Skip to content

Analysis: why catch rates swing in 2018-19 (blacktips down, lemons up) #120

Open

Description

@Luen

What. On the overview's "Catch rate per season", blacktip catch rate drops in 2018-19 while sicklefin lemon catch rate jumps (and stays high in 2019-20).

First look (aggregate, real data)

  • Not a change in sampling: the same nurseries and months were sampled, in similar proportions to the seasons around it.
  • The swing survives adjustment. In a negative-binomial model with nursery and day of the season:
    • seasons differ more than chance for lemons (p < 0.001) and blacktips (p ≈ 0.03);
    • 2018-19 lemons ≈ 2.3× the average season (95% CI ≈ 1.7–3.1), and 2019-20 ≈ 2.0×;
    • 2018-19 blacktips ≈ 0.76× (≈ 0.59–0.97).
  • Where it happened: the lemon high is concentrated in the lemon nurseries (above all Vaiane, also Tiki), but lemons were above their usual rate elsewhere too.
  • No sign of displacement: across seasons at the same lemon nursery, lemon-rich seasons are not significantly poorer for blacktips (Spearman ρ −0.15 to −0.35, all p > 0.25).

Open questions for the lab

Acceptance

  • The overview chart states whether seasons differ beyond chance (adjusted test) and names the standout seasons, computed live.
  • This issue records the lab's answer on fieldwork changes.

Part of #72.

Activity

  1. added
    questionFurther information is requested
    analyticsCharts, statistics and analyses
    on Nov 3, 2026
  2. Luen commented on Nov 3, 2026

    @Luen
    CollaboratorAuthor

    PR #122 adds the adjusted season test to the overview's "Catch rate per season" chart. The chart now names the seasons whose 95% interval excludes the average, computed live: 2018–19 for both species, and 2019–20 for lemons. This issue stays open for the lab's answer: did anything change in the 2018–19 or 2019–20 fieldwork (net placement at Vaiane or Tiki, nets, timing)?

  3. Luen commented on Nov 3, 2026

    @Luen
    CollaboratorAuthor

    @JodieRummer, two questions about the 2018-19 swing. PR #122 now names 2018-19 (both species), plus 2019-20 for lemons, as seasons that differ beyond chance once nursery and day of the season are accounted for.

    1. Did anything change in the 2018-19 or 2019-20 fieldwork? For example: where the nets went within Vaiane or Tiki (see Data: review nursery alias table (Tiki / Tiki2 / Tiki NEW / Tiki (PEN), Papetoai old/Woody …) #13), the net itself (length, mesh), set length or timing, or the crew.
    2. Would you expect sicklefin lemon pups to rise for two seasons running? Possible reasons include a biennial breeding cycle, or a few females returning to pup at Vaiane.

    If the fieldwork didn't change, the overview will keep calling these unusual seasons and won't suggest a cause.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    analyticsCharts, statistics and analysesquestionFurther information is requested

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions