The Puzzle of Insect Decline in a Protected Paradise

In Costa Rica’s Guanacaste Conservation Area, an ecological paradise, a tragic ecological mystery has been unfolding. Despite decades of protection from pesticides, urbanization, and industrial pollution, scientists like Professor Daniel Janzen (University of Pennsylvania) and Dr. Winnie Hallwachs who have been diligently studying the insects of Guanacaste for decades, have observed a sharp decline in insect life. Butterflies, moths, beetles, ants, and wasps that once flourished in the vibrant tropical forest now seem to be vanishing. Conventional culprits like habitat loss, pesticides, and chemical exposure don’t fit at all in the highly protected environment. So what is causing this slow-motion collapse?

We propose a new hypothesis proposes that the answer lies not in direct local human interference, but in a far more subtle and insidious force: the rising levels of carbon dioxide (CO₂) in the global atmosphere, and the disproportionate impact this high and rising CO2 has on the soil and leaf litter habitats where most insects begin their lives. The once glorious and abundant insects of this tropical paradise are suffocating in their nurseries beneath the forest floor.

A Hidden Crisis Beneath the Forest Floor

Insects commonly spend the earliest and most vulnerable larval stages of their lives underground or within layers of decomposing leaf litter. These environments are biologically rich but physically constrained. Oxygen is limited, and carbon dioxide—produced by microbes, fungi, roots, and decomposing material—accumulates at high concentrations naturally. In tropical forests like Guanacaste, where heat and humidity supercharge decomposition, these subsurface zones are naturally close to insect tolerance and CO₂-saturated .

What has changed over the last 50 years is the atmosphere above. Atmospheric CO₂ has risen from around 330 ppm in the 1970s to over 420 ppm today. This seemingly small increase has a disproportionately large effect on subsurface environments because it reduces the perfectly understandable physics of diffusion gradients that formerly allowed CO₂ to escape the soil. The result today: biological CO₂ in forest floor soils is trapped, building up to toxic levels that impact insect larvae before they ever get a chance to fly.

This mechanism hinges on gas diffusion physics. Under normal conditions, CO₂ generated in the soil escapes upward into the atmosphere through tiny air spaces between soil particles. The driving force for this diffusion is the concentration gradient between the higher CO₂ levels underground and the lower CO₂ levels in the air. When the atmosphere was at 330 ppm, the soil CO₂ had a steep gradient to escape into, allowing excess gas to be released relatively efficiently.

However, as atmospheric CO₂ rises—now over 420 ppm—this gradient flattens. The ability of CO₂ to escape into the atmosphere slows down, causing it to accumulate within the soil. This is not a linear relationship; a small change in the upper boundary condition (air CO₂) can produce a much larger buildup below when biological production remains high, especially in warm, moist environments like tropical forests. The denser and wetter the soil or leaf litter, the slower the gas exchange, and the more extreme the CO₂ buildup becomes.

While this diffusion barrier effect may be the primary unseen force in subsurface larval mortality, it does not act in isolation. As Janzen and Hallwachs and others have noted, the tropical insects of Guanacaste are also vulnerable to a suite of interacting stressors. Habitat fragmentation, erratic rainfall patterns, and seasonal mismatches in flowering and fruiting cycles compound the physiological stress of CO₂ exposure. For example, reduced rainfall or altered timing of rains can dry out the upper leaf litter layers, reducing microbial activity and altering decomposition dynamics. These shifts affect the availability and chemical composition of food sources for detritivorous larvae.

Moreover, the collapse of some insect populations can destabilize entire food webs. Pollinators vanish, affecting plant reproduction. Leaf-eating insects decline, leaving parasitoids and predators with no hosts. These cascading interactions suggest that the ecological collapse is both deep and systemic.

As Janzen and Hallwachs described in their 2021 PNAS paper: *”The visible insects have disappeared; the once-familiar species are now the exception rather than the rule.”  They caution that “we are losing them for reasons that we cannot yet fully explain but must urgently investigate.”

Rising CO₂ may be the spark, but it burns in a field already parched by climate volatility and ecological fragmentation (Janzen & Hallwachs, 2021).

Modeled Insect Lethality Thresholds from Soil CO₂ Accumulation

To investigate this possibility, we modeled soil CO₂ concentrations from 1950 to 2025. The model assumes a steady biological respiration rate in tropical soils, but introduces a feedback where rising atmospheric CO₂ restricts escape, causing an exponential rise in subsurface concentrations.

The chart below illustrates this process:

• Blue line: Projected soil CO₂ concentrations from 1950 to 2025.
• Green dashed line (right axis): Atmospheric CO₂ levels over the same period.
• Threshold lines:
  • LD₁₀ (~6,000 ppm): The level at which 10% of insect larvae may suffer harm. Set to match 1970 levels, a period known to be ecologically healthy.
  • LD₅₀ (10,000 ppm): Lethal to 50% of insect larvae, based on studies of bees and stored-product pests.
  • LD₉₀ (13,500 ppm): An estimated point of 90% near-total mortality for most soil-dwelling insects.

 

 

 

 

 

 

 

 

Timeline Comparison: PNAS Observations vs. CO₂ Model

Period	Janzen & Hallwachs (2021)	CO₂ Mortality Model
Pre-1970s	Insect life abundant; ecological balance intact	Soil CO₂ below LD₁₀
1974–1990s	Early declines observed; loss of some once-common species begins	Soil CO₂ crosses LD₁₀ (~6,000 ppm) in 1974
2000s	Widespread disappearance of visible insects accelerates	Soil CO₂ breaches LD₅₀ (~10,000 ppm around 2007)
2010s–2020	Severe biodiversity loss; many species nearly or completely absent	Soil CO₂ exceeds LD₉₀ (~13,500 ppm by ~2020)

This alignment between direct observation and modeled soil conditions lends strong support to the hypothesis that rising atmospheric CO₂—through its effects on subsurface respiration environments—is a major driver of the insect collapse seen in Guanacaste.

 Why CO₂ is More Dangerous Than Temperature

It’s important to note that temperature change alone cannot explain the severity or timing of insect collapse. Costa Rica’s climate has warmed modestly over the past 50 years—perhaps by 1°C. However, subsurface CO₂ levels have doubled or tripled during the same time. Moreover, CO₂ acts as a direct physiological stressor: it alters insect respiration, disrupts acid-base balance, and impairs development. For species whose eggs or larvae reside in the soil for weeks or months, prolonged exposure is fatal.

Bee researchers have observed similar patterns: in modern hives with limited ventilation, in-hive CO₂ levels can reach 5,000 to 10,000 ppm, correlating with increased larval mortality and colony collapse. Our model applies this insight to natural systems, where the same gas exchange physics operates silently—and lethally.

Conclusion: A Tipping Point Crossed

The tragedy of insect collapse in Guanacaste is not the result of poisoning or bulldozing—it is the outcome of a slowly tightening atmospheric grip. As CO₂ increases globally, the subsurface world that nurtures insect life becomes a suffocating chamber.

We often think of carbon dioxide as a climate gas, but it is also a biological agent—one that alters the invisible chemistry of the soil. The story here is not of a single catastrophic event, but of an ecological tipping point quietly crossed. The insects did not flee. They simply failed to emerge.

To reverse this, we must see rising CO₂ not only as an atmospheric burden but as a biological toxin—one that kills from below, long before the air feels any warmer. In doing so, we might yet protect the tiny lives that sustain the web of life.

The one and only hope for these and most of our vital insect co-habitants on this planet is to immediately reduce atmospheric CO2, and that means working not in the forests which beautiful and vital are a tiny sliver of the CO2 cycle compared to the vast ocean plankton pastures of this Blue Planet. Only by restoring the ocean pastures can we save life in our forests.

Reference:
Janzen, D. H., & Hallwachs, W. (2021). To us insectometers, it is clear that insect decline in our Costa Rican tropical forest is real, so let’s be kind to the survivors. Proceedings of the National Academy of Sciences, 118(2), e2002546117. https://doi.org/10.1073/pnas.2002546117

Biological Conservation Volume 233, May 2019, Pages 102-108
Perspective: Where might be many tropical insects?
Daniel H. Janzen, Winnie Hallwachs
https://doi.org/10.1016/j.biocon.2019.02.030

The Bees Are Speaking: Asking Us to Let Them Breathe
Russ George 2025
https://russgeorge.net/2025/05/15/the-bees-are-speaking-asking-us-to-let-them-breathe/