Critical Analysis of Fish Bone Data from Anakena Cove, Rapa Nui

Examining Claims of Subsistence Change Through Taphonomic and Methodological Lenses

Automatic Figure and Table Export

This document automatically saves all figures and tables as both PNG and SVG files when rendered. Files are saved to:

• PNG files: figures/png/
• SVG files: figures/svg/

These can be used for presentations, publications, or other documents.

1 Introduction

The narrative of progressive faunal resource depletion has dominated archaeological interpretations of Rapa Nui’s environmental transformation. Diamond (2005), Martinsson-Wallin and Crockford (2002), and Steadman et al. (1994) constructed a picture of ecological collapse based primarily on faunal assemblages from Anakena. According to these accounts, the archaeological record indicates a progression from an abundant seabird and offshore marine resource base to an impoverished subsistence base dependent on chickens and rats. This interpretation requires critical examination.

This analysis examines fish bone data from early deposits at Anakena Cove (A.D. 800-1000) to evaluate claims of temporal shifts in fishing practices. The central question is whether the data support interpretations of resource depletion or whether alternative explanations better account for observed patterns.

2 Data and Methods

2.1 Data Source

Analysis is based on Tables 7 and 8 from Martinsson-Wallin and Crockford (2002), representing fish remains from the early component at Anakena Cove, Trench C. The assemblage dates to approximately A.D. 800-1000.

Table 1: Fish bone data from early Anakena deposits

# Enter the fish bone data from Table 7
early_fish <- tibble(
  taxon = c("Engraulidae", "Holocentridae", "Scorpaenidae", "Serranidae",
            "Carangidae (undetermined)", "Alectis ciliaris", "Carangoides equula",
            "Decapterus scombrinus", "Scomberoides lysan", "Seriola dumerili",
            "Coryphaenidae", "Girellops nebulosus", "Kyphosus bigibbus",
            "Chaetodon litus", "Sphyraena helleri", "Bodianus oxycephalus",
            "Thalassoma purpureum", "Katsuwonus pelamis", "Thunnus albacares",
            "Gerreidae", "Unknown #3", "Unknown #4", "Unknown #5"),
  family = c("Anchovies", "Squirrelfishes", "Scorpionfishes", "Groupers",
             "Jacks", "Jacks", "Jacks", "Jacks", "Jacks", "Jacks",
             "Dorados", "Rudderfishes", "Rudderfishes", "Butterflyfishes",
             "Barracudas", "Wrasses", "Wrasses", "Tunas", "Tunas",
             "Mojarras", "Unknown", "Unknown", "Unknown"),
  mni = c(1, 2, 1, 1, 0, 3, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 2, 1, 2, 1, 1, 1),
  habitat = c("Coastal", "Inshore", "Inshore", "Inshore", "Variable", "Variable",
              "Variable", "Variable", "Variable", "Variable", "Offshore",
              "Inshore", "Inshore", "Inshore", "Coastal", "Inshore", "Inshore",
              "Offshore", "Offshore", "Coastal", "Unknown", "Unknown", "Unknown")
)

# Size distribution data from Table 8
size_distribution <- tibble(
  size_category = factor(c("Very Small (3-6\")", "Small (8-12\")", "Medium (16-20\")",
                    "Large (24-36\")", "Very Large (36-60\")"), 
                    levels = c("Very Small (3-6\")", "Small (8-12\")", "Medium (16-20\")",
                    "Large (24-36\")", "Very Large (36-60\")")),
  mni = c(0, 8, 11, 9, 2)
)

2.2 Analytical Approach

Fish taxa were classified by habitat preference (offshore/pelagic, inshore/reef, coastal, variable) based on species ecology. Minimum Number of Individuals (MNI) served as the primary quantification method. Analysis focused on:

1. Sample size adequacy
2. Taxonomic composition
   
3. Habitat representation
4. Size distribution patterns
5. Taphonomic considerations

3 Results

3.1 Sample Composition

# Calculate basic statistics
total_specimens <- 215
identified_specimens <- 56
total_mni <- sum(early_fish$mni)
identification_rate <- (identified_specimens / total_specimens) * 100

stats_summary <- tibble(
  Metric = c("Total Specimens (NSP)", "Identified Specimens", 
             "Identification Rate", "Total MNI"),
  Value = c(total_specimens, identified_specimens, 
            paste0(round(identification_rate, 1), "%"), total_mni)
)

# Save table as image
save_table(stats_summary, "table1_sample_stats", 
           "Table 1: Basic sample statistics")

quartz_off_screen 
                2 

stats_summary %>%
  kable(align = "lr") %>%
  kable_styling(bootstrap_options = c("striped"), full_width = FALSE)

Table 2: Basic sample statistics for early Anakena fish remains

Metric	Value
Total Specimens (NSP)	215
Identified Specimens	56
Identification Rate	26%
Total MNI	28

The assemblage exhibits a low identification rate (26%), with only 28 individuals identified from 215 specimens. This immediately raises concerns about sample representativeness.

3.2 Taxonomic Composition

# Aggregate by family
family_summary <- early_fish %>%
  group_by(family, habitat) %>%
  summarise(total_mni = sum(mni), .groups = 'drop') %>%
  arrange(desc(total_mni))

# Create family composition plot
p_taxonomic <- ggplot(family_summary %>% filter(total_mni > 0), 
       aes(x = reorder(family, total_mni), y = total_mni, fill = habitat)) +
  geom_bar(stat = "identity") +
  coord_flip() +
  scale_fill_manual(values = c("Offshore" = "#2166ac", 
                               "Inshore" = "#d6604d", 
                               "Coastal" = "#f4a582",
                               "Variable" = "#92c5de",
                               "Unknown" = "#bababa")) +
  labs(x = NULL, y = "Minimum Number of Individuals (MNI)",
       fill = "Habitat Type") +
  theme(legend.position = "bottom",
        panel.grid.major.y = element_blank())

# Save the plot
save_plot(p_taxonomic, "fig1_taxonomic_composition", width = 8, height = 6)

# Display the plot
p_taxonomic

Figure 1: Fish family representation in early deposits showing MNI by family and habitat association

Carangidae (jacks) dominate the assemblage with 7 MNI, followed by Labridae (wrasses) with 4 MNI. Most families are represented by only 1-3 individuals, limiting ecological interpretation.

3.3 Habitat Representation

# Aggregate by habitat zone
habitat_summary <- early_fish %>%
  filter(mni > 0) %>%
  group_by(habitat) %>%
  summarise(total_mni = sum(mni), .groups = 'drop') %>%
  mutate(percentage = round((total_mni / sum(total_mni)) * 100, 1))

# Reclassify for clearer analysis
habitat_grouped <- habitat_summary %>%
  mutate(habitat_group = case_when(
    habitat == "Offshore" ~ "Offshore/Pelagic",
    habitat == "Inshore" ~ "Inshore/Reef",
    habitat %in% c("Coastal", "Variable") ~ "Variable/Coastal",
    TRUE ~ "Unknown"
  )) %>%
  group_by(habitat_group) %>%
  summarise(mni = sum(total_mni), 
            percentage = sum(percentage), 
            .groups = 'drop')

# Create two panel figure
p1 <- ggplot(habitat_grouped, aes(x = habitat_group, y = mni)) +
  geom_bar(stat = "identity", fill = c("#2166ac", "#d6604d", "#bababa", "#92c5de")) +
  geom_text(aes(label = mni), vjust = -0.5, size = 4) +
  labs(y = "MNI", x = NULL,
       title = "A. Habitat distribution") +
  ylim(0, max(habitat_grouped$mni) * 1.15) +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

p2 <- ggplot(habitat_grouped, aes(x = "", y = percentage, fill = habitat_group)) +
  geom_bar(stat = "identity", width = 1) +
  coord_polar("y", start = 0) +
  scale_fill_manual(values = c("Offshore/Pelagic" = "#2166ac",
                               "Inshore/Reef" = "#d6604d",
                               "Unknown" = "#bababa",
                               "Variable/Coastal" = "#92c5de")) +
  labs(fill = "Habitat", title = "B. Proportional representation") +
  theme_void() +
  theme(legend.position = "right") +
  geom_text(aes(label = paste0(percentage, "%")), 
            position = position_stack(vjust = 0.5))

p_habitat <- p1 + p2

# Save the combined plot
save_plot(p_habitat, "fig2_habitat_distribution", width = 10, height = 5)

# Save individual panels as well
save_plot(p1, "fig2a_habitat_bars", width = 5, height = 5)
save_plot(p2, "fig2b_habitat_pie", width = 5, height = 5)

# Display the plot
p_habitat

Figure 2: Distribution of fish by generalized habitat zones

# Calculate effective sample sizes by habitat
habitat_n <- habitat_grouped %>%
  mutate(interpretation = case_when(
    mni < 5 ~ "Too small for inference",
    mni < 10 ~ "Very limited inference",
    mni < 20 ~ "Limited inference",
    TRUE ~ "Moderate inference possible"
  ))

# Save table as image
save_table(habitat_n, "table2_habitat_stats", 
           "Table 2: Statistical assessment by habitat zone")

quartz_off_screen 
                2 

habitat_n %>%
  kable(col.names = c("Habitat Zone", "MNI", "Percentage", "Statistical Assessment"),
        align = "lrrr") %>%
  kable_styling(bootstrap_options = c("striped"), full_width = FALSE)

Table 3: Statistical assessment of sample sizes by habitat zone

Habitat Zone	MNI	Percentage	Statistical Assessment
Inshore/Reef	10	35.7	Limited inference
Offshore/Pelagic	4	14.3	Too small for inference
Unknown	3	10.7	Too small for inference
Variable/Coastal	11	39.3	Limited inference

The assemblage shows mixed habitat exploitation, with no single zone dominating. All habitat categories have sample sizes too small for robust statistical inference.

3.4 Size Distribution

p_size <- ggplot(size_distribution, aes(x = size_category, y = mni)) +
  geom_bar(stat = "identity", fill = "#4575b4", alpha = 0.8) +
  geom_text(aes(label = mni), vjust = -0.5, size = 4) +
  labs(x = "Size Category", y = "MNI") +
  ylim(0, max(size_distribution$mni) * 1.15) +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

# Save the plot
save_plot(p_size, "fig3_size_distribution", width = 8, height = 5)

# Display the plot
p_size

Figure 3: Fish size distribution based on subjective size estimations

The size distribution shows predominance of small to medium specimens (8-20 inches), with few very large individuals. This pattern requires taphonomic evaluation before ecological interpretation.

4 Comparative Analysis: Early vs Late Deposits at Anakena

4.1 The Missing Comparison

While Martinsson-Wallin and Crockford (2002) claim to present data comparing early and late fish deposits, the tables provided in their paper show only early deposit data. However, Ayres (1985) provides the comparison they reference, with fish remains from both early (A.D. 1350-1500) and late (A.D. 1550-present) deposits at Anakena.

# Create comparison data from Ayres
anakena_early <- tibble(
  Family = c("Labridae", "Serranidae", "Muraenidae", "Congridae", "Brotulidae",
             "Balistidae", "Diodontidae", "Scaridae", "Ostraciontidae",
             "Holocentridae", "Unidentified"),
  Early_Percent = c(16.9, 16.9, 30.5, 1.7, 3.4, 1.7, 5.1, 3.4, 1.7, 1.7, 16.9),
  Early_MNI = c(10, 10, 18, 1, 2, 1, 3, 2, 1, 1, 10)
)

anakena_late <- tibble(
  Family = c("Labridae", "Serranidae", "Muraenidae", "Congridae", "Brotulidae",
             "Balistidae", "Diodontidae", "Scaridae", "Dasyatidae", "Ostraciontidae",
             "Scorpaenidae", "Holocentridae", "Unidentified"),
  Late_Percent = c(22.2, 11.1, 20.0, 3.7, 2.2, 2.2, 5.9, 1.5, 7.4, 4.4, 0.7, 4.4, 14.1),
  Late_MNI = c(30, 15, 27, 5, 3, 3, 8, 2, 10, 6, 1, 6, 19)
)

# Merge the datasets
anakena_comparison <- full_join(anakena_early, anakena_late, by = "Family") %>%
  replace_na(list(Early_Percent = 0, Early_MNI = 0, Late_Percent = 0, Late_MNI = 0)) %>%
  arrange(desc(Early_MNI + Late_MNI))

# Save table
save_table(anakena_comparison, "table5_anakena_comparison",
           "Table 5: Anakena fish remains comparison")

quartz_off_screen 
                2 

anakena_comparison %>%
  kable(col.names = c("Fish Family", "Early %", "Early MNI", "Late %", "Late MNI"),
        digits = 1, align = "lrrrr") %>%
  kable_styling(bootstrap_options = c("striped"), full_width = FALSE)

Table 4: Fish remains from early and late deposits at Anakena (Ayres 1985)

Fish Family	Early %	Early MNI	Late %	Late MNI
Muraenidae	30.5	18	20.0	27
Labridae	16.9	10	22.2	30
Unidentified	16.9	10	14.1	19
Serranidae	16.9	10	11.1	15
Diodontidae	5.1	3	5.9	8
Dasyatidae	0.0	0	7.4	10
Ostraciontidae	1.7	1	4.4	6
Holocentridae	1.7	1	4.4	6
Congridae	1.7	1	3.7	5
Brotulidae	3.4	2	2.2	3
Balistidae	1.7	1	2.2	3
Scaridae	3.4	2	1.5	2
Scorpaenidae	0.0	0	0.7	1

4.2 Visualizing the Comparison

# Prepare data for visualization
anakena_long <- anakena_comparison %>%
  select(Family, Early_MNI, Late_MNI) %>%
  pivot_longer(cols = c(Early_MNI, Late_MNI), 
               names_to = "Period", 
               values_to = "MNI") %>%
  mutate(Period = factor(str_replace(Period, "_MNI", ""),
                        levels = c("Early", "Late")))

# Create comparison plot
p_comparison <- ggplot(anakena_long, aes(x = reorder(Family, MNI), y = MNI, fill = Period)) +
  geom_bar(stat = "identity", position = "dodge", alpha = 0.8) +
  coord_flip() +
  scale_fill_manual(values = c("Early" = "#d73027", "Late" = "#4575b4"),
                    labels = c("Early (1350-1500)", "Late (1550-present)")) +
  labs(x = NULL, y = "Minimum Number of Individuals",
       title = "Fish Taxa Changes at Anakena") +
  theme(legend.position = "bottom")

# Save the plot
save_plot(p_comparison, "fig5_anakena_comparison", width = 10, height = 8)

# Display the plot
p_comparison

Figure 4: Comparison of fish family representation between early and late deposits at Anakena

4.3 Analysis of Changes

# Classify by general habitat
habitat_classification <- anakena_comparison %>%
  mutate(Habitat = case_when(
    Family %in% c("Labridae", "Muraenidae", "Congridae", "Serranidae", 
                  "Scaridae", "Brotulidae") ~ "Reef/Rocky",
    Family == "Dasyatidae" ~ "Sandy/Muddy",
    Family %in% c("Holocentridae", "Diodontidae", "Ostraciontidae",
                  "Balistidae", "Scorpaenidae") ~ "Mixed Reef",
    TRUE ~ "Unknown"
  )) %>%
  group_by(Habitat) %>%
  summarise(Early_Total = sum(Early_MNI),
            Late_Total = sum(Late_MNI), .groups = 'drop') %>%
  pivot_longer(cols = c(Early_Total, Late_Total),
               names_to = "Period",
               values_to = "MNI") %>%
  mutate(Period = str_replace(Period, "_Total", ""))

# Create habitat comparison
p_habitat <- ggplot(habitat_classification, aes(x = Habitat, y = MNI, fill = Period)) +
  geom_bar(stat = "identity", position = "dodge", alpha = 0.8) +
  scale_fill_manual(values = c("Early" = "#d73027", "Late" = "#4575b4")) +
  labs(x = "Habitat Type", y = "Total MNI",
       title = "Fish Habitat Representation Across Time") +
  theme(legend.position = "bottom")

# Save and display
save_plot(p_habitat, "fig6_habitat_persistence", width = 8, height = 6)
p_habitat

Figure 5: Persistence of inshore fishing focus across time periods

4.4 Key Findings from the Anakena Comparison

The actual data from Ayres (1985) reveals:

1. No evidence of offshore fishing in either period: Both assemblages completely lack pelagic species (no Scombridae, Coryphaenidae, or other offshore taxa)

2. Consistent inshore focus: Both periods are dominated by reef-associated species (Labridae, Muraenidae, Serranidae)

3. Minor compositional changes:

   • Slight increase in Labridae (wrasses): 16.9% → 22.2%
   • Decrease in Serranidae (groupers): 16.9% → 11.1%
   • Appearance of Dasyatidae (rays) in late deposit (7.4%)

4. Sample size differences: Late deposit has 2.3× more specimens (135 vs 59 MNI), suggesting:

   • More intensive shelter use
   • Longer accumulation period
   • Better preservation conditions
   • NOT necessarily dietary change

The data directly contradicts claims of a shift from offshore to inshore fishing. Instead, it shows remarkable continuity in inshore fishing practices with minor variations that could reflect local environmental changes, fishing spot preferences, or taphonomic factors.

4.5 Comparing Different Depositional Environments

A fundamental issue undermining temporal comparisons is the dramatic change in depositional context between early and late periods. If late deposits represent rapid dune accumulation, they constitute an entirely different formation process.

depositional_factors <- tibble(
  Factor = c("Deposition rate", "Sediment source", "Energy environment",
             "Preservation potential", "Activity representation",
             "Spatial extent", "Temporal resolution"),
  `Early Deposits` = c("Slow accumulation", "Mixed cultural/natural", 
                     "Low energy", "Higher", "Primary disposal",
                     "Concentrated", "High (palimpsest)"),
  `Late Dune Deposits` = c("Rapid accumulation", "Primarily aeolian",
                         "High energy", "Lower", "Secondary/disturbed",
                         "Dispersed", "Low (rapid burial)"),
  Comparability = c("Not comparable", "Not comparable", "Not comparable",
                    "Biased", "Different activities", "Different scales",
                    "Different processes")
)

# Save table as image
save_table(depositional_factors, "table3_depositional_factors", 
           "Table 3: Depositional context differences")

quartz_off_screen 
                2 

depositional_factors %>%
  kable(align = "llllc") %>%
  kable_styling(bootstrap_options = c("striped"), full_width = TRUE, font_size = 12) %>%
  column_spec(4, color = "darkred", bold = TRUE)

Table 5: Depositional differences between early and late contexts

Factor	Early Deposits	Late Dune Deposits	Comparability
Deposition rate	Slow accumulation	Rapid accumulation	Not comparable
Sediment source	Mixed cultural/natural	Primarily aeolian	Not comparable
Energy environment	Low energy	High energy	Not comparable
Preservation potential	Higher	Lower	Biased
Activity representation	Primary disposal	Secondary/disturbed	Different activities
Spatial extent	Concentrated	Dispersed	Different scales
Temporal resolution	High (palimpsest)	Low (rapid burial)	Different processes

4.6 Taphonomic Effects on Assemblage Composition

# Create preservation probability model
scenario_data <- tibble(
  Period = rep(c("Early deposits\n(slow accumulation)", 
                 "Late deposits\n(rapid dunes)"), each = 4),
  Fish_Type = rep(c("Large offshore", "Small offshore", 
                    "Large inshore", "Small inshore"), 2),
  Preservation_Probability = c(
    # Early deposits - better preservation overall
    0.7, 0.5, 0.6, 0.4,
    # Late deposits - size-biased preservation
    0.3, 0.1, 0.5, 0.2
  ),
  Original_Catch = rep(c(30, 20, 25, 25), 2)
) %>%
  mutate(Expected_Archaeological = Original_Catch * Preservation_Probability,
         Fish_Category = factor(Fish_Type, 
                               levels = c("Large offshore", "Small offshore",
                                         "Large inshore", "Small inshore")))

# Create visualization
p_preservation <- ggplot(scenario_data, aes(x = Fish_Category, y = Expected_Archaeological, 
                          fill = Period)) +
  geom_bar(stat = "identity", position = "dodge", alpha = 0.8) +
  geom_hline(yintercept = 5, linetype = "dashed", color = "red", alpha = 0.5) +
  annotate("text", x = 3.5, y = 6, label = "Detection threshold", 
           color = "red", size = 3) +
  labs(x = "Fish Type",
       y = "Expected Archaeological Visibility\n(Original catch × Preservation probability)",
       fill = "Depositional Context") +
  scale_fill_manual(values = c("#4575b4", "#d73027")) +
  theme(legend.position = "bottom",
        axis.text.x = element_text(angle = 45, hjust = 1))

# Save the plot
save_plot(p_preservation, "fig4_preservation_model", width = 10, height = 6)

# Display the plot
p_preservation

Figure 6: Model showing how identical fishing practices could produce different archaeological signatures due to depositional context

This model demonstrates how identical fishing behavior in both periods could produce an apparent shift from offshore to inshore species due solely to differential preservation in different depositional environments.

deposition_effects <- tibble(
  Aspect = c("Element size", "Element density", "Species representation",
             "Spatial association", "Cultural behavior"),
  Effect = c("Small elements lost in rapid deposition",
             "Light bones transported differently",
             "Robust inshore species overrepresented",
             "Mixing of unrelated deposits",
             "Different disposal patterns"),
  `Impact on Interpretation` = c("Size bias against small fish",
                               "Density bias against bird/fish bones",
                               "False habitat shift pattern",
                               "Loss of behavioral context",
                               "Cannot infer subsistence change")
)

# Save table as image
save_table(deposition_effects, "table4_taphonomic_effects", 
           "Table 4: Taphonomic effects of rapid dune deposition")

quartz_off_screen 
                2 

deposition_effects %>%
  kable(align = "lll") %>%
  kable_styling(bootstrap_options = c("striped"), full_width = TRUE) %>%
  column_spec(3, color = "darkred")

Table 6: Specific effects of rapid dune deposition on faunal assemblages

Aspect	Effect	Impact on Interpretation
Element size	Small elements lost in rapid deposition	Size bias against small fish
Element density	Light bones transported differently	Density bias against bird/fish bones
Species representation	Robust inshore species overrepresented	False habitat shift pattern
Spatial association	Mixing of unrelated deposits	Loss of behavioral context
Cultural behavior	Different disposal patterns	Cannot infer subsistence change

5 Discussion

5.1 Sample Size Limitations

With only 30 identified individuals, the early deposit sample lacks statistical power to characterize fishing practices, let alone detect temporal changes. The 26% identification rate further suggests significant information loss through poor preservation or analytical limitations.

5.2 Alternative Interpretations

Rather than representing baseline fishing practices, the observed patterns could reflect:

1. Opportunistic exploitation: Mixed offshore/inshore species suggest targeting seasonal aggregations rather than daily fishing patterns
2. Social deposition practices: Differential disposal based on status, ceremony, or specific events
3. Preservation biases: Systematic loss of smaller, more fragile elements
4. Sampling effects: Single site may not represent island-wide practices

5.3 The Depositional Context Problem

The comparison of early cultural deposits with late aeolian deposits violates basic principles of archaeological comparison. High-energy dune environments:

• Preferentially destroy small, light elements (most fish bones)
• Create size-sorted assemblages through winnowing
• Mix materials from different sources and times
• Break the relationship between behavior and archaeological patterning

Finding fewer offshore fish in late dune deposits provides no evidence for resource depletion when the assemblages are taphonomically incomparable.

6 Conclusions

The fish bone data from early Anakena deposits cannot support claims of temporal change in fishing practices or marine resource depletion. Critical issues include:

1. Inadequate sample sizes: 30 MNI from Martinsson-Wallin and Crockford is insufficient for robust analysis
2. No valid comparison: Their tables show only early deposit data, not the claimed comparison with late deposits
3. Taphonomic incomparability: Different depositional environments preclude direct comparison
4. Alternative explanations: Multiple non-ecological factors could produce observed patterns
5. Methodological overreach: Limited data interpreted through predetermined collapse narrative

When we examine the actual comparative data from Ayres (1985) for Anakena, we find: - No offshore fishing in either early or late periods - Consistent focus on inshore reef species across time - Minor variations in species proportions that do not indicate fundamental changes - Complete absence of pelagic species in both assemblages

The absence of deep-sea fish in late dune deposits cannot indicate changing fishing practices when: - The deposits themselves are fundamentally different in formation - Both early and late assemblages lack offshore species entirely - The “shift” from offshore to inshore fishing appears to be a narrative construction unsupported by data

Proper testing of the depletion hypothesis requires: - Comparable depositional contexts - Adequate sample sizes (minimum 100+ MNI per period) - Multiple sites across the island - Taphonomic controls and corrections - Independent environmental proxies - Integration with fishing technology data - Recognition that absence of offshore species may reflect technology/preference rather than depletion

Until these requirements are met, narratives of fishing depletion and ecological collapse remain unsupported speculation rather than demonstrated fact. The tendency to interpret limited archaeological data through collapse narratives should be replaced with more rigorous taphonomic and statistical analysis. The Anakena data, when properly examined, reveals continuity rather than collapse in marine subsistence practices.

7 References

Diamond, J. (2005). Collapse: How societies choose to fail or succeed. Viking Press.

Martinsson-Wallin, H., & Crockford, S. J. (2002). Early settlement of Rapa Nui (Easter Island). Asian Perspectives, 40(2), 244-278.

Steadman, D. W., Casanova, P. V., & Ferrando, C. C. (1994). Stratigraphy, chronology, and cultural context of an early faunal assemblage from Easter Island. Asian Perspectives, 33(1), 79-96.

8 Exported Figures and Tables

All figures and tables have been saved in the following directories:

• PNG files: figures/png/
• SVG files: figures/svg/

8.1 List of exported files:

Figures: - fig1_taxonomic_composition - Fish family representation by habitat - fig2_habitat_distribution - Combined habitat distribution plots - fig2a_habitat_bars - Habitat distribution bar chart - fig2b_habitat_pie - Habitat distribution pie chart - fig3_size_distribution - Fish size distribution - fig4_preservation_model - Differential preservation model

Tables: - table1_sample_stats - Basic sample statistics - table2_habitat_stats - Statistical assessment by habitat - table3_depositional_factors - Depositional context differences - table4_taphonomic_effects - Taphonomic effects of dune deposition