inds)) %>%
  dplyr::group_by(catch_taxon) %>%
  dplyr::summarise(
    inds = dplyr::first(inds_100)
  ) %>%
  dplyr::left_join(models) %>%
  dplyr::rename(interagency_code = catch_taxon) %>%
  dplyr::left_join(timor.nutrients::catch_groups, by = "interagency_code") %>%
  dplyr::left_join(top_groups, by = "catch_name") %>%
  dplyr::select(fish_group, catch_name, price, inds, catch_percent)

revenue_plot <-
  revenue_dat %>%
  dplyr::mutate(fish_group = factor(fish_group, levels = c(
    "Small pelagics", "Large pelagics", "Small demersals", "Large demersals",
    "Sharks and rays", "Marine invertebrates"
  ))) %>%
  # dplyr::filter(!catch_name %in% c("Stingrays")) %>%
  ggplot(aes(inds, price, size = catch_percent, color = fish_group)) +
  ggpubr::theme_pubr() +
  geom_point(stroke = 2, alpha = 0.5) +
  geom_point(size = 1) +
  geom_text(aes(label = catch_name, color = fish_group), size = 3.5, vjust = -2.2, show.legend = FALSE) +
  scale_size(range = c(1, 15)) +
  coord_cartesian(
    ylim = c(0, 12),
    xlim = c(65, 145)
  ) +
  scale_x_continuous(labels = scales::label_percent(scale = 1), n.breaks = 5) +
  scale_y_continuous(labels = scales::label_dollar(), n.breaks = 8) +
  scale_color_manual(values = groups_palette) +
  theme(
    panel.grid = element_blank(),
    legend.position = "bottom",
    plot.subtitle = element_text(face = "bold")
  ) +
  guides(
    size = "none",
    colour = guide_legend(override.aes = list(size = 5))
  ) +
  labs(
    x = "",
    y = "Price (USD / kg)",
    subtitle = "Affordability and nutrient density",
    color = "Functional\ngroup"
  )


groups_leg <- ggpubr::get_legend(revenue_plot)
revenue_plot <-
  revenue_plot +
  theme(legend.position = "none")

habitat_nutrients <-
  timor.nutrients::kobo_trips %>%
  dplyr::select(habitat, weight:Vitamin_A_mu) %>%
  rename_nutrients_mu() %>%
  tidyr::pivot_longer(-c(habitat, weight), names_to = "nutrient") %>%
  dplyr::left_join(RDI_tab) %>%
  dplyr::mutate(
    value = value / weight,
    inds_kg = (value * 1000) / conv_factor
  ) %>%
  dplyr::group_by(habitat, nutrient) %>%
  dplyr::summarise(inds_kg = median(inds_kg, na.rm = T) / 10) %>%
  dplyr::mutate(inds_kg = inds_kg * 100) %>%
  dplyr::filter(!nutrient == "selenium") %>%
  dplyr::mutate(
    habitat = ifelse(habitat == "Deep", "Pelagic", habitat),
    nutrient = stringr::str_to_title(nutrient),
    nutrient = dplyr::case_when(
      nutrient == "Omega3" ~ "Omega-3",
      nutrient == "Vitamina" ~ "Vitamin-A",
      TRUE ~ nutrient
    )
  )
habitat_plot <-
  ggplot2::ggplot() +
  ggpubr::theme_pubr() +
  ggchicklet::geom_chicklet(habitat_nutrients,
    mapping = ggplot2::aes(
      y = inds_kg,
      x = reorder(habitat, inds_kg),
      fill = nutrient,
      color = nutrient
    ),
    position = ggplot2::position_stack(reverse = FALSE),
    alpha = 0.8,
    width = 0.8,
    radius = grid::unit(3, "pt")
  ) +
  ggplot2::geom_text(habitat_nutrients,
    mapping = ggplot2::aes(
      y = inds_kg,
      x = reorder(habitat, inds_kg),
      label = round(inds_kg, 0),
      group = nutrient
    ),
    position = ggplot2::position_stack(0.5, reverse = FALSE),
    color = "white",
    size = 2.5
  ) +
  ggplot2::scale_fill_manual(values = timor.nutrients::palettes$nutrients_palette) +
  ggplot2::scale_color_manual(values = timor.nutrients::palettes$nutrients_palette) +
  ggplot2::scale_y_continuous(labels = scales::label_percent(scale = 1)) +
  ggplot2::coord_flip(expand = FALSE) +
  ggplot2::theme(
    legend.position = "",
    panel.grid = ggplot2::element_blank(),
    axis.text.y = element_text(size = 7.75),
    plot.subtitle = element_text(face = "bold")
  ) +
  ggplot2::labs(x = "", fill = "", y = "Cumulative nutrient density score", subtitle = "Habitat and nutrient density") +
  ggplot2::guides(
    alpha = "none",
    color = "none",
    fill = guide_legend(override.aes = list(size = 7))
  )

gear_nutrients <-
  timor.nutrients::kobo_trips %>%
  dplyr::select(gear_type, weight:Vitamin_A_mu) %>%
  rename_nutrients_mu() %>%
  tidyr::pivot_longer(-c(gear_type, weight), names_to = "nutrient") %>%
  dplyr::left_join(RDI_tab) %>%
  dplyr::mutate(
    value = value / weight,
    inds_kg = (value * 1000) / conv_factor
  ) %>%
  dplyr::group_by(gear_type, nutrient) %>%
  dplyr::summarise(inds_kg = median(inds_kg, na.rm = T) / 10) %>%
  dplyr::mutate(inds_kg = inds_kg * 100) %>%
  dplyr::filter(!nutrient == "selenium") %>%
  dplyr::mutate(
    nutrient = stringr::str_to_title(nutrient),
    nutrient = dplyr::case_when(
      nutrient == "Omega3" ~ "Omega-3",
      nutrient == "Vitamina" ~ "Vitamin-A",
      TRUE ~ nutrient
    )
  ) %>%
  dplyr::mutate(gear_type = stringr::str_to_title(gear_type)) %>%
  dplyr::ungroup()

gear_nutrients <-
  ggplot2::ggplot() +
  ggpubr::theme_pubr() +
  ggchicklet::geom_chicklet(gear_nutrients,
    mapping = ggplot2::aes(
      y = inds_kg,
      x = reorder(gear_type, inds_kg),
      fill = nutrient,
      color = nutrient
    ),
    position = ggplot2::position_stack(reverse = FALSE),
    alpha = 0.8,
    width = 0.8,
    radius = grid::unit(3, "pt")
  ) +
  ggplot2::geom_text(gear_nutrients,
    mapping = ggplot2::aes(
      y = inds_kg,
      x = reorder(gear_type, inds_kg),
      label = round(inds_kg, 0),
      group = nutrient
    ),
    position = ggplot2::position_stack(0.5, reverse = FALSE),
    color = "white",
    size = 2.5
  ) +
  ggplot2::scale_fill_manual(values = timor.nutrients::palettes$nutrients_palette) +
  ggplot2::scale_color_manual(values = timor.nutrients::palettes$nutrients_palette) +
  ggplot2::scale_y_continuous(labels = scales::label_percent(scale = 1)) +
  ggplot2::coord_flip(expand = FALSE) +
  ggplot2::theme(
    legend.position = "",
    panel.grid = ggplot2::element_blank(),
    axis.text.y = element_text(size = 7.75),
    plot.subtitle = element_text(face = "bold")
  ) +
  ggplot2::labs(x = "", fill = "", y = "Cumulative nutrient density score", subtitle = "Gear type and nutrient density") +
  ggplot2::guides(
    alpha = "none",
    color = "none",
    fill = guide_legend(override.aes = list(size = 7))
  )


second_row_plot <- cowplot::plot_grid(
  groups_plot, #+ theme(plot.margin = unit(c(0, 0, 0, +2.2), "cm")),
  revenue_plot,
  ncol = 1,
  align = "hv",
  labels = c("B", "C"),
  rel_heights = c(1.2, 1)
)

p1 <-
  cowplot::plot_grid(
    long_plot,
    second_row_plot,
    ncol = 2,
    align = "hv",
    rel_widths = c(1.75, 1),
    labels = "AUTO"
  )

bottom_row_plot <- cowplot::plot_grid(
  habitat_plot + labs(y = ""),
  gear_nutrients + labs(y = ""),
  ncol = 2,
  align = "hv",
  labels = c("", ""),
  rel_heights = c(1, 1)
)

p2 <- cowplot::plot_grid(
  p1 + theme(plot.margin = unit(c(0, 0, 0, 0), "cm")),
  bottom_row_plot + theme(plot.margin = unit(c(-0.5, 0, 0, 0), "cm")),
  nrow = 2,
  labels = c("A", "D"),
  rel_heights = c(1, 0.25),
  align = "v" # align vertically
)

legends <-
  cowplot::plot_grid(
    long_legend, groups_leg
  )

x_label <- cowplot::draw_label("Cumulative Nutrient Density Score (NDS)", x = 0.55, y = 0.0825)

cowplot::plot_grid(
  p2,
  legends + theme(plot.margin = unit(c(0.19, 0, 0, -2), "cm")),
  nrow = 2,
  align = "h",
  rel_heights = c(1.3, 0.1)
) +
  x_label

Code

parallel_plot <-
  timor.nutrients::kobo_trips %>%
  dplyr::mutate(habitat = ifelse(habitat == "Deep", "Pelagic", habitat)) %>% 
  dplyr::select(habitat, gear_type, Selenium_mu:Vitamin_A_mu) %>%
  rename_nutrients_mu() %>%
  tidyr::pivot_longer(-c(habitat, gear_type), names_to = "nutrient", values_to = "concentration_g") %>%
  dplyr::group_by(habitat, gear_type, nutrient) %>%
  dplyr::summarise(concentration_g = sum(concentration_g, na.rm = T)) %>%
  # get relative values
  dplyr::group_by(nutrient) %>%
  dplyr::mutate(
    nutrient_sum = sum(concentration_g),
    concentration_g = concentration_g / nutrient_sum * 100
  ) %>%
  dplyr::ungroup() %>%
  dplyr::mutate(dplyr::across(habitat:nutrient, ~ as.factor(.x))) %>%
  dplyr::mutate(
    nutrient = as.factor(nutrient)
  ) %>%
  dplyr::mutate(
    nutrient = stringr::str_to_title(nutrient),
    nutrient = dplyr::case_when(
      nutrient == "Omega3" ~ "Omega-3",
      nutrient == "Vitamina" ~ "Vitamin-A",
      TRUE ~ nutrient
    )
  ) %>%
  dplyr::rename(
    "Nutrient" = nutrient,
    "Habitat" = habitat,
    "Gear type" = gear_type
  ) %>%
  ggforce::gather_set_data(c(3, 1, 2))

parallel_plot$x <- factor(parallel_plot$x,
                          levels = c("3", "1", "2"),
                          labels = c("Nutrient", "Habitat", "Gear type")
)

parallel_plot$y <- factor(parallel_plot$y, levels = c(
  "Protein", "Omega-3", "Calcium", "Iron", "Vitamin-A", "Zinc", "Selenium",
  "Reef", "Pelagic", "FAD", "Beach", "Traditional FAD", "Mangrove", "Seagrass",
  "gill net", "seine net", "spear gun", "hand line", "long line", "manual collection",
  "beach seine", "trap", "cast net"
))

parallel_plot %>%
  dplyr::filter(!Nutrient == "Selenium") %>%
  na.omit() %>%
  ggplot(aes(x, id = id, split = y, value = concentration_g)) +
  ggforce::geom_parallel_sets(aes(fill = Nutrient), alpha = 0.7, axis.width = 0.1) +
  ggforce::geom_parallel_sets_axes(axis.width = 0.25, fill = "grey95") +
  ggforce::geom_parallel_sets_labels(colour = "grey30", angle = 0, size = 3.5) +
  scale_x_discrete(name = NULL, expand = c(0, 0.2)) +
  scale_y_continuous(breaks = NULL, expand = c(0.05, 0)) +
  theme_minimal(12) +
  scale_fill_manual(values = timor.nutrients::palettes$nutrients_palette) +
  # scale_fill_manual(values = c("#508AA8", "grey95", "#BD9391")) +
  theme(
    axis.line = element_blank(),
    axis.ticks = element_blank(),
    legend.position = "bottom",
    panel.grid.major = element_blank(),
    axis.text = element_text(size = 12, color = "black")
  ) +
  labs(fill = "")