Reassessing display behavior from Bels et al. (2025) given the complexity of anthropogenic hybridization and intraspecific diversity in Iguana iguana

This paper reanalyzes and critiques Bels et al. (2025) by demonstrating that their oversimplified classification of non-native *Iguana iguana* populations obscured significant behavioral variation driven by anthropogenic hybridization and intraspecific diversity, ultimately revealing that display patterns in Guadeloupe retain characteristics from their specific South American origins.

The Gist

Imagine the Caribbean islands of Guadeloupe as a bustling, crowded neighborhood where two different families of lizards live side-by-side.

One family is the Native Family (Iguana delicatissima), who have lived there for thousands of years. They are rare, endangered, and very particular about their traditions. The other family is the Newcomer Family (Iguana iguana), who arrived from South America. They are strong, adaptable, and have been moving in for over a century.

Because these families live so close together, they have started having children together. These mixed-heritage lizards are called Hybrids.

The Original Study: A Flawed Map

Recently, a group of scientists (Bels et al., 2025) tried to understand how these lizards communicate. Lizards talk by doing a specific dance called a "head bob" (nodding their heads up and down).

The original researchers looked at the lizards and said, "Okay, we have two groups: The Natives and The Newcomers." They put all the Newcomers and all the Hybrids into one big bucket labeled "Non-Native." They then compared the dance moves of the Natives against this big bucket.

The Problem: This was like trying to understand a complex family reunion by only asking, "Are you a guest or a host?" It ignored the fact that the "guests" were actually a mix of pure guests, half-guests, and people who had been there for generations. By lumping everyone together, they missed the subtle differences in how the different groups danced.

The New Study: A Better Look

The authors of this new paper (Van den Burg and Thibaudier) said, "Hold on, we need a better map." They took the original data and reorganized it into three distinct groups:

1. Pure Natives
2. Pure Newcomers (the ones with no Native blood)
3. Hybrids (the mixed family)

They also fixed a few mistakes in how the data was counted (like making sure everyone was counting the dance steps in the same order).

What They Found: The Dance Reveals the Truth

When they looked at the data with this new, three-group lens, the picture changed dramatically:

• The Dance is Unique: The Pure Natives have a very specific rhythm. The Pure Newcomers have a different rhythm. But the Hybrids? They have their own unique rhythm that is different from both parents.
• The "Bucket" Mistake: The original study missed this because they treated the Hybrids and Pure Newcomers as the same. It turns out, the Hybrids don't just copy the Newcomers; they have their own style.
• Where Did the Newcomers Come From? The researchers also looked at videos of lizards in South America (the Newcomers' home). They discovered that the lizards in Guadeloupe are dancing exactly like the lizards from French Guiana (their specific home), not like lizards from other parts of South America. This means the Newcomers brought their specific "dance moves" with them and haven't changed them much, even after mixing with the Natives.

The Big Takeaway: Why This Matters

Think of the Native lizards as a rare, ancient library of books. The Newcomers are a flood of new books. When they mix, the library gets messy.

The original study tried to clean up the library by saying, "We have the old books and the new books." But this new study says, "No, we have old books, new books, and mash-up books."

Why does this matter for saving the lizards?
If you want to protect the Native lizards, you need to know exactly who is who. If you think a Hybrid is just a "Newcomer," you might treat them the same way. But this study shows that Hybrids are a distinct group with their own behaviors.

The Lesson for Science:
When nature is messy and complex (like a family tree with lots of mixing), you can't use simple, black-and-white categories. You have to look at the shades of gray. If you oversimplify the data, you miss the real story.

In short: The lizards are doing a complex dance. The first scientists tried to count the steps by grouping everyone into two piles and got the rhythm wrong. The new scientists separated the piles into three, fixed the counting, and realized that the mixed-up lizards have their own unique beat. To save the rare Native lizards, we need to understand that beat.

Technical Summary

1. Problem Statement

The paper addresses critical conceptual and methodological flaws in a recent study by Bels et al. (2025), which compared Display-Action-Patterns (DAP), specifically head-bobbing behaviors, between the endangered native Iguana delicatissima and non-native Iguana iguana in the Guadeloupe Archipelago.

The authors identify three primary issues with the original study:

• Oversimplified Taxonomy: Bels et al. (2025) treated all non-native iguanas on the main islands (Basse-Terre and Grande-Terre) as a single "non-native" group. However, genetic data indicates these populations are a complex mix of pure non-native, hybrid, and introgressed individuals. Merging these groups obscures behavioral variation.
• Inconsistent Data Handling: The original study treated head-bob sequences inconsistently between groups. For I. delicatissima, they plotted the first head bob of every sequence; for non-natives, they right-aligned the data (plotting the last head bob of sequences with 8 bobs in the first column), making direct comparisons of sequential duration invalid.
• Biased Comparative Baselines: Bels et al. (2025) compared Guadeloupe non-natives to mainland I. iguana from Panama and Colombia (Clade II). However, genetic evidence suggests Guadeloupe non-natives originate from French Guiana (Clade IV). The authors argue that comparing Clade IV behavior to Clade II baselines introduced geographic bias, leading to incorrect hypotheses about behavioral evolution.

2. Methodology

The authors reanalyzed the raw data provided in Bels et al.'s supplementary materials using the R environment, adhering to the original statistical framework (Linear Mixed Models) while introducing necessary corrections.

• Data Reclassification: The "species" variable was restructured from a binary (Native vs. Non-native) to a three-level categorical variable:
  1. Native (I. delicatissima): Colibri site (CO).
  2. Hybrid: Malendure (MA) and Carangaise (CA) sites, where hybridization is known to occur.
  3. Non-native: Gosier site (GO), representing a population of pure non-native origin.
• Statistical Modeling:
  • Three response variables were analyzed: (1) Duration of individual head bobs, (2) Number of head bobs per DAP, and (3) Total duration of the DAP.
  • Models were built using log-transformed response variables.
  • Two competing explanatory variables were tested: "Species" (3 levels) vs. "Site" (4 levels).
  • Model fit was evaluated using Akaike Information Criterion (AIC).
  • Post-hoc pairwise comparisons were conducted using the emmeans package with Tukey adjustments.
• Data Correction: The authors corrected the sequential plotting error by ensuring all groups were plotted based on the chronological order of head bobs (1st, 2nd, 3rd, etc.) rather than relative position.
• External Validation: To address the geographic bias, the authors conducted a systematic review of online videos (YouTube, TikTok, etc.) from South American mainland populations (Clades II and IV). They analyzed the sequence of short vs. long head bobs in male I. iguana from French Guiana (Clade IV) and compared them to the Guadeloupe non-native population.

3. Key Contributions

• Refined Taxonomic Resolution: Demonstrated that treating hybrid populations as a distinct category is statistically necessary to accurately model behavioral traits in hybridizing systems.
• Correction of Data Visualization: Identified and rectified a critical error in how sequential data was plotted, allowing for valid comparisons of head-bob duration progression.
• Geographic Contextualization: Provided the first behavioral comparison between Guadeloupe non-natives and their true source population (French Guiana/Clade IV), rather than the previously used Clade II populations.
• Quantification of Individual Variation: Highlighted significant intra-individual variation in the number of head bobs per DAP, a metric previously ignored in the literature.

4. Results

A. Model Fit and Variable Importance

• Head Bob Duration: The model using "Site" as the explanatory variable had the best fit (lowest AIC), suggesting that local population admixture (hybridization levels) drives duration differences more than broad species labels.
• Number of Head Bobs: The model using "Species" (including the hybrid category) was the best fit, indicating that the number of bobs is a distinct trait separating native, hybrid, and non-native groups.
• Total DAP Duration: The model using "Species" (including hybrid) was the best fit.

B. Behavioral Differences

• Native vs. Non-Native/Hybrid: Native I. delicatissima exhibited significantly longer individual head bobs (specifically the first three) and fewer total head bobs per DAP compared to both hybrid and non-native groups.
• Hybrid vs. Non-Native: Significant differences were found between hybrids and pure non-natives:
  • Hybrids had longer 4th and 5th head bobs than non-natives.
  • Hybrids had a significantly shorter total DAP duration than non-natives.
• Sequential Pattern: Native I. delicatissima head bob duration decreased throughout the sequence. In contrast, non-native and hybrid groups showed increasing duration for the first few bobs.

C. Mainland Comparison (Clade IV)

• Analysis of online videos confirmed that Clade IV populations (French Guiana, Suriname, etc.) exhibit a DAP sequence starting with short head bobs followed by long head bobs.
• This pattern matches the Guadeloupe non-native population, contradicting the pattern of Clade II (Panama/Colombia) populations, which start with long bobs and end with short ones.

5. Significance and Implications

• Conservation Management: The study underscores that behavioral traits in hybridizing populations are not uniform. Conservation strategies for I. delicatissima must account for the distinct behaviors of hybrids versus pure non-natives, as these groups may interact differently with native populations.
• Hypothesis Refinement: The findings refute two hypotheses proposed by Bels et al. (2025) regarding the evolution of DAP in hybrids. Instead, the data supports two remaining possibilities:
  1. Sexual selection favors I. iguana-like DAPs in hybrids.
  2. Display behavior is not under sexual selection, and the observed I. iguana traits in hybrids are due to skewed introgression (genetic swamping).
• Methodological Standard: The paper establishes that studies on hybridization must explicitly acknowledge and categorize admixture levels. Ignoring the "hybrid" category leads to oversimplified models and potentially erroneous biological conclusions.
• Behavioral Plasticity: The high variation in the number of head bobs per individual suggests that DAP is not a rigid, fixed trait, necessitating future studies to record multiple DAPs per individual to ensure statistical robustness.

In conclusion, van den Burg and Thibaudier demonstrate that accurate behavioral analysis in complex hybrid zones requires precise genetic categorization, consistent data handling, and geographically appropriate comparative baselines. Their reanalysis reveals that non-native iguanas in Guadeloupe have retained the ancestral DAP characteristics of their Clade IV source populations, rather than evolving new traits through hybridization.