Presence of Dirofilaria immitis in domestic cats and dogs in Chile: Assessment of monthly infection risk as a preventive tool

Abstract

Dirofilaria immitis is the causative agent of heartworm disease, a cosmopolitan vector-borne zoonotic disease that affects domestic and wild mammalian animals, and can be fatal, affecting the vascular and pulmonary systems, as well as other systems such as the renal system. The aim of this study was to analyze the monthly risk of D. immitis infection in Chile as a disease prevention tool using Ecological Niche Models (ENM), validated with a seroepidemiological study. Vector presence data (Culex spp. and Aedes spp.) and bioclimatic variables were collected to model habitat suitability using MaxEnt, integrating them with the calculation of potential generations of the parasite. For validation, samples from 323 pets (178 cats and 145 dogs) from north-central Chile were analyzed. The ENM showed excellent predictive capacity (Area Under the Curve, AUC > 0.89), with the human footprint being the variable with the highest contribution. The model showed marked seasonality, with maximum risk in the austral summer (January-February) dropping to near-zero levels in winter due to thermal restrictions and the Andes Mountains barrier. Serological analysis revealed D. immitis prevalence of 12.9% in cats and 5.5% in dogs, reporting the parasite’s circulation in the country. All positive animals were geolocalized in high or very high-risk areas. Projections to 2100 warn of a massive territorial expansion, estimating suitable habitat gains from 62.8% (Shared Socio-economic Pathways (SSP) scenario SSP1-2.6) to 103.5% (scenario SSP5-8.5) toward the center-south. It is necessary to incorporate the generated dynamic colorimetric maps that alert about the risk of infection into chemoprophylaxis policies and active surveillance in Chile under the One Health approach.

Graphical abstract

Highlights

• •First report of Dirofilaria immitis in domestic dogs and cats in Chile.
• •Ecological niche models predict the monthly risk of Dirofilaria immitis infection in Chile.
• •Infection risk peaks in the southern summer and drops to near zero in winter.
• •Projections to 2100 warn of a massive expansion towards the centre-south of the country.

1. Introduction

Heartworm disease, caused by the parasitic nematode Dirofilaria immitis, represents one of the most important parasitic diseases in global veterinary medicine. It is a vector-borne zoonosis in which domestic and wild canids and felids act as definitive hosts, with domestic dogs serving as the primary reservoirs. Furthermore, culicid mosquitoes act as vectors for the disease, while humans serve as accidental hosts (Noack et al., 2021).

The biological cycle of D. immitis depends intrinsically on the presence and natural state of reservoirs, and extrinsically on environmental climatic conditions and the presence of water, which lead to vector proliferation. In dogs, development takes 6–7 months, whereas in cats, it takes between 7 and 8 months. Adult D. immitis worms are located between the right ventricle and the pulmonary artery of their definitive host. In dogs, dozens of adult worms can be found, living between 5 and 7 years; in cats, typically only 1–3 worms are present, with a shorter lifespan of up to 4 years, and the infection is generally amicrofilaremic (Simón et al., 2012).

Adult ovoviviparous females release microfilariae into the bloodstream. When the vector ingests blood during feeding, it simultaneously suctions microfilariae, which undergo two molts to reach the infective larval stage (L₃) in the Malpighian tubules. This process is strictly dependent on ambient temperature. It has been established that larval development ceases below 14 °C, and the maturation rate increases proportionally with rising temperatures, which is measured using Dirofilaria Development Units (DDUs). The persistence of suitable temperatures and sufficient humidity for the vector’s ecological niche are determining factors for effective transmission (Genchi and Kramer, 2020). This close link to the climate makes heartworm disease a sensitive biomarker of climate change, where warmer and more prolonged seasonal periods favor the expansion of the disease to latitudes previously free of the parasite (Momčilović et al., 2025). When the vector feeds again on a new host, the L₃ larvae are deposited in a drop of hemolymph onto the bite wound and penetrate the definitive host, eventually returning to the bloodstream where they molt to L₄, reach the pulmonary microarteries, and mature into adults in their final definitive location (Hattendorf and Lühken, 2025).

Dirofilaria immitis induces a chronic pathology that leads to severe vascular and pulmonary disease, ultimately resulting in heart failure. This is characterized by proliferative endarteritis, remodeling, edema, pulmonary hypertension, thromboembolism, and congestive heart failure, which can be fatal (Carretón et al., 2020). This leads to progressive clinical manifestations ranging from clinically healthy animals, in many cases, to the gradual onset of chronic cough, exercise intolerance, dyspnea, weight loss, and heart murmurs. In cats, the clinical signs are more respiratory and digestive (vomiting and asthma) and are often fatal (Nelson and Johnson, 2024).

Heartworm disease is a cosmopolitan illness reported in domestic animals across different continents. However, in the context of South America, the distribution of D. immitis, although widespread, is heterogeneous and understudied. In Brazil, Venezuela, Argentina, Bolivia, Peru and Colombia, among others, considerable prevalences are reported, linked to river basins and tropical climate zones (Labarthe and Guerrero, 2005; Bronson et al., 2008; Vezzani et al., 2011; Simón et al., 2012; Labarthe et al., 2014; Portela et al., 2024; Rojas et al., 2023; Cuervo et al., 2024; Chocobar et al., 2026; Félix et al., 2026; Julca et al., 2026; Morchón et al., 2026). Nevertheless, despite sharing borders and similar climate dynamics in certain regions, there is a critical gap in scientific information from Chile. To date, there are no updated and comprehensive epidemiological studies describing the presence of heartworm disease in domestic dog and cat populations within Chilean territory, creating diagnostic and preventive uncertainty for veterinary professionals.

Disease dynamics and transmission risk are influenced by a multitude of factors, including territorial size, climate variability, topography, temperature, rainfall, the distribution of mosquito vectors, and the introduction of invasive species. To this must be added the increased movement of people and pets, as well as anthropogenic activity reflected in urban development, irrigation, and the creation of stagnant water bodies (Maggi and Krämer, 2019; Morchón et al., 2022, 2026). Within a One Health framework, heartworm prevention and control strategies include the diagnosis and clinical management of infected animals, the administration of chemoprophylaxis, and the implementation of mosquito exposure reduction measures (https://www.heartwormsociety.org; https://www.esda.vet/; Montoya-Alonso et al., 2022). Risk maps based on Ecological Niche Models (ENM) are being used to estimate the potential distribution of vectors, infected hosts, and parasite behavior within the vector through the number of Dirofilaria spp. generations, allowing for a more realistic evaluation of infection risk in specific geographical areas, not only annually but even on a monthly basis (González-Mohino et al., 2025; Rodríguez-Escolar et al., 2025; Morchón et al., 2026).

The objective of the present study is to analyze the monthly infection risk of D. immitis in Chile using ENM and to validate it by investigating the presence of D. immitis in domestic cats and dogs from various regions of the country, as no such data exist to date. With all this, and under a One Health perspective, the aim is to provide a reliable visual tool to anticipate the expansion of this zoonosis and strengthen public health and veterinary policies in the Southern Cone.

2. Materials and methods

2.1. Study area

Chile is located on the southwestern margin of South America and consists of a continental strip with more than 8000 km of coastline, four insular regions of volcanic origin located in the South Pacific Ocean, and part of the Antarctic territory, with a total surface area of 2,006,096.3 km² (Fig. 1). This study focuses on the continental region (31°45′40″S and 71°19′7″W), which features an elongated continental geography made up of coastal plains, the Coast Range, the Intermediate Depression, and the Andes Mountains. From north to south, the terrain changes drastically, shifting from arid zones and cliffs to wide, fertile valleys where the majority of the population resides, ending where the Coast Range meets the southern sea. This entire system is dominated by the Andes, the great mountain range that forms the backbone of the country, reaching altitudes of over 6000 m and extending even into Antarctic territory. Regarding Chile’s hydrography, in the north, water hardly reaches the sea because it evaporates in the desert or remains stagnant in salt flats and lagoons, with the Loa River being the only river that manages to cross the desert. In the central zone, rivers receive water from both rain and mountain snowmelt and are essential for agriculture. In the south, rivers have a slower flow regulated by lakes, down to the far south, where constant rainfall creates short, highly voluminous rivers that flow forcefully from the Andes into the ocean (Biblioteca del Congreso Nacional de Chile, 2026).

Fig. 1.

Localization and orography in Chile.

The climate of Chile varies drastically due to latitude and topography, moving through the Köppen classification from north to south. In the north, dry climates (B) predominate, ranging from absolute desert (BW) to semi-arid (BS). Upon reaching the central zone, the climate becomes temperate Mediterranean (Cs), with distinct seasons and winter rains, while in the south and austral zones, temperate rainy and oceanic climates (Cf) dominate, with abundant precipitation influenced by the Polar Front. Finally, in the high peaks and Antarctica, polar and alpine climates (E) appear, where intense cold and snow are a constant norm (Climate Shifts, 2019; Biblioteca del Congreso Nacional de Chile, 2026).

2.2. Vector presence data

Mosquito presence points were obtained belonging to two of the most abundant genera in Chile, Aedes (Ae. albifasciatus and Ae. aegypti) and Culex (Cx. pipiens and Cx. quinquefasciatus), which have been demonstrated to be transmitters of D. immitis in neighboring countries (Labarthe and Guerrero, 2005; Vezzani et al., 2011; Dantas-Torres and Otranto, 2020; Cano-Pérez et al., 2022). These occurrence points were downloaded from databases such as the Global Biodiversity Information Facility (https://www.gbif.org/; accessed on 2 February 2026) and iNaturalist (https://www.inaturalist.org; accessed on 2 February 2026) as well as from published articles (Labarthe et al., 1998; González et al., 2005, 2016; Labarthe and Guerrero, 2005; Vezzani et al., 2011; Cancino-Faure et al., 2024). The geolocalized presence points obtained for the vectors were 116 for the genus Culex and 43 for the genus Aedes.

To avoid biases due to uneven sampling or excessive proximity between data points, all presence records were cleaned and spatially standardized. Using a 1 km² grid, points were filtered to retain only one observation per cell, eliminating duplicates and artificial clusters. This process guaranteed data independence and allowed working with a homogeneous scale in the construction of ecological niche models. At the end of the process, the number of presence points was 99 for Culex spp. and 39 for Aedes spp.

2.3. Bioclimatic and environmental variables

Bioclimatic variables [current years (1970–2000) and future years (2081–2100)] were downloaded from the WorldClim v.2.1 database (https://www.worldclim.org/data/bioclim.html; accessed on 2 February 2026) with a spatial resolution of approximately 1 km² (Fick and Hijmans, 2017). Regarding future variables, we selected two Shared Socioeconomic Pathways (SSP1-2.6 and SSP5-8.5), which represent two scenarios with different greenhouse gas emission concentrations. To avoid cross-correlation and improve model calibration, a multicollinearity test was performed in R software (R Core Team, 2026) using Pearson’s correlation coefficient. Of the 19 current variables downloaded, those with a correlation equal to or greater than 0.8 were excluded (Cobos et al., 2019). The variables selected for Aedes spp. in the study area were: annual mean temperature (BIO₁); mean diurnal range (BIO₂); isothermality (BIO₃); temperature seasonality (BIO₄); annual precipitation (BIO₁₂); precipitation of the driest month (BIO₁₄); and precipitation seasonality (BIO₁₅). For Culex spp. the variables were: BIO₁; BIO₂; BIO₃; BIO₄; maximun temperature of the warmest month (BIO₅); BIO₁₂; BIO₁₅; and precipitation of the driest quarter (BIO₁₇). Thermal variables (BIO₁, BIO₂, BIO₃, BIO₄ and BIO₅) influence the survival and development of the culicids, and precipitation-related variables (BIO₁₂, BIO₁₄, BIO₁₅ and BIO₁₇) are biological determinants as they condition the presence of surface water, which is essential for the presence of immature stages of both genera (Roiz et al., 2011; Samy et al., 2026; Ferraccioli et al., 2023; Morchón et al., 2026).

Additionally, the environmental variables included in the model were: rivers (USGEO, 2026), water bodies (Biblioteca del Congreso Nacional de Chile, 2026), irrigated crops (FAO, 2025), human footprint (Venter et al., 2018), and herbaceous and shrubby vegetation density (Tuanmu and Jetz, 2014). All variables were downloaded in raster format and processed at a resolution of 1 km² (30 arc-seconds), with the same spatial extent (study area) and coordinate system (GCS_WGS_1984), using QGIS 3.44.7.

2.4. Habitat suitability models

To determine the habitat suitability of Aedes spp. and Culex spp. in Chile, ENM were generated using the MaxEnt algorithm (Phillips et al., 2006). The execution and automation of these analyses were carried out using the kuenm package in R (Cobos et al., 2019), which allowed for a systematic evaluation of the potential distribution of both vectors.

To find the best models, 102 candidate models were generated and evaluated for each genus. This was done from a single set of variables with 17 regularization multiplier configurations (ranging from 0.1 to 0.9 and 1 to 10) and six feature class types (l, q, lq, lp, qp, and lqp). The selection of the best model was performed through kuenm under three criteria: statistical significance (Proc < 0.05); an omission rate of 5%; and a minimum complexity based on the corrected Akaike Information Criterion (AICc ≤ 2). Likewise, general predictive capacity was analyzed using the Area Under the Curve (AUC). These values were calculated internally by the program following the standard MaxEnt protocol, using presence records and 10,000 random background (pseudo-absence) points chosen at random. Finally, models that met all requirements - low omission, high significance, simplicity, and robust performance in AUC - were selected as definitive models.

Subsequently, the pixels of the suitability maps for both culicid species were weighted by applying different weighting levels (0–100, 25–75, 50–50, 75–25, and 100–0). In this expression, ENMc and ENMa represent the habitat suitability values for the Culex spp. and Aedes spp. models, respectively, and variables C and A correspond to the relative contribution of each species to the final model:

$$ENMweighted=\frac{ENMc\ast C+ENMa\ast A}{100}$$

2.5. Generations of Dirofilaria immitis

The annual and monthly development of the parasite within the vector was analyzed based on the calculation of temperature-dependent generations (extrinsic development), in order to integrate these results with the vector habitat suitability models and thus obtain infection risk maps. The calculation of D. immitis generations was carried out using a custom script in R, adapting the methodology of Genchi et al. (2005), González-Mohino et al. (2025), and Rodríguez-Escolar et al. (2025) to obtain both annual and monthly estimations. For these calculations, average daily temperatures for the 1990–2016 period obtained from the CHELSA database (CHELSA Climate, 2026) were used. The complete development of the larvae requires an accumulation of 130°-days (GDD) within a maximum period of 30 days, a process that is only possible when the daily mean temperature exceeds the critical threshold of 14 °C.

2.6. Average risk map and its validation

The habitat suitability maps of the two genera weighted in different proportions (Aedes spp. – Culex spp.; 0–100, 25–75, 50–50, 75–25, 100–0) were integrated through an equal weighting (50–50) with the number of D. immitis generations calculated on both an annual and monthly basis.

To choose the best-performing risk map and validate it, first, the Jenks Natural Breaks classification method with ArcGIS v.10.8 was used to divide the data into five risk classes (“very high”, “high”, “medium”, “low”, and “very low”). This classification is a standard technique in cartography and statistical analysis. This algorithm groups values in such a way as to minimize the internal variance of each class and maximize the disparity between them. Secondly, the cats and dogs infected by D. immitis were overlaid through their geolocation, which allowed calculating the percentage of infected individuals within each risk class.

The geolocation of the infected animals by D. immitis was obtained by carrying out a seroepidemiological study of the presence of D. immitis in 323 animals (178 cats and 145 dogs) from Calama (n = 125), Arica (n = 140) and Concepción (n = 58), all of them collected in 2024. These animals had visited veterinary clinics for routine and periodic check-ups for prescribed diagnostic purposes or for official follow-up studies, and their data were subsequently made available to this study. The inclusion criteria were that the animals were over 6 months old, had not travelled outside their province, and had no history of heartworm infection. A complete record was kept for each animal, including identification details (age, sex, breed, and habitat), clinical history and geolocation data.

Blood samples were collected from the cephalic or jugular vein, placed in 3 ml serum tubes and centrifuged. Serum samples were kept at −80 °C until tests were performed. Dog samples were tested for the detection of D. immitis antigens following immunochromatograpic technique (Uranotest® Dirofilaria, Uranovet, Barcelona, Spain) following instructions from the manufacturer. Briefly, 20 μl of serum or plasma was added along with 2 drops of reagent to each of the test strips. The sensitivity of the tests declared by the manufacturer was 94%, and the specificity was 100% (vs necropsy). Cats were tested for serological techniques for anti-D. immitis and anti-Wolbachia antibody detection, as described by Morchón et al. (2004) and other studies (Esteban-Mendoza et al., 2020; Zumaquero et al., 2020; Montoya-Alonso et al., 2022). The cats were considered seropositive when anti-D. immitis and anti- Wolbachia Surface Protein (WSP) antibodies presented jointly.

2.7. Monthly risk maps

To evaluate the seasonality of the cardiopulmonary dirofilariosis infection risk throughout the year, monthly-scale risk maps were developed. To generate these monthly colorimetric risk maps, the vector habitat suitability model that demonstrated the highest predictive capacity during the validation of the average risk map was used. The procedure consisted of weighting this suitability model with the monthly maps of the number of D. immitis generations. Thanks to this weighting, a total of 12 colorimetric infection risk maps were obtained, one for each month of the year.

2.8. Projection to the year 2100 and rank-change analysis

For future projections for the 2081–2100 period in the two climate scenarios, ENM were generated again with MaxEnt using the same parameters obtained from kuenm in the current models, but with the climate variables for 2100 under both climate scenarios. The HadGEM3-GC31-LL model was used to determine the effect of climate change under the Shared Socioeconomic Pathways (SSP1-2.6 and SSP5-8.5) from the Coupled Model Intercomparison Project Phase 6 (CMIP6), representing scenarios with low and high greenhouse gas emission concentrations, respectively (Kuhlbrodt et al., 2018). HadGEM3-GC31-LL is one of the most widely used and evaluated general circulation models (GCM) in ecological niche modeling (Zelinka et al., 2020; Palmer et al., 2023).

To quantify changes in the distribution of both vectors, the developed models and the different future scenarios were converted into presence-absence maps. These binary maps were obtained by applying the 10th percentile threshold of the current model using the raster calculator. Subsequently, the biomod2 library in R was employed to perform a range-change analysis, which allowed quantifying the proportion of pixels that will gain or lose climatic suitability for both vectors in both climate scenarios projected to 2100 (Thuiller et al., 2009).

3. Results

3.1. Ecological niche models

The final model selected by kuenm for Culex spp. was M_3_F_qp (AUC = 0.964), with a habitat suitability range varying from 0 (low) to 0.9 (high). For its part, M_8_F_lqp (AUC = 0.899) was chosen for Aedes spp., with values oscillating from 0 to 0.8. Culex spp. presented higher habitat suitability in the northern and central areas of the country, mainly in the valleys and basins of the central zone. Suitability decreased drastically moving towards southern regions, being practically null in the southernmost areas. Regarding Aedes spp., habitat suitability closely followed the country’s hydrographic network, with the highest suitability found in the central zone. These final models were used to create the 5 models for both genera weighted in different proportions (Supplementary Fig. S1).

Regarding variable importance, for the Culex spp. suitability model, those presenting the highest contribution percentage were the human footprint (52.8%), BIO₁ (18.4%), and BIO₅ (10.3%). The remaining variables had a contribution percentage of less than 4%. Regarding the Aedes spp. model, the human footprint (36%), rivers (29.3%), and BIO₃ (21.7%) presented the highest contributions, with the rest of the variables also contributing less than 4% (Table 1).

Table 1.

Analysis of the contribution of the 15 environmental and bioclimatic variables to the ecological niche model for Aedes spp. and Culex spp.

Variable	Variable contribution (%)
	Culex spp.	Aedes spp.
Human footprint	52.78	36.02
BIO1 (Annual mean temperature)	18.43	2.81
BIO5 (Max. temperature of warmest month)	10.27	0
Water bodies	3.29	0.21
BIO15 (Precipitation seasonality)	2.93	2.22
Rivers	2.52	29.33
BIO2 (Mean diurnal range)	2.28	3.66
BIO4 (Temperature seasonality)	1.99	1.53
BIO3 (Isothermality)	1.73	21.71
BIO17 (Precipitation of driest quarter)	1.72	0
Irrigated crops	0.77	0.01
Shrub density	0.62	0.01
BIO12 (Annual precipitation)	0.46	0.02
Herbaceous density	0.21	0
BIO14 (Precipitation of driest month)	0	2.47

3.2. Annual and monthly generations of Dirofilaria immitis

The average annual map of extrinsic generations of D. immitis in vectors across Chile revealed that the areas with the highest number of generations are located in the northern and north-central regions of the country. In these areas, it reaches a maximum of 4.48 generations, concentrated mainly in valleys and lower-altitude zones. In mountain systems, such as the Andean region, and in the south of the country, a null or very low number of generations was observed due to altitude and low temperatures.

Regarding the monthly dynamics of generations, January and February, during the austral summer, showed the highest number, reaching a maximum value of 5.45 generations in the far north and center, except for high mountain areas (Supplementary Fig. S2 and S3). Conversely, the period between June and August recorded the lowest values, reaching zero across almost the entire country, except for some northern coastal areas where the temperature allows for parasite development. In spring, during September and October, the number of generations begins to gradually increase toward the north and central regions.

3.3. Average infection risk map for Dirofilaria immitis and its validation

For the selection and validation of the average infection risk map for D. immitis in Chile, 323 animals (178 cats and 145 dogs) were analyzed. A seroprevalence of 12.9% was recorded in cats (23/178) and a prevalence of 5.5% in dogs (8/145) for D. immitis. Seropositive cats were reported in all three areas analyzed (Arica, Calama and Concepción), whilst infected dogs were detected only in Calama and Concepción (Table 2).

Table 2.

Prevalence (dogs) and seroprevalence (cats) of Dirofilaria immitis in Chile.

Localization	Dogs			Cats
	N	n	Prevalence (%)	N	n	Seroprevalence (%)
Calama	84	6	7.14	41	6	14.63
Concepción	33	2	6.06	25	8	32.00
Arica	28	0	0	112	9	8.04
Total	145	8	5.52	178	23	12.92

Abbreviations: N, number of samples; n, number of positive samples.

Of the 5 annual infection risk maps for D. immitis generated for Chile by weighting the ENM of both vectors with different contributions (Aedes spp. – Culex spp.; 0–100, 25–75, 50–50, 75–25, and 100–0) and the number of parasite generations, the risk map with the 50Ae50Cx contribution (50% Aedes spp. – 50% Culex spp.) was chosen and validated (Fig. 2). This map best represented the average annual risk of D. immitis infection in Chile, achieving very high performance based on the percentage of infected and geolocalized animals within infection risk zones from high to low risk. In fact, of the 31 positive hosts for D. immitis, 29.1% were geolocalized in very high-risk areas and 70.9% in high-risk areas, with no infected animals located in medium, low, or very low-risk areas (Supplementary Table S1).

Fig. 2.

Average infection risk map for heartworm infection in Chile, constructed from the ecological niche model (ENM) of Aedes spp. and Culex spp. suitability and the average number of extrinsic generations of the parasite within the vectors.

Areas with a very high infection risk for D. immitis in Chile are located mainly in the north and north-central parts of the country, areas that also have suitable habitat for the vectors and a high number of parasite generations. The highest risk values (2.43) were found in the northernmost tip (Arica and Parinacota region) and extend through the valleys and coastal areas of the center. Other high- and medium-risk areas are the south-central regions. On the contrary, higher altitude regions along the Andes Mountains, which act as a natural barrier, presented low risk values. Finally, in the southernmost latitudes of the country, including the fjords and Patagonia, the infection risk was low to very low.

3.4. Monthly infection risk maps for Dirofilaria immitis

Monthly risk maps in Chile showed a marked seasonal pattern that varies by latitude and southern hemisphere seasons (Fig. 3). During the austral summer, the risk reached its maximum expansion. In January and February, the highest levels were recorded with maximum risk values of 2.95 and 2.69, respectively, with the northern and central regions presenting a very high risk, while the Andean peaks and the southernmost regions maintained moderate or low levels.

Fig. 3.

Monthly infection risk maps for heartworm infection in Chile generated by combining the ecological niche model (ENM) of Aedes spp. and Culex spp. suitability with the estimated monthly number of extrinsic parasite generations.

Upon the arrival of autumn, starting in March, a progressive decline in risk began from the highest latitudes toward the north. It is during April and May when the greatest decrease in risk was observed, remaining restricted mainly to northern areas and the northern coast. During the austral winter (June, July, and August), the risk dropped to minimum levels across almost the entire territory, with values close to zero. In September and October, with the arrival of spring, the risk began to increase again from the north toward the center, and it is in November and December when very high-risk levels were re-established in the basins of the north and north-central parts of the country.

3.5. Projection of habitat suitability for vectors in 2100

The future projection of habitat suitability for the year 2100 for both genera weighted under the SSP1-2.6 climate scenario showed an expansion in Chilean territory, with a territory gain of 62.8% compared to a loss of 2.3% (Fig. 4). Under the high-emissions scenario SSP5-8.5, suitability showed a much more pronounced growth, recording a gain of 103.5% against a loss of 4.9%.

Fig. 4.

Analysis of range change of 50% Aedes spp. and 50% Culex spp. (50Ae50Cx) under SPP1-2.6 and SPP5-8.5 scenarios in 2100.

Range change maps for both genera reveal that the gain in climatic suitability is predominantly concentrated in the south-central regions and the foothills (precordillera), reaching latitudes and altitudes that currently act as thermal barriers for vectors (Supplementary Fig. S4). The same pattern of expansion was observed in both scenarios, although it manifested much more drastically in the high-emissions projection.

4. Discussion

The application of ENM using the MaxEnt algorithm has proven to be a robust methodology to delineate the habitat suitability of vectors involved in D. immitis transmission, allowing for highly accurate evaluations of both annual and monthly infection risk in a given region (Rodríguez-Escolar et al., 2023, 2025). Through this methodological approach, the present study consistently demonstrates and validates the risk of D. immitis infection in Chile as an emerging vector-borne disease. Furthermore, the presence of infected animals within the national territory of the Southern Cone is reported for the first time.

The differentiated modeling of suitable habitat for the genera Culex spp. and Aedes spp., the most abundant taxa with the highest presence records in the study area, revealed complex and specific ecological dynamics. A fundamental finding is that the human footprint constituted the variable with the highest contribution percentage for Culex spp. habitat suitability and positioned itself as one of the main factors for Aedes spp. This marked synanthropic behavior and high ecological plasticity correspond directly to the geospatial patterns described in other studies (Wilke et al., 2021; Rodríguez-Escolar et al., 2025; Morchón et al., 2026), in which mosquito ecology is linked to uncontrolled urbanization and the availability of artificial breeding sites rich in organic matter in various temperate and tropical regions of the world, as is the case with Cx. quinquefasciatus and species of the genus Aedes.

Anthropogenic activities, uncontrolled urbanization, and anthropogenic landscape modification profoundly alter local ecosystems (Kache et al., 2022; Romanello et al., 2024). This phenomenon gives rise to stable microclimates and artificial water collections that act as optimal breeding sites for vector culicids, thereby facilitating the expansion and persistence of dirofilariosis in urban nodes densely populated by definitive hosts (Noack et al., 2021). However, the models discriminated subtle ecological differences: while thermal variables such as annual mean temperature (BIO₁) and maximum temperature of the warmest month (BIO₅) critically determine the niche of Culex spp., proximity to hydrographic networks is vital for the establishment of Aedes spp. in the arid and semi-arid regions of northern Chile.

The incorporation of the biological development model of the parasite conditioned by temperature (measured through cumulative potential generations based on the biological threshold of 14 °C and 130°-days of development or GDD) allows for a fundamental qualitative leap from purely vector suitability maps to actual infection risk maps (Genchi et al., 2005; González-Mohino et al., 2025). Through the colorimetric maps generated for Chile, marked spatial heterogeneity and strict temporal seasonality were observed, combined with absolute geographical restrictions imposed by the country’s climatological topography. On one hand, areas with a very high infection risk are concentrated in the north and north-central parts of the territory, reaching maximum values during January. In this sense, the altitude of the Andes Mountains plays a crucial role as a natural thermal barrier, since the extreme cold of the high peaks completely limits the extrinsic development of D. immitis, causing a clear spatial decoupling between the physical presence of the mosquito and the biological viability of the parasite. On the other hand, the risk is concentrated almost exclusively during the austral summer (January and February). In marked contrast, during winter (June to August), low ambient temperatures cause the transmission risk to drop to practically zero levels across the entire country, persisting residually only in very specific coastal sectors of the far north. This restricted temporal window delineates a markedly intermittent and seasonal transmission scenario, which differs from the permanent continuity observed in the lowlands of equatorial and tropical regions of the subcontinent (Dantas-Torres et al., 2023; Atkinson et al., 2025; Morchón et al., 2026).

At the seroepidemiological level, the finding of an alarming 12.9% seroprevalence in cats and a 5.5% prevalence in dogs in the studied urban centers (Calama, Arica, and Concepción) provides unprecedented data for the Southern Cone. Although canine prevalence sits in intermediate ranges compared to hyperendemic foci on the Atlantic coast or tropical Latin American basins (Labarthe and Guerrero, 2005; Maggi and Krämer, 2019; Esteban-Mendoza et al., 2020; Badillo-Viloria et al., 2023; Monteiro et al., 2026), the high positivity rate in domestic felines represents a serious epidemiological warning. Cats are traditionally considered more resistant hosts than canids against D. immitis infection. However, due to their lifestyle habits and significantly lower use of prophylactic therapies, they act as excellent biological sentinels of vector activity in urbanized environments (Noack et al., 2021). For example, in the locality of Arica, no infected dogs were detected, but seropositive felines were reported. This clearly demonstrates that omitting sampling in the feline species induces a dangerous underestimation of the real local transmission risk. One potential limitation of the study lies in the sampling strategy, as the animals analyzed were limited to those from which a sample could be obtained for testing. It is likely that with a larger number of samples, especially if they were more widely distributed, the seroprevalence and prevalence data would differ.

Future projections for Chile warn of an expansion of the climatically suitable territory for dirofilariosis transmission. Under the sustainable development and low-emissions scenario (SSP1-2.6), a net territory gain of suitable habitat of 62.8% is estimated, accompanied by a loss of barely 2.3%. This phenomenon is alarmingly exacerbated under the high-emissions scenario (SSP5-8.5), where the territorial gain doubles to reach 103.5%, compared to a suitability loss of only 4.9%. Range change maps clearly show that the colonization of new niches will be predominantly concentrated in the south-central regions of the country and in the foothills (precordillera). These geographical zones currently operate as stable thermal barriers due to their low winter and nighttime temperatures. However, projected global warming will cause the breakdown of these thermal limits, shifting the risk toward higher latitudes and altitudes and exposing pet populations to parasite expansion (Hattendorf and Lühken, 2025; Momčilović et al., 2025).

Under the One Health approach, these findings demonstrate that dirofilariosis can no longer be considered an exotic disease in Chile. It is urgent that public health and veterinary policies in the Southern Cone incorporate these monthly risk maps as dynamic predictive tools, establishing well-defined seasonal chemoprophylaxis guidelines (coinciding with the months of thermal GDD accumulation) or continuous guidelines in the far north, and implementing active surveillance programmes to protect animal health and prevent potential zoonotic cases and pulmonary nodules in the human population.

5. Conclusions

This study demonstrates, through integrated, predictive, and validated modeling (ENM and GDD), that there is a seasonal and intermittent risk of D. immitis infection in Chile, concentrated in the summer period in the northern and north-central regions, remaining strictly restricted only by the altitude of the Andes Mountains and the cold southern areas. Furthermore, the presence of animals infected by D. immitis in Chile is reported for the first time. The generated colorimetric maps can serve as an epidemiological tool for health surveillance planning, selective vector control, and anticipated mitigation of climate change effects on animal health and public health. This highlights the need to design and implement structured and smart chemical chemoprophylaxis regimens (seasonal or continuous depending on the geographical region) and to expand monthly risk maps to other zoonotic vector-borne diseases under the unified prism of One Health.

Ethical approval

For this study, no ethical approval was required. All blood samples were routinely collected for prescribed diagnostic purposes or official monitoring studies and subsequently made available to this study. The study was conducted in accordance with current Chilean legislation on animal welfare.

Funding

The presented study was supported by the General Foundation of the University of Salamanca (art. 60 - LOSU), Spain. Iván Rodríguez-Escolar was supported by the University of Salamanca-Banco Santander as predoctoral scholarship. Elena Infante González-Mohino and Manuel Collado-Cuadrado were supported by the CLAVE programme and the General Foundation of the University of Salamanca. Alfonso Balmori-de la Puente was supported by Margarita Salas grants for the training of young doctors (University of Barcelona), funded by the Ministerio de Universidades and Plan de Recuperación, Transformación y Resiliencia (Spain) and by Next Generation (European Union).

Declaration of competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix B. Supplementary data

The following are the Supplementary data to this article:

Data availability

The data supporting the conclusions of this article are included within the article and its supplementary files.