Skip to main content

Maritimes Ecosystem Indicator Catalogue

12 Structure & Function: Top of the Food Web

Data Type: Tabular Data (within eco_indicators)

Spatial Scope: Maritimes

Duration 1970-2022

Source: Bundy et al. 2017

12.1 Introduction to Indicator

The eco_indicators dataset provides indicators of ecosystem structure with focus on the top of the food web (Bundy, Gomez, and Cook 2017).

  • Large Fish Indicator: Proportion of large fish (>35cm) in the community.
  • Proportion of Predators: Proportion of predatory fish (feeding on fish or non-planktonic invertebrates).

These indicators provide data on how the marine community is structured, and have relevance to top-down controls on the ecosystem.

12.2 View Data 


library(tidyr)
library(plotly)
library(stringr)

plotly_df <- data@data %>% inner_join(global_cols3)

# function to create plot with dropdown menu ------------------------------
make_topfoodweb_dropdown_plot <- function(df,
                                      year_col   = "year",
                                      region_col = "region",
                                      value_suffix = "_value") {

  # convert to long format
  long <- df %>%
    janitor::clean_names() %>%
    pivot_longer(
      cols = ends_with(value_suffix),
      names_to = "metric",
      values_to = "value"
    ) %>%
    # remove suffix
    mutate(
      metric = str_remove(metric, "_value")
    ) %>%
    # drop NAs (some regions don't have data for some variables or years)
    tidyr::drop_na(value)

  # find all metrics and regions
  metrics <- unique(long$metric)
  regions <- unique(long[[region_col]])

  # assign regions to consistent colors
 # region_colors <- setNames(hcl.colors(length(regions), palette = "Dark 3"), regions)

  # clean names for dropdown panels, helper
  pretty_label <- function(x) str_to_title(gsub("_", " ", x)) 

  # build plot -----------------
  p <- plot_ly()

  # Add bar traces: metric1 has region1..K, metric2 has region1..K, ...
  for (metric_i in seq_along(metrics)) {
    m <- metrics[metric_i]
    
    for (region_i in regions) {
      dat <- long %>%
        filter(metric == m, .data[[region_col]] == region_i) %>%
        group_by(.data[[year_col]],color, linetype, linewidth, region_group, region_group_label) %>%     # in case you have multiple rows per year
        summarise(value = sum(value), .groups = "drop") %>%
        arrange(.data[[year_col]])
      
      group_name <- unique(dat$region_group)
      color <- unique(dat$color)
      linetype <- unique(dat$linetype)
      width <- unique(dat$linewidth)
      
      # If a region truly has no data for that metric, add an empty trace
      # (keeps trace indexing stable)
      if (nrow(dat) == 0) {
        dat <- tibble::tibble(!!year_col := integer(0), value = numeric(0))
      }
      
      p <- p %>% add_lines(
        data = dat,
        x = ~.data[[year_col]],
        y = ~value,
        name = as.character(region_i),
        legendgroup = group_name,
        legendgrouptitle = list(
          text =
            ifelse(group_name == "ESS",
                   "Eastern Scotian Shelf Zones",
                   "Western Scotian Shelf Zones"
                    )),
        showlegend = (metric_i == 1),
        visible = (metric_i == 1),
        line = list(color = color, dash = linetype),
        hovertemplate = paste0("<b>", region_i,":</b> ","%{y:.3f}<extra></extra>")      )
      
    }
  }

  n_regions <- length(regions)
  n_traces  <- length(metrics) * n_regions

  buttons <- lapply(seq_along(metrics), function(metric_i) {
    vis <- rep(FALSE, n_traces)
    shl <- rep(FALSE, n_traces)

    idx_start <- (metric_i - 1) * n_regions + 1
    idx_end   <- metric_i * n_regions
    vis[idx_start:idx_end] <- TRUE
    shl[idx_start:idx_end] <- TRUE

    list(
      method = "update",
      args = list(
        list(visible = vis, showlegend = shl),
        list(
          title = pretty_label(metrics[metric_i]),
          yaxis = list(title =  pretty_label(metrics[metric_i]))
        )
      ),
      label = pretty_label(metrics[metric_i])
    )
  })

  p %>%
    layout(
      barmode = "stack",
      hovermode = "x unified",
      title = pretty_label(metrics[1]),
      xaxis = list(title = str_to_title(year_col)),  # keep one bar per year
      yaxis = list(title = pretty_label(metrics[1]), fixedrange = TRUE),
      legend = list(
        x = 1.02, xanchor = "left",
        y = 1,    yanchor = "top",
        groupclick = "toggleitem",
        itemdoubleclick = FALSE
      ),
      updatemenus = list(list(
        type = "dropdown",
        x = -.1, xanchor = "left",
        y = 1.15, yanchor = "top",
        buttons = buttons
      )),
      margin = list(r = 180, t = 80)
    )
}

# usage:
p <- make_topfoodweb_dropdown_plot(plotly_df)
p <- p %>% config(displayModeBar= F)
  
p

Figure 12.1: Top of the food web indicators in all NAFO divisions and Scotian Shelf regions overall. Use dropdown menu to select indicator, and click legend to toggle regions.

12.4 Relevance to Research and Stock Assessments

Fishing activities disproportionately target large fish and predator species, so top of the food web indicators are expected to respond to fishing activities in exploited systems.

In Atlantic Canada, removal of predator species via fishing activity can lead to trophic cascades, affecting marine ecosystems through top-down change (Frank et al. 2005). Declines of large-bodied predator species can result in “fishing down the food web” strategies in marine fisheries, which might result in a phase of increased landings or revenue from lower-trophic level species, but ultimately unsustainability for the marine ecosystem (Pauly et al. 1998).

12.5 Variable Definitions

variable description unit
year Year of trawl survey
region Region over which values are summarized
LargeFishIndicator_value Proportion of fish >35cm Proportion
PREDATORS_ALL_value Proportion of predators Proportion

12.6 Additional Data

Top of Food Web Variables (within eco_indicators) contain data for 4VN, 4VS, 4W, 4X, ESS, and WSS. All are shown on this page, but note that NAFO divisions are nested within Scotian Shelf regions.

12.7 Get the Data 

library(marea)
data('eco_indicators')
plot(eco_indicators)

References

Bundy, Alida, Catalina Gomez, and Adam M. Cook. 2017. “Guidance Framework for the Selection and Evaluation of Ecological Indicators.” Canadian Technical Report of Fisheries and Aquatic Sciences 3232. Dartmouth, Nova Scotia, Canada: Fisheries; Oceans Canada, Bedford Institute of Oceanography. https://waves-vagues.dfo-mpo.gc.ca/Library/40648886.pdf.
Frank, Kenneth T, Brian Petrie, Jae S Choi, and William C Leggett. 2005. “Trophic Cascades in a Formerly Cod-Dominated Ecosystem.” Science 308 (5728): 1621–23.
Pauly, Daniel, Villy Christensen, Johanne Dalsgaard, Rainer Froese, and Francisco Torres Jr. 1998. “Fishing down Marine Food Webs.” Science 279 (5352): 860–63.