Tropical Interiors: Why the Urban Heat Island Is the Wrong Model for Pest Ecology in a Subarctic City
The literature on urban warming was written in temperate cities where pavement can rescue a marginally hardy insect. Winnipeg is colder than that mechanism works, which makes the building envelope, not the city, the ecological boundary that matters
Abstract
Urban pest ecology is dominated by a single explanatory framework: the urban heat island, in which impervious surfaces raise local temperatures, accelerate development, improve overwintering survival and increase pest density. The evidence for this framework is strong and we review it. We then argue that it transfers poorly to Winnipeg, and that the reason is quantitative rather than conceptual. Work on overwintering survival found that impervious surfaces buffered egg mortality until temperatures reached or fell below roughly minus twenty two degrees Celsius, at which point the buffering failed. Winnipeg passes that threshold in an ordinary January. The urban heat island in a subarctic city does not convert a hostile outdoor winter into a survivable one; it makes an uninhabitable winter marginally less uninhabitable, which for an insect is the same thing. What does the work instead is the building envelope. We argue that Winnipeg's principal structural pests are better understood not as urban-warmed temperate insects but as tropical and subtropical species occupying artificial refugia that are thermally disconnected from the region they sit in, with the German cockroach as the limiting case, a species that has never been found in natural habitat anywhere on Earth. We draw out the consequences: the absence of a seasonal reset, the irrelevance of outdoor weather to interior population dynamics, and the reframing of the building envelope as the operative ecological boundary. We also examine a finding that supports the argument from an unexpected direction, namely that temperate urban insects appear to face stronger selection from cold than from heat.
1. Introduction: a framework built in warmer cities
Ask why cities have more pests and the standard answer is the urban heat island. Impervious surfaces absorb and re-emit heat, urban temperatures run above the surrounding countryside, and insects being ectotherms benefit accordingly.
The framework is well supported. Temperature is probably the single most important environmental factor influencing pest behaviour, distribution, development, survival and reproduction, because insects are cold-blooded and their body temperature approximates that of the immediate environment.5 For arthropods, any source of warming can speed development, increase overwinter survival and heighten fecundity, and for pest species this can lead to increased population densities in warmer areas.4
This paper accepts all of that and argues it is the wrong lens for this city.
The central argument The urban heat island helps marginally hardy insects survive winters they would otherwise almost survive. Winnipeg's winter is not marginal. Past a certain cold, urban surfaces stop buffering, and this city spends much of the winter past it. The refuge that matters here is not the city. It is the inside of the building.
2. What the urban heat island literature establishes
The effect sizes in this literature are large, which is worth establishing before arguing about their applicability.
In a study of the scale insect Parthenolecanium quercifex across an urban heat island, overwintering second instars were 13 times more abundant on hot than on cold trees. Ovisacs deposited by the same generation the following April were 5.5 times more abundant on hot trees, and by June the next generation of first instars was over seven times more abundant on hot trees.2
Related work found populations of the same species seven times denser in trees that were only 2.4 degrees Celsius warmer.4
2.1 Adaptation, not just ambient warmth
A common garden experiment sharpened the finding. Individuals from hot urban areas became almost four times more abundant than those from cold urban areas when both were placed in hot greenhouses, indicating local acclimation or adaptation to urban thermal conditions rather than a simple response to ambient temperature. The authors attribute the effect to survival rather than fecundity, having found no differences in fecundity between hot-site and cold-site insects.2
2.2 The refugia argument
The mechanism most relevant to a cold climate is overwintering rescue. Work on evergreen bagworms concluded that impervious surfaces in urban areas can provide refugia for marginally hardy insects and improve their chances of surviving the cold of winter.1
Note the qualifier in that sentence: marginally hardy. It is doing a great deal of work, and §3 is about what happens when an insect is not marginally anything.
2.3 The range expansion argument
There is also a documented mechanism by which warming creates pests rather than merely multiplying them. Increased temperatures due to climate change and urbanisation can allow disease vectors such as mosquitoes and ticks, and invasive species generally, to expand into areas previously unfavourable.3 Urban heat islands can influence pests by increasing their activity and by making non-pest species move into warmer centres and become pests.5
3. The threshold where buffering fails
The single most useful number in this literature for a cold-climate reader is a limit rather than an effect size.
In the bagworm overwintering study, egg mortality was not buffered by impervious surfaces at temperatures at or below minus 21.67 degrees Celsius.1
A ceiling on the mechanism The urban heat island is not an unlimited subsidy. It is a few degrees of buffering, and those few degrees stop mattering once the ambient cold is deep enough. Past roughly minus twenty two, asphalt and concrete no longer rescued the eggs in this study.
3.1 Why a threshold exists at all
Mechanistically this is unsurprising. Impervious surfaces do not generate heat; they store and re-emit absorbed solar energy and slow the rate of loss. The effect is a modest offset from ambient, on the order of a few degrees. An offset of that size converts a lethal minus eighteen into a survivable minus fifteen. It cannot convert a lethal minus thirty three into anything.
The literature is also clear that arthropod responses to urban warming are not uniformly positive. Arthropod abundance does not always increase as cities experience warming,4 and the urban heat island may push some species beyond a critical threshold rather than assisting them.4
4. Winnipeg is past it
Winnipeg's January mean daily low sits in the region of minus 22 degrees Celsius, and cold snaps into the minus thirties are an ordinary feature of the season rather than an exceptional event.
Set against a buffering limit of minus 21.67,1 the implication is direct. Winnipeg does not spend a few nights past the threshold at which impervious surfaces stopped helping. It spends much of the winter there, and the deep cold events go far beyond it.
4.1 What this does to the framework
The urban heat island mechanism does not disappear in a cold city. Winnipeg is warmer than the farmland around it, and that is real. But the ecological function of that warmth changes. In Raleigh or Columbus, a few degrees can move an insect from the dead side of a survival threshold to the living side. In Winnipeg, for a species with no meaningful cold hardiness, the same few degrees moves it from impossible to impossible.
For species that are cold-adapted, and most of Manitoba's outdoor arthropod fauna is, urban warmth is not a rescue either. They were already going to survive.
The heat island therefore does its most consequential work at the margins, on species whose tolerance sits near the local winter minimum. In a temperate city that is a large group. Here it is a narrow one.
5. The cold tolerance finding
A study of urban ants supports this argument from a direction we did not expect, and it is worth reporting carefully because it contradicted its own authors' predictions.
Researchers measured heat and cold tolerance in 14 ant species across urban and peri-urban areas, and measured foraging activity across eight sites over three consecutive years, expecting the urban heat island to drive higher thermal tolerance.6
Contrary to prediction, ants exposed to the urban heat island did not have higher heat tolerance than peri-urban ants. Instead cold tolerance varied across habitats, with ants from cooler peri-urban habitats able to tolerate lower temperatures. Ant activity was almost ten times higher in urban sites and was best predicted by cold tolerance, not heat tolerance.6
The authors conclude that predictions about urban heat islands increasing insect heat tolerance need rethinking, since cold tolerance may be a more plastic or adaptable trait, particularly in the temperate zone.6 The paper's title states the position plainly: insects in temperate urban parks face stronger selection pressure from the cold than the heat.6
5.1 Why this matters for the argument
If cold is the stronger selective force in a temperate city, it is overwhelmingly the stronger force in a subarctic one. The variable that structures Winnipeg's arthropod community is not how warm the summer gets or how much warmer downtown is than Headingley. It is what survives February, and for a substantial set of species the answer outdoors is nothing.
6. The building as the real refuge
Which brings us to the argument this paper actually wants to make.
A heated Winnipeg building in January maintains an interior around 21 degrees Celsius against an exterior around minus 25. That is a sustained gradient of some 46 degrees across a single wall assembly, held continuously for months.
No natural thermal boundary on the Canadian prairie does this. The building envelope is not a gradient in the ecological sense. It is a discontinuity, and on the warm side of it sits an environment with no seasonal analogue anywhere in the surrounding landscape.
6.1 The species this explains
Winnipeg's principal structural pests are not cold-climate insects making the best of a hard winter. They are warm-climate species living in artificial tropics.
The German cockroach requires sustained warmth and cannot persist outdoors here. The pharaoh ant requires roughly 27 to 30 degrees Celsius and gets exactly that year-round from heated mechanical spaces, as discussed elsewhere in this journal. Bed bugs develop on a five to seven week cycle in heated buildings and would face a very different timetable in an unheated one.
For all three, outdoor Manitoba is not a suboptimal habitat. It is not habitat.
7. The limiting case
Blattella germanica deserves separate treatment because it represents the extreme end of this relationship.
As noted elsewhere in this journal, the species is obligately associated with human-built structures, and it has never been found in a natural habitat anywhere on Earth. There is no wild population. There is no source habitat from which buildings are colonised.
This is a genuinely unusual ecological situation and it is worth stating clearly. An organism with a global distribution and no natural range is not an invasive species in the ordinary sense, because invasion implies an origin. The German cockroach's habitat is a human artefact, and its biogeography is the distribution of heated buildings rather than the distribution of any climate zone.
A species that lives in architecture For the German cockroach the relevant map of Manitoba is not a climate map. It is a map of heated structures. Winnipeg and Churchill are, from the species' perspective, the same habitat delivered at different latitudes.
8. Consequences of a missing winter
Several practical consequences follow from a population that never experiences a cold season.
No seasonal reset. Outdoor insect populations in Manitoba face an annual bottleneck that kills most individuals and resets density. Indoor populations do not. A German cockroach infestation in February is not a remnant of a summer population; it is a population that has been reproducing continuously.
Continuous generation turnover. Short generation times running year-round mean more generations per calendar year than an equivalent outdoor species achieves. As argued elsewhere in this journal, generation count is a primary driver of resistance evolution, so the indoor thermal regime is also a resistance accelerator.
Seasonality reflects human behaviour, not insect biology. Apparent seasonal patterns in indoor pest activity track the calendar of human life: heating, holiday travel, lease turnover, food storage. Attributing these to the insect's biology is a category error.
Outdoor weather is nearly irrelevant to interior dynamics. For obligate indoor species, the only thing a cold snap changes is the behaviour of animals trying to get in.
9. The envelope as ecological boundary
If the building is the habitat, the envelope is the habitat boundary, and that reframes what exclusion work actually is.
In conventional terms, sealing a gap is maintenance. In the terms of this paper it is the management of a boundary between two incompatible thermal environments, and it is the single intervention that operates on the actual limiting factor.
9.1 Why exclusion is disproportionately powerful here
In a temperate city, a pest excluded from a building persists outdoors and re-enters later. Exclusion redistributes the problem. In Winnipeg, for an obligate indoor species, exclusion during winter is close to absolute, because the alternative habitat does not exist.
This is an argument for cold-climate operators weighting exclusion more heavily than the general literature suggests, and for treating envelope integrity as the core of a programme rather than as a recommendation appended to a treatment.
9.2 The asymmetry of autumn
It also explains why autumn pressure is so sharp here. The September and October movement of mice, rats, cluster flies and boxelder bugs toward heated structures is not a preference. For animals without the cold hardiness to overwinter outdoors, the building is the only survivable option within reach. The intensity of the push reflects the severity of the alternative.
10. What the urban heat island still explains here
We should not overstate the argument. The framework retains real explanatory power in Winnipeg for outdoor species.
Summer development rates. Urban warmth accelerates development during the growing season, and the mechanism linking warming to faster development and higher fecundity applies as much here as anywhere.4
Range expansion at the margin. The mechanism by which warming allows vectors and invasive species to expand into previously unfavourable areas3 is directly relevant to the northward movement of blacklegged ticks discussed elsewhere in this journal.
Species near their northern limit. Where a species' cold tolerance sits close to Winnipeg's winter minimum, a few degrees of urban buffering could be decisive. The literature's marginally hardy category is small here but not empty, and identifying which local species fall into it would be genuinely useful work.
Microclimate within the city. Sun-facing walls, south elevations and sheltered courtyards create real microclimatic variation, which is why boxelder bug aggregation is so strongly associated with warm south and west walls.
11. Practical reframing
Four consequences for how an operator or a building manager in this climate might think.
Stop expecting winter to help. Clients frequently assume cold weather will suppress an indoor problem. For obligate indoor species it will not, and saying so early prevents a January of wasted waiting.
Treat the envelope as the programme. Where the habitat boundary is a wall, the wall is the highest-leverage object in the building.
Read seasonality as human behaviour. Indoor pest calendars track heating, travel, moving and storage, so anticipate the human events rather than the weather.
Expect faster resistance indoors. Continuous year-round breeding means more generations under selection than an outdoor population accumulates, which supports rotating modes of action on a shorter cycle than general guidance implies.
12. Limitations and open questions
The threshold is one study, one species. The minus 21.67 figure comes from bagworm egg mortality.1 Different species have different cold hardiness and the threshold at which buffering fails will vary. Our use of it as a general limit is an extrapolation from a specific result, and we have treated it as illustrative of a ceiling rather than as a universal constant.
Winnipeg climate figures are contextual. The temperatures in §4 are widely known regional values used to frame the comparison. We have not sourced them to a climate dataset, and a reader needing precision should consult Environment and Climate Change Canada.
No local studies exist. We are not aware of any published urban ecology work on arthropod thermal regimes in Winnipeg, indoor or outdoor. The argument is constructed from literature generated elsewhere plus reasoning about local conditions.
The indoor environment is not uniform. We have treated building interiors as a single thermal regime. Real buildings contain substantial variation, and mapping that variation against species requirements is unexamined here.
The ant study is temperate. Its finding that cold selection exceeds heat selection6 was generated in temperate parks. We argue it applies more strongly at higher latitude, which is reasonable but untested.
13. Conclusion
The urban heat island is a genuine and well-evidenced force in pest ecology. Overwintering survival, development rate and abundance all respond to a few degrees of urban warmth, with effects running to thirteen-fold differences in abundance between hot and cold sites within a single city.2
But the mechanism has a ceiling. Impervious surfaces stopped buffering egg mortality at or below roughly minus twenty two degrees,1 and Winnipeg spends much of its winter on the far side of that line. A subsidy of a few degrees is decisive when survival is marginal and irrelevant when it is not.
What performs the ecological work here is the building envelope, which maintains a discontinuity of forty degrees or more between a stable interior and an exterior that would kill the occupants in hours. The species that matter most in Winnipeg structural work are not cold-climate insects assisted by urban warmth. They are warm-climate species inhabiting artificial refugia, with the German cockroach as the limiting case: a species with a global distribution and no natural habitat anywhere.
The consequences are practical. Populations inside never get a winter, so they never reset, and they accumulate generations and therefore resistance faster than outdoor populations do. Apparent seasonality indoors is a record of human behaviour rather than insect biology. And exclusion, usually filed as maintenance, is in this climate the management of a habitat boundary, which is why it does more here than the general literature would lead an operator to expect.
The urban ecology written in Raleigh and Columbus is about how cities help insects survive a winter they nearly survive anyway. The urban ecology of Winnipeg is about a wall.
References
- Urban microclimate warming improves overwintering survival of evergreen bagworms. Journal of Urban Ecology, 8(1), juac014 (2022). Source for the finding that egg mortality was not buffered by impervious surfaces at temperatures at or below minus 21.67 degrees Celsius, and for the conclusion that impervious surface can provide refugia for marginally hardy insects. https://academic.oup.com/jue/article/8/1/juac014/6678090
- Meineke, E.K., Dunn, R.R., Sexton, J.O. & Frank, S.D. Urban Warming Drives Insect Pest Abundance on Street Trees. PLOS ONE. Source for overwintering second instars being 13 times more abundant on hot than cold trees, ovisacs 5.5 times more abundant, next-generation first instars over seven times more abundant, and the common garden result showing roughly four-fold greater abundance attributable to survival rather than fecundity. https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0059687
- Potential for urban warming to postpone overwintering dormancy of temperate mosquitoes. Source for the review of urban heat island effects across insect groups and for the finding that increased temperatures from climate change and urbanisation can allow disease vectors and invasive species to expand into previously unfavourable areas. https://www.sciencedirect.com/science/article/abs/pii/S0306456523001353
- The functional microclimate of an urban arthropod pest: urban heat island temperatures in webs of the western black widow spider. Source for the general statement that warming speeds development, increases overwinter survival and heightens fecundity in arthropods, the report of populations seven times denser in trees 2.4 degrees warmer, and the qualification that arthropod abundance does not always increase with urban warming. https://www.sciencedirect.com/science/article/abs/pii/S0306456524000329
- Review of Climate Change Impacts on Urban Pests. International Conference on Urban Pests. Source for temperature as the single most important environmental factor influencing pest behaviour, distribution, development, survival and reproduction, and for the observation that urban heat islands can increase pest activity and cause non-pest species to move into warmer centres and become pests. https://www.icup.org.uk/media/w1kgal0a/icup1179.pdf
- Insects in temperate urban parks face stronger selection pressure from the cold than the heat. Source for the 14-species ant study across urban and peri-urban sites over three years, the finding that urban ants did not show higher heat tolerance, that cold tolerance varied with peri-urban ants tolerating lower temperatures, that urban ant activity was almost ten times higher and best predicted by cold rather than heat tolerance, and the conclusion that predictions about urban heat islands raising heat tolerance require rethinking. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11333530/
How to cite this article
APC Exterminators Research Division (2026). Tropical Interiors: Why the Urban Heat Island Is the Wrong Model for Pest Ecology in a Subarctic City. APC Review, Urban Ecology & Pest Biology. Retrieved from https://apcexterminators.com/insights/indoor-thermal-refugia-urban-ecology-subarctic-city