Draft abstract. Small pelagic fisheries may contribute to the resilience of coastal communities in coral reef systems by diversifying the fishery portfolio available to households. As with diversification in other risk-prone systems, the value of this portfolio effect may not be higher average returns, but lower exposure to shocks affecting a single resource base. In small-scale fisheries, where income volatility, food insecurity, governance constraints, and climate exposure often overlap, reducing risk is central to resilience. This article uses Rare’s Impact, Monitoring, Evaluation, and Learning (IMEL) Framework to connect fishery portfolio composition with social resilience metrics in community-led fisheries management areas. Using historical catch and Household Survey data from the Philippines, Indonesia, and Mozambique, we classify the contribution of small pelagic taxa to recorded catch and relate these patterns to social equity, trust in local leadership, collective efficacy, participation, financial sufficiency, and food security. Because the IMEL Framework is being implemented in 2026, the historical data used here were not originally collected with the sampling coverage and precision required for full management use. The analysis should therefore be read as a proof of concept: it shows how the IMEL approach can connect ecological, fisheries, and social information to support adaptive management, while also illustrating why fit-for-purpose sampling is necessary for future resilience monitoring.

Central claim. Small pelagic fisheries can contribute to community resilience when they expand the portfolio of resources available to coastal households, but their resilience value depends on effective, equitable, and legitimate co-management. Fishery diversification is therefore not only an ecological or economic condition; it is also a social and institutional challenge.

Introduction

Coastal communities in coral reef systems often depend on a mix of reef-associated and pelagic fisheries for food, income, and cultural continuity. This mix matters because different fisheries are exposed to different ecological, climatic, market, and governance pressures. A community that can access both reef-associated species and small pelagics may have a more diversified fishery portfolio than a community dependent on a narrower resource base.

The idea is intuitive: diversification can reduce risk. In finance, a diversified portfolio may not always produce the highest possible return, but it can reduce exposure to a single shock. A similar logic can apply to small-scale fisheries. If one resource becomes temporarily less available, less profitable, or more heavily regulated, access to another resource may help households smooth food and income. In contexts where social vulnerability and climatic risk exposure are high, risk reduction itself is a resilience outcome.

However, diversification does not automatically produce resilience. A more diversified fishery portfolio can still generate conflict, unequal benefits, overharvesting, or unstable incomes if the social and institutional foundations of management are weak. The resilience value of small pelagics therefore depends on whether communities can govern access, participate in decisions, trust local leadership, manage rules legitimately, and maintain household well-being under changing conditions.

Rare’s Impact, Monitoring, Evaluation, and Learning (IMEL) Framework provides a practical way to make this argument measurable. The framework treats resilience as multidimensional: social conditions such as equity, trust, participation, collective efficacy, financial sufficiency, and food security are measured alongside ecological and fisheries dimensions such as fisheries productivity and the protection of critical habitats. In this article, the emphasis is on social metrics, while catch composition is used to represent the fishery portfolio context in which those social conditions operate.

Why this analysis is a proof of concept

Rare is implementing the IMEL Framework in 2026 as a more structured approach for measuring resilience and using evidence in adaptive management. This matters for the interpretation of the analysis. The historical catch and Household Survey data used here are valuable, but they were collected before the current IMEL system was designed and were not originally intended to meet its full management requirements.

Data-readiness framing. The analysis below demonstrates the analytical logic of the IMEL approach, but it should not be interpreted as a final assessment of resilience across sites. Historical catch and HHS collections do not necessarily provide the coverage, timing, precision, or consistency required for a management-ready resilience framework. The 2026 IMEL sampling design is intended to address this gap by producing data with sufficient coverage and precision to support more reliable ABM-level interpretation and adaptive management.

This distinction changes the purpose of the article. The main contribution is not to claim that past data can already provide a complete resilience assessment. The contribution is to show how the framework can be used: combining fishery portfolio information with social metrics to identify resilience opportunities, vulnerabilities, and management questions that future IMEL data collection can answer more rigorously.

Analytical approach

This article asks three questions:

  1. Where are small pelagics important in recorded fishery portfolios?
  2. What is the social resilience profile of those communities?
  3. How can combining catch composition and social metrics support adaptive management?

The analysis is descriptive and exploratory. It is not designed to estimate the causal effect of small pelagic fisheries on community resilience. Instead, it develops an empirical basis for using the IMEL Framework as a management and learning tool.

Unit of analysis. The main unit is the area-based management unit (ABM), summarized from ABM-year observations where catch and Household Survey data overlap. Catch data are used to describe fishery portfolio composition. Household Survey responses are used to estimate social resilience indicators. The ABM-year join is used for reproducibility, but the main article figures emphasize ABM-level patterns because the goal is to support management interpretation.

Data and measures

The analysis uses two data sources. The first is a catch dataset containing recorded catch weight, value, taxonomic information, date, fisher identifier, and ABM name. The second is a Household Survey dataset containing selected social and livelihood indicators. The catch data are used to characterize fishery portfolio composition. The Household Survey data are used to estimate social resilience indicators.

After cleaning and restricting the analysis to the Philippines, Indonesia, and Mozambique, the catch dataset contains 189,211 records across 77 ABMs and approximately 4,664,196 kg of recorded catch. A strict small-pelagic classification represents 14% of recorded catch weight, while a broader pelagic grouping represents 33%.

The Household Survey dataset contains 29,023 responses in the same three countries. Of these, 15,969 responses could be linked to an ABM using the matching approach in this document. The final joined dataset contains 50 ABM-year observations, of which 31 ABM-year observations across 20 ABMs meet the initial analytical screen used for the main figures.

Fishery portfolio composition

Small pelagic contribution is measured as the share of recorded catch weight classified as strict small pelagic taxa. The strict classification focuses on small schooling pelagic taxa such as sardines, anchovies, selected scads, mackerel scads, and related groups. A broader pelagic classification is retained as a sensitivity metric but is not used as the main result.

The main fishery portfolio variables are:

  • Strict small pelagic share: strict small pelagic catch weight divided by total recorded catch weight.
  • Broad pelagic share: broader pelagic catch weight divided by total recorded catch weight.
  • Catch diversity: Shannon diversity of recorded catch families by ABM-year.

All catch records not classified as strict small pelagic or broader pelagic are grouped as other recorded catch in the figures. This keeps the visual focus on the role of small pelagics within the broader fishery portfolio.

This article focuses on catch composition rather than CPUE. Fisheries productivity can be incorporated in future iterations of the analysis, but the central question here is whether small pelagics form part of a diversified fishery portfolio and how that portfolio context relates to social resilience.

Social resilience metrics

The social metrics are derived from six Household Survey fields. Each indicator is converted to a 0–100 score where higher values represent more favorable resilience conditions.

indicator_table <- tibble::tribble(
  ~`Resilience dimension`, ~`Article indicator`, ~`Household Survey field`,
  "Collective action", "Social equity", "g8_fishery_benefit_equal",
  "Collective action", "Trust in local leadership", "g8_trust_local_decision",
  "Collective action", "Collective efficacy", "g8_my_community_ability",
  "Collective action", "Empowerment and participation", "g12_agreement_community_participation",
  "Livelihoods and food security", "Financial sufficiency", "g13_hh_ends_meet",
  "Livelihoods and food security", "Food security", "g11_food_worry"
)

knitr::kable(indicator_table)
Resilience dimension Article indicator Household Survey field
Collective action Social equity g8_fishery_benefit_equal
Collective action Trust in local leadership g8_trust_local_decision
Collective action Collective efficacy g8_my_community_ability
Collective action Empowerment and participation g12_agreement_community_participation
Livelihoods and food security Financial sufficiency g13_hh_ends_meet
Livelihoods and food security Food security g11_food_worry

Three composite scores are used for interpretation:

  • Collective action score: average of social equity, trust, collective efficacy, and participation.
  • Livelihood and food security score: average of financial sufficiency and food security.
  • Social resilience score: average of all six social indicators.

These composites are used as analytical summaries for this article. They are meant to support interpretation and visualization, not replace the underlying indicators.

Results

Small pelagics are unevenly distributed across fishery portfolios

Recorded catch composition differs across countries and ABMs. Some ABMs show substantial small pelagic contribution, while others are dominated by other recorded catch groups. This variation is important because it means small pelagics are not a uniform feature of all coral reef community fisheries. They are better understood as one possible component of a broader fishery portfolio.

catch_country_summary %>%
  mutate(
    fishery_group = factor(
      fishery_group,
      levels = c(
        "Strict small pelagic",
        "Other pelagic / mobile schooling",
        "Other recorded catch"
      )
    )
  ) %>%
  ggplot(aes(x = country, y = share, fill = fishery_group)) +
  geom_col(width = 0.72, position = position_stack(reverse = TRUE)) +
  scale_fill_manual(values = fishery_palette, breaks = names(fishery_palette)) +
  scale_y_continuous(labels = percent_format(accuracy = 1), expand = expansion(mult = c(0, 0.03))) +
  labs(
    x = NULL,
    y = "Share of recorded catch weight",
    fill = NULL,
    title = "Recorded catch composition varies across countries",
    subtitle = "Small pelagics are one component of a broader fishery portfolio."
  ) +
  theme_article()
Figure 1. Recorded catch composition by country. Strict small pelagics are highlighted in blue and placed at the bottom of each bar to make their contribution easier to compare.

Figure 1. Recorded catch composition by country. Strict small pelagics are highlighted in blue and placed at the bottom of each bar to make their contribution easier to compare.

At the ABM level, the distribution is more informative than the country average. ABMs with a higher small pelagic share may represent places where communities have access to a more diversified fishery portfolio. ABMs with lower small pelagic contribution may still be resilient, but their resilience may depend more strongly on reef-associated fisheries, non-fishery livelihoods, or other local conditions.

catch_abm_summary %>%
  filter(total_catch_kg > 0) %>%
  ggplot(aes(x = strict_small_pelagic_share, y = reorder(ma_name, strict_small_pelagic_share))) +
  geom_col(width = 0.72, fill = article_blue) +
  facet_wrap(~ country, scales = "free_y") +
  scale_x_continuous(labels = percent_format(accuracy = 1), expand = expansion(mult = c(0, 0.03))) +
  labs(
    x = "Strict small pelagic share of recorded catch weight",
    y = NULL,
    title = "Small pelagic contribution differs sharply across ABMs",
    subtitle = "This variation creates a useful basis for examining fishery portfolio diversification."
  ) +
  theme_article(base_size = 10)
Figure 2. Strict small pelagic share of recorded catch weight by ABM. ABMs are grouped by country and ordered within each panel by small pelagic contribution.

Figure 2. Strict small pelagic share of recorded catch weight by ABM. ABMs are grouped by country and ordered within each panel by small pelagic contribution.

Social resilience varies across dimensions, not only across places

The Household Survey indicators show that social resilience is multidimensional. A community can score relatively high on participation but lower on financial sufficiency, or relatively high on collective efficacy but lower on food security. This is why a resilience framework is useful: it prevents the analysis from reducing community resilience to a single outcome or to catch composition alone.

For readability, the figure below aggregates available analytical observations to the ABM level. Where an ABM appears in more than one year, scores are averaged across years using Household Survey sample size as the weight.

social_long_abm %>%
  mutate(
    indicator = factor(
      indicator,
      levels = c(
        "Social equity",
        "Leadership trust",
        "Collective efficacy",
        "Participation",
        "Financial sufficiency",
        "Food security"
      )
    )
  ) %>%
  ggplot(aes(x = indicator, y = reorder(ma_name, social_resilience_score), fill = score)) +
  geom_tile(color = "white", linewidth = 0.25) +
  facet_wrap(~ country, scales = "free_y") +
  scale_fill_gradient(limits = c(0, 100), low = "#f7fbff", high = article_blue_dark, na.value = "gray90") +
  labs(
    x = NULL,
    y = NULL,
    fill = "Score",
    title = "Social resilience profiles differ across ABMs",
    subtitle = "The same community can show strengths in one resilience dimension and weaknesses in another."
  ) +
  theme_article(base_size = 10) +
  theme(axis.text.x = element_text(angle = 35, hjust = 1))
Figure 3. Social resilience indicators by ABM. Scores are aggregated across available analytical years using HHS sample size as the weight. Each cell reports a score from 0 to 100, with higher values indicating more favorable resilience conditions.

Figure 3. Social resilience indicators by ABM. Scores are aggregated across available analytical years using HHS sample size as the weight. Each cell reports a score from 0 to 100, with higher values indicating more favorable resilience conditions.

The country-level summary below is included only to orient the reader. The main interpretation should remain at the ABM level, because community resilience and management decisions are organized locally.

hhs_summary_country <- social_abm %>%
  group_by(country) %>%
  summarise(
    `ABMs` = n(),
    `Median social resilience` = median(social_resilience_score, na.rm = TRUE),
    `Median collective action` = median(collective_action_score, na.rm = TRUE),
    `Median livelihoods and food security` = median(livelihood_food_security_score, na.rm = TRUE),
    .groups = "drop"
  )

knitr::kable(hhs_summary_country, digits = 1)
country ABMs Median social resilience Median collective action Median livelihoods and food security
IDN 9 59.4 68.2 40.5
MOZ 7 65.6 81.8 19.4
PHL 4 53.5 74.5 12.7

Fishery portfolio composition is not the same thing as resilience

The main empirical result is not that higher small pelagic contribution automatically corresponds to higher social resilience. In the analytical ABM-year dataset, the rank correlation between strict small pelagic share and the overall social resilience score is 0.1. This should be interpreted as an exploratory association, not as evidence of impact.

This finding is useful. It suggests that small pelagics should not be treated as a simple resilience shortcut. Instead, they should be treated as a potential diversification pathway whose value depends on the social and institutional conditions surrounding the fishery.

analysis_main %>%
  ggplot(aes(x = strict_small_pelagic_share, y = social_resilience_score)) +
  geom_point(aes(size = hhs_n), alpha = 0.75, color = article_blue) +
  geom_smooth(method = "lm", se = FALSE, linewidth = 0.8, color = article_gray_dark) +
  facet_wrap(~ country) +
  scale_x_continuous(labels = percent_format(accuracy = 1)) +
  scale_y_continuous(limits = c(0, 100)) +
  labs(
    x = "Strict small pelagic share of recorded catch weight",
    y = "Social resilience score",
    size = "HHS n",
    title = "Small pelagic contribution does not automatically translate into social resilience",
    subtitle = "The relationship is exploratory and should be interpreted as a management-learning signal."
  ) +
  theme_article()
Figure 4. Relationship between strict small pelagic contribution and social resilience. Points are ABM-years in the analytical sample. The line is a descriptive linear smoother and should not be interpreted causally.

Figure 4. Relationship between strict small pelagic contribution and social resilience. Points are ABM-years in the analytical sample. The line is a descriptive linear smoother and should not be interpreted causally.

Looking across individual social indicators provides a more nuanced picture. This helps identify whether small pelagic contribution is more closely aligned with some social dimensions than others, such as collective efficacy, participation, financial sufficiency, or food security.

analysis_main %>%
  select(
    country,
    ma_name,
    year,
    strict_small_pelagic_share,
    social_equity,
    leadership_trust,
    collective_efficacy,
    empowerment_participation,
    financial_sufficiency,
    food_security
  ) %>%
  pivot_longer(
    cols = c(
      social_equity,
      leadership_trust,
      collective_efficacy,
      empowerment_participation,
      financial_sufficiency,
      food_security
    ),
    names_to = "indicator",
    values_to = "score"
  ) %>%
  mutate(
    indicator = recode(
      indicator,
      social_equity = "Social equity",
      leadership_trust = "Leadership trust",
      collective_efficacy = "Collective efficacy",
      empowerment_participation = "Participation",
      financial_sufficiency = "Financial sufficiency",
      food_security = "Food security"
    ),
    indicator = factor(
      indicator,
      levels = c(
        "Social equity",
        "Leadership trust",
        "Collective efficacy",
        "Participation",
        "Financial sufficiency",
        "Food security"
      )
    )
  ) %>%
  ggplot(aes(x = strict_small_pelagic_share, y = score)) +
  geom_point(alpha = 0.65, color = article_blue) +
  geom_smooth(method = "lm", se = FALSE, linewidth = 0.7, color = article_gray_dark) +
  facet_wrap(~ indicator) +
  scale_x_continuous(labels = percent_format(accuracy = 1)) +
  scale_y_continuous(limits = c(0, 100)) +
  labs(
    x = "Strict small pelagic share of recorded catch weight",
    y = "Indicator score",
    title = "The portfolio-resilience relationship differs by social dimension"
  ) +
  theme_article(base_size = 10)
Figure 5. Strict small pelagic contribution and individual social resilience indicators. This figure helps identify which dimensions may be most relevant for follow-up interpretation.

Figure 5. Strict small pelagic contribution and individual social resilience indicators. This figure helps identify which dimensions may be most relevant for follow-up interpretation. 

correlation_table <- correlation_results %>%
  filter(fishery_metric == "strict_small_pelagic_share") %>%
  mutate(
    `Social metric` = recode(
      social_metric,
      social_equity = "Social equity",
      leadership_trust = "Leadership trust",
      collective_efficacy = "Collective efficacy",
      empowerment_participation = "Participation",
      financial_sufficiency = "Financial sufficiency",
      food_security = "Food security",
      collective_action_score = "Collective action score",
      livelihood_food_security_score = "Livelihoods and food security score",
      social_resilience_score = "Social resilience score"
    )
  ) %>%
  select(`Social metric`, `Complete ABM-years` = n_complete, `Spearman rho` = spearman_rho) %>%
  arrange(desc(abs(`Spearman rho`)))

knitr::kable(correlation_table, digits = 2)
Social metric Complete ABM-years Spearman rho
Food security 31 -0.24
Collective action score 31 0.22
Participation 31 0.20
Livelihoods and food security score 31 -0.18
Financial sufficiency 31 -0.14
Social equity 31 0.13
Social resilience score 31 0.07
Collective efficacy 31 0.07
Leadership trust 31 -0.01

Interpretation. The absence of a strong simple association is not a weak result. It supports a more realistic argument: fishery portfolio diversification may create resilience potential, but social and institutional conditions determine whether that potential becomes resilience in practice. It also reinforces the need for the 2026 IMEL sampling design: legacy data can illustrate the logic, while fit-for-purpose data are needed to guide management with greater confidence.

An initial IMEL lens for adaptive management

The value of this analysis is not only in describing patterns. It also shows how IMEL metrics can be used for adaptive management. By combining fishery portfolio composition and social resilience scores, managers can begin to distinguish between different types of situations that may require different responses.

The figure below is intentionally presented as an initial management lens, not as a definitive classification. It uses median cutoffs in the ABM-level analytical dataset to separate ABMs into four broad signals:

  • Diversification opportunity: higher small pelagic contribution and higher social resilience.
  • Potential dependence risk: higher small pelagic contribution but lower social resilience.
  • Strong social foundation: lower small pelagic contribution but higher social resilience.
  • Foundational resilience priority: lower small pelagic contribution and lower social resilience.
typology_df %>%
  ggplot(aes(x = strict_small_pelagic_share, y = social_resilience_score)) +
  geom_vline(xintercept = pelagic_cutoff, linetype = "dashed", color = "gray45") +
  geom_hline(yintercept = social_cutoff, linetype = "dashed", color = "gray45") +
  geom_point(aes(size = hhs_n_total), alpha = 0.75, color = article_blue) +
  facet_wrap(~ country) +
  scale_x_continuous(labels = percent_format(accuracy = 1)) +
  scale_y_continuous(limits = c(0, 100)) +
  labs(
    x = "Strict small pelagic share of recorded catch weight",
    y = "Social resilience score",
    size = "HHS n",
    title = "A practical IMEL lens for adaptive management",
    subtitle = "Combining portfolio composition and social resilience helps distinguish opportunity from vulnerability."
  ) +
  theme_article()
Figure 6. Initial management-learning lens combining fishery portfolio composition and social resilience. Dashed lines show sample medians. This is intended as a practical discussion tool, not a final classification.

Figure 6. Initial management-learning lens combining fishery portfolio composition and social resilience. Dashed lines show sample medians. This is intended as a practical discussion tool, not a final classification.

This lens helps move the management question away from a narrow focus on whether small pelagics are present. A more useful question is whether communities have the social foundations needed to manage a diversified fishery portfolio fairly and adaptively.

For example, a site with high small pelagic contribution and strong social resilience may be a good candidate for more advanced adaptive management, monitoring, or value-chain work. A site with high small pelagic contribution but weaker social resilience may require more attention to participation, trust, benefit-sharing, and household risk before fishery portfolio diversification can be treated as a resilience asset.

Discussion

This article uses legacy data to demonstrate an approach that the IMEL Framework is designed to make more actionable from 2026 onward. That distinction is central. The historical data can show how catch composition and social metrics may be brought together, but they cannot fully deliver the management-ready precision and coverage that the framework now requires. The analysis should therefore be read as a foundation for future IMEL implementation, not as a final assessment of site performance.

Within that framing, the analysis supports three main points.

First, small pelagics should be understood as part of a broader fishery portfolio. Their contribution varies substantially across ABMs, which means their potential role in resilience is also context-specific. In some places, they may represent a meaningful additional resource base. In others, they may be marginal or not central to the recorded fishery portfolio.

Second, community resilience cannot be inferred from fishery composition alone. The social indicators show variation across equity, trust, collective efficacy, participation, financial sufficiency, and food security. These dimensions shape whether communities can convert access to multiple fisheries into more stable livelihoods and more adaptive management.

Third, the IMEL approach is useful because it links measurement to management. It allows teams to look at fishery portfolio composition alongside the social conditions that enable legitimate and adaptive co-management. It also provides a structure for integrating this social evidence with ecological and fisheries indicators, including fisheries productivity and critical habitat protection. This is especially important for economic growth strategies in small-scale fisheries, where growth that is not socially grounded can increase risk, inequality, or conflict.

The key implication is that small pelagic fisheries can contribute to resilience, but not automatically. Their contribution depends on whether diversification is governed through institutions that are trusted, participatory, equitable, and capable of adapting to change.

Conclusion

Small pelagic fisheries may help coastal communities in coral reef systems reduce risk by diversifying the fishery portfolio available for food and income. But diversification alone is not resilience. The resilience value of small pelagics depends on the social and institutional conditions that determine who benefits, who participates, how rules are adapted, and whether households are better able to withstand shocks.

By combining catch composition with social resilience metrics, Rare’s IMEL Framework provides a practical way to understand this relationship and use it for adaptive management. The analysis developed here should be treated as a proof of concept using historical data. Its main contribution is to show how social metrics can be integrated into fisheries strategies so that small pelagic fisheries are evaluated not only as a production opportunity, but as part of a broader community resilience agenda. As the IMEL Framework is implemented in 2026, improved sampling coverage and precision should allow this approach to move from demonstration toward more reliable management use.