Ecosystem Effects of Invasive Hippopotamus Populations in Colombia

Background & Ecological Context

• Human‐mediated defaunation and species introductions have reshaped global biogeography.
• Megafaunal extinction (e.g., end-Pleistocene losses in the New World & Australia) removed key nutrient vectors and ecosystem engineers.
• Large animals (marine mammals, migrating ungulates, anadromous fishes) historically transported nutrients both horizontally & vertically, influencing primary productivity and ecosystem states.
• Re-wilding debate: can non-native megafauna restore lost functions? Hippos in Colombia provide a real-world test of this idea.

The Colombian Hippo Invasion

• Species: Common Hippopotamus Hippopotamus amphibius.
• Origin: 4 individuals imported by Pablo Escobar to Hacienda Nápoles (Magdalena Medio, Antioquia) ~1980s; escaped after Escobar’s death (1993).
• Current estimate (2019): 65–80 individuals.
• Distribution: Small artificial lakes at Hacienda Nápoles + Magdalena River & tributaries (up to 150 km downstream).
• Growth projection (exponential, no control):
  • $N_{1993}=4$
  • $N_{2019}=65$
  • Annual intrinsic rate $r = \frac{\ln(65)-\ln(4)}{26\;\text{yr}} \approx 0.11\;\text{yr}^{-1}$ (≈11 % yr⁻¹)
  • $N<em>{2040}=785$, $N</em>{2060}=7089$ if $N<em>t = N</em>0 e^{rt}$ and density dependence absent.
• Social–ecological mismatch: charismatic appeal & tourism value hinder lethal control; management controversial nationally & internationally.

Research Questions & Hypotheses

• Do invasive hippos replicate their African role as ecosystem engineers in Neotropical aquatic systems?
• Predictions for lakes with hippos:
  • Heavier POM $\delta^{13}C$ (more terrestrial carbon).
  • Elevated eutrophication indicators & ecosystem metabolism (e.g., dissolved-O₂ swings).
  • Shifts in community composition (phyto-, bacterio-, zoo- & benthic invertebrates).

Study Area & Design

• Location: 14 artificial impoundments at Hacienda Nápoles (Averages: shallow, turbid, cattle-grazed catchments).
• Hippo status: 2 lakes with permanent populations (Lake 1: 20–30 hippos, 78 000 m²; Lake 10: 3–5 hippos, 13 000 m²); 12 lakes without hippos.
• Sampling campaigns: April 2017 (wet), July 2017 (dry), September 2018 (wet).
• All sampling conducted from shore (hippo hazard).

Measurements & Analytical Methods

• Water chemistry: chlorophyll-a, total C & N, pH, conductivity, Secchi depth.
• Stable isotopes: POM $\delta^{13}C$ & $\delta^{15}N$.
• Ecosystem metabolism: HOBO loggers recorded dissolved O₂ every 15 min; daily amplitude = mean(dusk) – mean(dawn).
• Biotic communities:
  • Phytoplankton: 60 mL Lugol-fixed, Utermöhl counting to genus/division.
  • Bacterioplankton: 16S-V4 amplicon sequencing (phylum → ASV resolution).
  • Zooplankton: 45 µm net tows, ethanol preserved, species-level IDs.
  • Macroinvertebrates: 10 × 1 m sweeps with dip net, family/genus IDs.
• Statistics: Two-way ANOVA (hippo presence × sampling date); PERMANOVA (community matrices, 1000 permutations); NMDS for ordination; rarefaction using vegan R package.

Key Numerical Findings

• Stable isotopes:
  • Mean $\delta^{13}C$: No-hippo lakes ≈ $-24.9\,‰$; Lake 10 ≈ $-22.7\,‰$; Lake 1 ≈ $-19.2\,‰$.
  • ANOVA: hippo presence significant (P < 0.01); $\delta^{15}N$ unaffected (P = 0.53).
• Dissolved O₂ daily range:
  • Hippo lakes: 3.6 mg L⁻¹ (mean), night minima often <4 mg L⁻¹ (fish-kill threshold).
  • No-hippo lakes: 2.1 mg L⁻¹; minima never <4 mg L⁻¹.
  • ANOVA: hippo presence significant; lake area dropped by stepwise AIC, indicating hippos stronger predictor than size.
• Other chemistry (chlorophyll-a mean 58.3 µg L⁻¹, range 5.6–248; total C, total N, pH, conductivity, Secchi) → no significant hippo effect (all P > 0.1).
• Phytoplankton:
  • Divisions dominated by Chlorophyta, Cyanophyta, Charophyta.
  • Cyanophyta proportion higher in hippo lakes; PERMANOVA hippo effect P = 0.02.
  • Richness (division, genus) similar; slight division-level increase in hippo lakes (P = 0.046).
• Bacteria, Zooplankton, Macroinvertebrates:
  • Dominant bacterial phyla: Proteobacteria, Planctomycetes, Actinobacteria, Bacteroidetes, Verrucomicrobia.
  • Community composition NS between lake categories (PERMANOVA P ≥ 0.56); strong temporal (seasonal) signal.
  • Zooplankton: rotifers, small cladocerans, copepods; no hippo effect.
  • Macroinvertebrates: Ephemeroptera, Hemiptera, annelids, ostracods; no hippo effect.

Interpretations & Significance

• Hippos act as cross-ecosystem nutrient & carbon vectors in Colombia, enriching lakes with terrestrial C and stimulating net metabolism.
• Enhanced daily O₂ swings (& occasional hypoxia) mirror African patterns but magnitude muted—likely due to:
  • Higher annual precipitation (46 – 187 % greater than East-African study sites) → stable water levels, lower hippo crowding.
  • Early invasion stage: population still well below carrying capacity.
• Community responses limited to phytoplankton (Cyanobacteria dominance), suggesting bottom-up effects currently modest.
• If population grows to projected hundreds/thousands, eutrophication, harmful algal blooms, and fish kills could escalate—posing threats to water quality & fisheries in the Magdalena Basin.

Management & Socio-Ethical Considerations

• Trade-offs:
  • Ecological risk (eutrophication, hypoxia, native species interactions) vs. Tourism & cultural fascination with ‘cocaine hippos’.
• Lethal control socially contentious; need for humane, publicly acceptable strategies (e.g., fertility control, translocation).
• Example of a broader “social–ecological mismatch” where charismatic exotics limit conventional invasive-species management.
• Policymakers must integrate ecological data, economic valuation, and public sentiment into adaptive management plans.

Broader Connections to Ecological Theory

• Supports theory that megafaunal nutrient transport is functionally conserved across biogeographic contexts.
• Provides empirical data for re-wilding debates: non-native megafauna CAN restore certain processes but may create disservices (eutrophication).
• Highlights context dependency (Subalusky & Post 2019): climatic seasonality, water residence time, and hippo density modulate ecosystem effects.
• Illustrates early-stage invasion dynamics: rapid exponential growth before density dependence manifests.

Equations & Statistical References

• Exponential growth: $N<em>t = N</em>0 e^{rt}$.
• Intrinsic rate estimation: $r = \frac{\ln(N<em>t) - \ln(N</em>0)}{\Delta t}$.
• Hypoxia threshold for fish: [O₂] < $4\,\text{mg L}^{-1}$.
• PERMANOVA & NMDS used for multivariate community analysis (999–1000 permutations).

Study Limitations & Future Research

• Only 2 lakes with permanent hippos ⇒ unbalanced design; effects may be underestimated/overestimated.
• Shore-based sampling constrained spatial coverage (safety), especially for deeper mid-lake metrics.
• Lack of wind speed & mixed-layer depth data prevented partitioning net vs. gross production & respiration.
• Future needs:
  • Quantitative regional surveys of hippo abundance & distribution.
  • Year-round high-frequency monitoring of nutrients, metabolism, and algal toxins.
  • Experimental manipulations or mesocosms to disentangle nutrient vs. organic-matter pathways.
  • Socio-economic studies on stakeholder perceptions and cost-benefit of management options.

Key Terminology

• Ecosystem Engineer: species that physically or chemically modifies habitats, affecting resource availability to other organisms.
• Subsidy Overload: excessive allochthonous inputs surpass ecosystem processing capacity, leading to negative outcomes (e.g., hypoxia).
• PERMANOVA: Permutational Multivariate Analysis of Variance, non-parametric test for community differences.
• $\delta^{13}C$ / $\delta^{15}N$: Stable isotope ratios indicating carbon/nitrogen source signatures.
• Rarefaction: Statistical standardization of richness to equalize sample size.

Take-Home Messages

• Even at early invasion stages, hippos measurably alter carbon flow and metabolic dynamics in Colombian lakes.
• Current effects modest but trajectory suggests mounting risk to water resources if population unchecked.
• Management must balance ecological integrity with socio-economic values, utilizing humane & scientifically informed strategies.
• Case study enriches understanding of megafaunal roles, rewilding potential, and social-ecological conflicts in conservation science.