Advertisement

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Endemic and zoonotic cycles of cutaneous leishmaniasis depend on vector feeding preferences: An epidemiological model for Southeastern Mexico

Abstract

Introduction

Cutaneous leishmaniasis (CL) is a neglected tropical disease caused by Leishmania spp. parasites. It poses significant public health challenges among economically marginalized communities in endemic regions like Southeastern Mexico. Here, we developed a mathematical model to describe the transmission dynamics of CL in the Yucatan Peninsula region, focusing on two key sand fly vectors: Lutzomyia cruciata and Bichromomyia olmeca.

Methods

Transmission was modelled as frequency dependent with Susceptible-Infected dynamics for vectors and rodents, and Susceptible-Exposed-Infected-Susceptible for humans, and accounts for the impact of available blood meals on vector populations. The model was parameterized from published literature for vector and reservoir growth rates and carrying capacities, transmission efficiency, and vector feeding preferences.

Result

The simulations highlight the importance of both vector species in CL transmission, with Lu. cruciata showing a higher preference for human blood, while Bi. olmeca is more frequently associated with rodents. Sensitivity and scenario analyses reveal that the system is highly sensitive to transmission rates and vector feeding preferences, suggesting that Bi. olmeca is necessary to maintain infection among reservoirs, while Lu. cruciata contributes to zoonotic transmission.

Conclusions

Overall, we emphasize that vector species have different roles on leishmaniasis epidemiology due to feeding preferences. Therefore, roviding funding to improve basic knowledge into the epidemiology of this disease, will improve our understanding of its dynamics. Once achieved, policymakers can develop targeted interventions to reduce CL incidence in affected regions to improve the outcomes of public health interventions.

Author summary

Cutaneous leishmaniasis (CL) is a skin disease caused by parasites and spread by the bites of infected sand flies. It affects vulnerable populations in rural and marginalized communities, such as those in Southeastern Mexico. Although it can cause long-lasting skin sores and social stigma, many aspects of how this disease spreads in the region remain poorly understood. We developed a mathematical model to better understand the transmission of CL in the Yucatan Peninsula, focusing on two sand fly species that feed, at different degrees, on both humans and wild animals. Our results show that the two sand fly species play different roles in disease transmission: one species maintains wild animal infection, while the other transmits the disease to humans. These results derive from data scattered across the study region, which carries significant uncertainty, and represent important information gaps. These limitations highlight the need for further research on how local sand fly populations interact with human and animal hosts, thus improving modelling efforts for disease control.

Citation: Martín G, González EJ, Loría-Cervera EN, Montes de Oca-Aguilar AC (2025) Endemic and zoonotic cycles of cutaneous leishmaniasis depend on vector feeding preferences: An epidemiological model for Southeastern Mexico. PLoS Negl Trop Dis 19(12): e0013786. https://doi.org/10.1371/journal.pntd.0013786

Editor: Indrajit Ghosh, Ahmedabad University, INDIA

Received: June 17, 2025; Accepted: November 20, 2025; Published: December 31, 2025

Copyright: © 2025 Martín et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All data and model code are available in the GitHub repository https://github.com/gerardommc/Leish-transmission.git, and its Zenodo archive https://doi.org/10.5281/zenodo.17717116.

Funding: Funding was provided by the Secretaría de Ciencia, Humanidades, Tecnología e Innovación, project number CBF-2025-I-2041 (granted to GM, EJG, ENLC and ACMOA). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Cutaneous leishmaniasis (CL) is a neglected tropical disease caused by Leishmania spp. protozoan parasites [1]. Clinically, CL is characterized by chronic skin lesions which may develop into face disfigurement [2]. In endemic areas, CL is a serious public health problem contributing with 0.58 disability adjusted life years (DALYs) per 100,000 people [3], especially among economically marginalized communities [4]. Although CL typically has very low mortality, its DALYs have a significant contribution to the poverty trap [4]. The annual CL incidence in the Americas is estimated to fluctuate between 187,200 and 307,800 cases [5], with Brazil, Colombia and Venezuela having the highest burden [6]. Controlling leishmaniasis transmission has been limited because it is a zoonotic disease involving multiple wild non-human mammalian reservoirs and insect vector species of the Phlebotominae subfamily (Diptera: Psychodidae) [7]. Thus, multiple complications arise from our poor understanding of its transmission cycles including the interactions between various vectors and reservoir species.

In the Neotropics, humans typically acquire CL near tropical forests, notably while working on agriculture, timber and non-timber forest products or during activities related to land conversion. Thus, working, living or settling in forest stands is the main risk factor for acquiring CL in the Americas [8]. In this way, humans become accidental, frequently dead-end hosts [9], and therefore they are considered irrelevant for the long-term persistence of CL in endemic areas [10].

The Americas are experiencing regional changes in the epidemiology of CL and mucocutaneus leishmaniasis (ML). In the past years (2016–2022), there has been a decline in the number of cases and incidence of CL and ML in 12 of the 17 countries considered to have high endemic transmission [6,11–17]. In 2016 alone, 48,915 cases were reported, of which 94–97% were CL, with an incidence of 21.71/100,000 habitants, while 41,617 cases and an incidence of 18.78/100,000 habitants were recorded in 2019 [6]. However, Mexico, together with four other countries that do not belong to the high transmission ranking, exhibit a non-linear increase from 76 to 146% in CL-ML cases [6,11–13,17–19]. In Mexico, CL accounts for 99% of the cases, of which ~60% are located in the biogeographic province of the Yucatan Peninsula [20].

The transmission of Leishmania mexicana mexicana in the Yucatán Peninsula is hyperendemic [21]. Currently several outbreaks of CL have been identified in areas with no historical records [20,22,23]. In this region, rural Mayan communities are the most affected since they have established a strong historical relationship with the tropical forest [24,25]. In these rural populations, CL is expressed as a single, rounded, and painless ulcer mainly located in the ear where it tends to become chronic if untreated [25]. At the time of writing, the annual incidence has been estimated at ~508/100,000 inhabitants [26]. However, given that 18.9–27.6% of the cases are asymptomatic, actual incidence in active foci could even be higher [22,25]. Regardless of the actual figures, CL control measures implemented in affected communities by the Mexican ministry of health, continue to be based on endorsing prevention via suggesting appropriate clothing for risk-prone activities and occupations [27].

The enzootic cycle of Le. mexicana is seasonal, beginning after the rainy period and persisting for approximately five months [28]. There, three wild rodent species act as primary reservoirs, Ototylomys phyllotys [29], Heteromys gaumeri [30] and Peromyscus yucatanicus [30–32]. Research on CL has so far identified at least eight sand fly species which could play an important role as vectors [23,33–36]. Among these sand flies, Bichromomyia. olmeca olmeca [37] is the only vector proven to be both infected by Leishmania spp. parasites and transmit them. However, this sand fly is considered more zoophilic than anthropophilic, as it is collected more frequently in traps baited with rodents than those baited with humans [33,34,38–40]. In addition, Bi. olmeca populations are smaller in disturbed landscapes [39,40]. In contrast, the sand fly Lutzomyia. cruciata (Coquillett), another suspected important vector, with confirmed high Leishmania infection prevalence [23,33,34], has higher preference for feeding on humans, [23,38], and appears to be more tolerant to human-dominated landscapes in the region, compared with Bi. olmeca [41–43].

Given the historical and contemporary epidemiological pattern of CL in the Yucatan Peninsula, exploring and understanding the mechanisms that regulate infection prevalence and incidence is still necessary. In this sense, mathematical models have been proposed previously to describe CL dynamics in the Americas, focusing on: 1) developing analytical expressions for the basic reproductive number [44]; 2) estimate critical epidemiological parameters from data at different rural localities in Venezuela [45]; and 3) analyze the dynamic properties of CL to test whether certain mitigation strategies could fail in Peru [46]. Mathematical models can provide insights to identify which prevention and control strategies are more effective and identify knowledge gaps to focus future research efforts [47]. Therefore, in this study, we present a simple mathematical model to describe the transmission dynamics of CL in the Yucatan Peninsula with the participation of two well-documented vectors (the sand fly species Bi. olmeca and Lu. cruciata) with different feeding behaviors and responses to the type of available bloodmeals from reservoir hosts (O. phyllotis, P. yucatanicus and H. gaumeri), and with humans as incidental hosts. Our study represents the first attempt to model the transmission dynamics of one of the most neglected but prevalent diseases in areas of high poverty and marginalization in the Yucatan Peninsula.

Methods

Model setting

We propose a CL transmission model comprising a rodent reservoir species, two sand fly vectors, Lu. cruciata and Bi. olmeca, and humans as incidental hosts with a set of nine ordinary differential equations (two equations for each vector species, two for reservoirs and three for humans). Among the rodent reservoir and the sand fly vectors, CL has Susceptible-Infected (SI) without recovery epidemiological dynamics. In turn, human CL follows Susceptible-Exposed-Infected-Susceptible (SEIS) dynamics. Transmission from rodents to sand flies, and from sand flies to rodents and humans is assumed to be frequency dependent. Furthermore, we assume that the two sand fly vectors are the most relevant species for the transmission of CL.

In humans, once CL infection occurs, there is an incubation period of 20 days before the onset of clinical symptoms. Afterwards, humans recover in approximately 21 days without protective immunity and re-enter the susceptible population. During the infectious period, we assume that humans are not capable of transmitting the infection neither to vectors nor to other humans, which is based on the most recent evidence of leishmaniasis epidemiology in the Yucatán Peninsula [25,48].

To represent the effect of available bloodmeals on sand fly species, we hypothesise that both human and rodent population sizes determine their carrying capacities in a logistic growth fashion. The second consequence of these different feeding ecologies is the variability of transmission among the two sand fly species.

Model equations.

The population of each vector species i (Vi, with i = 1 for Lu. cruciata, and i = 2 for Bi. olmeca) is divided into susceptible (S) and infected (I) compartments (Fig 1). The S → I transition depends on the probability that among its daily meals a vector bites an infected rodent reservoir. Hence the denominator of the frequency dependent transmission term (eqn. 1a) accounts for the proportion of meals taken from both humans and rodents, where subscripts N indicate the entire population, including the different epidemiological compartments.

thumbnail
Download:
Fig 1. Flow diagram of the model.

Square boxes correspond to state variables, while circles indicate model parameters. Solid arrows indicate in or outflows from each compartment, while dashed lines or arrows indicate the flows that are regulated by parameters interacting with the state variables connected to them. R = reservoir host, Vi = vector species, with i = 1, Lu. cruciata, and i = 2, B. olmeca, and H = humans. Subindices S = susceptible, E = exposed. and I = infected, identify the epidemiological status of each compartment.

https://doi.org/10.1371/journal.pntd.0013786.g001

Regarding their population dynamics, the model assumes that both S and I vectors are equally fertile, and that the carrying capacity of vectors (KVi, eqn. 1c) increases linearly with rodent (RN) and human (HN) total population densities, with vector species-specific constants, ai and bi, representing the number of sand flies that each rodent or human individual usually maintains alive, respectively [43]. Therefore, the equations describing these dynamics are:

(1a)(1b)(1c)

Tables 1–3 present the state variables and parameters associated with these equations, respectively.

thumbnail
Download:
Table 1. State variables associated to the cutaneous leishmaniasis transmission model, along with their initial values used in the simulations. Epidemiological compartments: S = susceptible, E = exposed, and I = infected.

https://doi.org/10.1371/journal.pntd.0013786.t001

thumbnail
Download:
Table 2. Parameters associated to the cutaneous leishmaniasis transmission model, along with the fixed values used in the simulations and the source of these values. i = 1 corresponds to vector species Lu. cruciata and i = 2 to vector species Bi. olmeca.

https://doi.org/10.1371/journal.pntd.0013786.t002

thumbnail
Download:
Table 3. Parameters of the sand fly population growth model derived from Castillo et al. [49].

https://doi.org/10.1371/journal.pntd.0013786.t003

Similarly to the vector population dynamics, the rodent dynamics are assumed to be unaffected by infection and all individuals are born susceptible. Epidemiologically, rodents are also divided into susceptible and infected, and the S → I transition after a bite from an infected vector is also frequency-dependent. The transmission rate from each vector species is the product of a general transmission rate (βR) and the proportion of bites taken by vector species i from rodents (1 – pi). These dynamics are represented by the equations:

(2a)(2b)

The population dynamics parameters rR and KR were estimated from field data (details below), and the latter was assumed to remain constant.

For humans, population growth is not restricted by carrying capacity, but by the balance between rH and μH. The S → I transition is frequency dependent, with a vector species-specific contact rate and proportion of meals taken from humans as the reciprocal of meals taken from rodents (pi). The epidemiological parameters controlling the transitions E → I and I → S are σ and ω respectively. The dynamics for humans are represented by the following equations:

(3a)(3b)(3c)

Demographic parameters

The parameters in the model associated with the dynamics of the state variables were the vector and reservoir population growth rates, rV,i and rR, respectively, and the reservoir carrying capacity KR. We approximated their values as described below. The model includes further parameters for which we obtained their values from the literature (Table 1).

Vector population dynamics.

To approximate a value for rV,1, we used laboratory-based life tables published by [49] for Lu. cruciata; we assumed that the corresponding parameter value for Bi. olmeca in the model is the same. The demographic dynamics was described through a continuous-time matrix population projection model, where vector populations were structured by development stage: 1) egg, 2) larvae, 3) pupae, and 4) reproductive adult. The value of rV was the real part of the dominant eigenvalue of the 4 × 4 matrix L:

where n is the average number of eggs deposited, pV is the fraction that hatches, mj are the maturation rates (1/time spent in development stage j, in days), and μj is the state-specific mortality rate (Table 2).

Reservoir population dynamics.

To approximate values for rR and KR, we used published demographic data on the three most common rodent species of the Yucatan Peninsula, P. yucatanicus, H. gaumeri and O. yucatanensis [52–54] (F). In all cases, data consists of a time series of population density estimates in quadrats of 1 ha, in separate locations across the state of Yucatan (Hernandez-Betancourt et al. 2004, 2006, 2008). These time series were divided into two phases: an r-dominated and a K-dominated. K-dominated phases were those where population densities were larger than the 50% of the maximum population density recorded, and r-dominated phases were those where population densities were smaller than 50%. Then, we calculated population change between time points and used the exponential model to solve for r:

In this way, we obtained a series of values for r, which we then averaged to obtain a point estimate to use in the transmission model; the values of KR were also averaged. Both values of rR and KR represent the average for all rodent species taken together.

Epidemiological parameters

The parameters controlling transition S → I are the transmission rates βV,i from reservoir to vector species i, βR from vectors to reservoir and βH,i for vectors to humans (Fig 1). To estimate a range of values for βH,i we used the approach developed by [57], where:

with n the daily number of vector-human contacts, and ε the transmission efficiency, both taken from the literature (Table 3). For βR, to use as much data from field studies we used the same values derived for humans (Ravinovich and Feliciangeli, 2004) but weighted by the proportion of bites on rodents. Also, in light of the absence of data for βV,i, we used the transmission rates for visceral leishmaniasis of [50].

For the fraction of Lu. cruciata bites on rodents, pi, we assumed that it was proportional to the fraction of skin surface area available to draw a blood meal among humans and rodents. For Bi. olmeca, pi is 20 times the value for Lu. cruciata (see Table 3 for epidemiological parameters).

Simulation setup and scenarios

Simulation scenarios were designed to show the range of possible outcomes during leishmaniasis one transmission season on five months duration (~150 days), in response to those parameters for which fewer information is available, namely the transmission rates β parameters. To begin, we assumed that all β’s are a function of the transmission efficiency ε estimated for other leishmaniasis endemic areas in the Americas [45].

The initial population values used in these simulations were the approximate human population found in Xpujil of ~6000 people [58] (Tatem et al. 2017), Campeche, the township where CL epidemiology is best understood in the study region, assuming 1, 5 and 10% of the population exposed to Le. mexicana vectors via various occupational risk factors. The initial values of vector populations, Vi(0), were calculated at equilibrium level with the formula for KVi (eqn. 1c) and were zero in scenarios where either or both sand fly species were absent.

Finally, to identify how the different parameters affect the system’s behaviour we performed a a general sensitivity analysis using the Sobol-Martinez technique. To generate parameter uncertainty, we used a uniform distribution with minimum and maximum values corresponding to default values + /- 20%, to match the local sensitivity analysis. These analyses were performed using the ODEsensitivity R package [59].

Results

The final model comprised nine differential equations with 11 parameters for the disease dynamics among vector species, four for the rodent reservoir dynamics and six for the transmission to humans (Fig 1). The transmission dynamics represented by the model were highly influenced by the vector species present. The scenario in which only Lu. cruciata was present, prevalence in one transmission season only reached a maximum of ~0.3% among humans (60 people), or 3 cases per 1000 population, ~ 100 days after one infected rodent was introduced in a 100% susceptible population, and after this point in time, prevalence decreased (Fig 2). However, in the scenario with only Bi. olmeca present, prevalence continued increasing and reached its peak at the end of the 150 day simulation with a ~ 1% among humans. In the scenario with both vector species present, prevalence at the end of the simulation was ~ 2%. In the long term, prevalence stabilised around prevalence of 34% and 45% of the human population respectively (S1 Fig). These results indicate that with the parameter values used, Bi. olmeca is necessary for the maintenance of the parasite among reservoirs, while Lu. cruciata increases transmission of CL from rodents to humans (Fig 2). Likewise, the difference between the scenarios with and without Lu. cruciata on CL prevalence among rodents in the span of one transmission season was negligible.

thumbnail
Download:
Fig 2. Trajectory of infection prevalence in each species of the CL system in three scenarios based on the vector species present.

Initial conditions and variable parameters were H(0) = 60, ε = 0.13, R = 1500. Vi(0) populations were calculated at equilibrium level with the formula for Ki, and were zero in scenarios where either are absent.

https://doi.org/10.1371/journal.pntd.0013786.g002

CL prevalence among humans also depends on the size of the exposed human population, where understandably, higher prevalence is reached faster in smaller populations (Fig 3) due to: 1) immediate prevalence is higher once infection is introduced; and 2) human populations increase the carrying capacity of Lu. cruciata, increasing its population size more than that of Bi. olmeca. In the absence of human population growth, prevalence stabilises at around zero for both rodents and humans when only Lu. cruciata is present; it stabilises at around ~34, 46 and 52% among humans, rodents and Bi. olmeca, respectively, when only this sand fly is present; and around 44, 47, 53 and 18% among humans, rodents, B. olmeca and Lu. cruciata, respectively, when both species are present (S2 Fig).

thumbnail
Download:
Fig 3. Effect of increasing transmission efficiency (ε) on infection prevalence among humans, in each of the three possible scenarios of human population density.

Initial conditions and variable parameters were H(0) = 60, ε = [0.01 - 0.13], RS(0) = 1500, RI(0) = 1. Vi(0) populations were calculated at equilibrium level with the formula for Ki, = aiR(0) + biH(0), and were zero in scenarios where either are absent.

https://doi.org/10.1371/journal.pntd.0013786.g003

Global sensitivity analysis revealed that leishmaniasis transmission to humans was most sensitive to the transmission rates involving rodents and Bichromomyia olmeca. Using the global sensitivity analysis, leishmaniasis transmission to humans was most sensitive to the transmission rates to and from rodents and Bi. olmeca (βR and βV,2, respectively), the fraction of bites of Bi. olmeca on rodents (p2), the transmission efficiency (ε), and the vector mortality rate (μ). The contribution of humans to the carrying capacity of both sand fly species (a1, a2, b1, and b2) had a very weak effect on infection prevalence on humans (Fig 4; full Sobol sensitivity analyses are in S3 Fig).

thumbnail
Download:
Fig 4. Sobol sensitivity analysis of the system using a uniform distribution with minimum and maximum correspoding to the default values ±20%.

The height of the columns in the y-axis shows the Sobol total sensitivity index, which already takes into acccount the variance of all parameters, for each state variable (colour). The initial conditions were H(0) = 60, RS(0) = 1500, RI(0) = 1, ε = 0.13.

https://doi.org/10.1371/journal.pntd.0013786.g004

Discussion

Our simulations suggest that the presence of Bi. olmeca is necessary for Le. mexicana infection to persist among the rodent reservoir populations, when using the parameter values in Table 1. Bi. olmeca has long been considered one of the most important Le. mexicana vectors in the region (Biagi 1965), albeit with controversy. To address this issue, we included a second vector species, Lu. cruciata, found to be consistently present in leishmaniasis-endemic areas [33,34,39,41,60]. In contrast with Bi. olmeca, Lu. cruciata increases transmission to humans, as it does not have a preference for feeding on either humans or rodents. Hence, while Bi. olmeca maintains Leishmania infection, Lu. cruciata, via generalist feeding habits increases the range of infected hosts. This is evidenced by the difference observed between scenarios with, without, or including both vector species. Nevertheless, the absence of Bi. olmeca could be offset by an increase in the life expectancy of Lu. cruciata, as there would be more time for infected sand flies to contact susceptible rodents or humans. The end result is that the leishmaniasis transmission system in the Yucatan Peninsula is highly sensitive to the feeding ecology of the different sand fly species and the underlying carrying capacity determined by the availability and abundance of potential hosts as bloodmeals.

To the best of our knowledge, this is the first study to integrate the effect of sympatric Leishmania vector species on modelling its transmission dynamics. While previous studies have documented active infections on seven Phlebotominae (sand fly) species in the Yucatan Peninsula and Mexico [33,34,36,41,61], their role on infection maintenance and transmission to humans had not been evaluated. On the contrary, most studies on transmission dynamics in the Americas have focused on the effect of reservoir species [44,60], but none of them in our study area. Compared with the cited studies, we acknowledge that our study area identifying system properties under the available evidence is highly determinant of the levels of transmission to humans. As this approach stems from the current level of understanding and data quality, a comprehensive research agenda we identify is resolving the role of vector and reservoir diversity on the leishmaniasis transmission system throughout its distribution in the Americas.

Even though the modelling exercise presented here has rendered novel insights, these should be taken with caution, the main reason being that knowledge on sand fly feeding preferences and demographic parameters remains limited, even with ~60 years of medical entomology research on leishmaniasis vectors in the study area [62]. Strinkingly, studies in the Yucatan Peninsula largely continue to focus on estimating infection prevalence, while profiles on feeding ecology are still poorly described for either Bi. olmeca or Lu. cruciata. Also, Lu. cruciata is the only sand fly species for which there is life history information in Mexico [49]. This undoubtedly limits the parameterization of our model and the interpretation of its results. A natural solution to this problem would be to generate basic feeding and population ecology information for Bi. olmeca, both of which will further help to understand the impact of this neglected tropical disease.

In a similar way that the vector feeding ecology affects leishmaniasis maintenance and transmission to humans, the diversity of reservoir species influences both epidemiological phenomena. In Venezuela, for instance, dogs and donkeys are its main reservoirs, but the latter increase leishmaniasis’ reproductive number more than dogs [60]. Such studies evidence that introduced domestic species can become the most important reservoirs of human leishmaniasis. This is possible by a transmission cycle separate from the sylvatic (with native wild mammal reservoirs; [63]). In the Yucatan Peninsula, the full range of reservoir species have not been identified yet, although there are clinical leishmaniasis cases involving domestic dogs and cats [64]. This is important because isolated evidence has emerged of human infection within rural communities where Bi. olmeca populations are small and activities traditionally considered risk factors are not practiced (e.g., activities in the forest, [23,65]). Above, we highlighted the importance of life expectancy of both reservoir and vector species on maintaining infection. In this sense, and under our model system, human leishmaniasis in areas were Bi. olmeca populations are scarce may indicate the potential presence of alternative reservoir hosts with longer life expectancy such as dogs and cats, where Lu. cruciata may be able to maintain the transmission cycle. Non-sylvatic cycles imply that Leishmania spp. can infect multiple domestic animals in the Americas, further indicating that the diversity of the Phlebotominae feed sources are an important source of Leishmania spp.’s adaptation to new domestic hosts [66,67]. The process of adaptation to new hosts implicates that filling the knowledge gaps on feeding ecology, demography and reservoir host diversity is critical for disease management, as the preventive and mitigation strategies are radically different for zoonoses than diseases with human-to-human transmission.

Leishmaniasis in Europe, Asia and Africa is an anthropozoonosis, whereas in the Americas it is still considered a zoonosis where the main risk factor is carrying out forest-related activities (i.e., agriculture, catlle raising, etc), making it a significant occupational disease. Currently, there is no evidence that humans are Leishmania spp. reservoirs in the Neotropics, although the high prevalence of asymptomatic human cases in the Yucatan Peninsula [22,26], and the frequent detection of live parasites in these [68] begs the question whether humans are in fact able to transmit Le. mexicana to vectors. If humans are proven to be competent reservoirs, public health policies should then target human-to-human and zoonotic transmission. In terms of our model, this novel characteristic would change the role of Lu. cruciata from driving zoonotic transmission and potential dilution (e.g., [69]) to maintaining Le. mexicana populations among humans and domestic reservoir hosts.

Leishmaniasis incidence among humans is still poorly characterised. Early analyses have found 508 cases per 100,000 inhabitants, but prevalence as high as 43% [26]. The prevalence and incidence figures produced by our model are nowhere near those values. First because we did not seek to estimate them but to show the effect of feeding preferences of different vector species while highlighting its epidemiological implications. As shown in the supplementary materials, if the simulation is let run for many more time units, the prevalence among humans is ~ 45% (S2 Fig), although this does not account for seasonality for the above reasons. Policy-wise, we do not recommend using our model for decision making, but to incentivise reports and the funding of epidemiological studies to clarify if such figures are still patent among populations at risk.

As seen above, the role of Lu. cruciata and other sand fly species should not be minimised in spite of current evidence, suggesting that Bi. olmeca maintains Le. mexicana infection (with the parameters in Table 1). The high prevalence observed in Lu. cruciata, its widespread distribution and high adaptability to human-modified landscapes [39,42] where domestic animal bloodmeals are abundant, are compelling arguments to consider it the pivotal vector towards a non-sylvatic transmission cycle. Further empirical evidence of the possibility of non-sylvatic cycles has been recorded in recent outbreaks associated with sand fly species not considered in this study in Yucatán [39,42]. The ability of Le. mexicana to infect a wide range of vectors and both reservoir and dead-end hosts has evolutionary consequences. For instance, we infer that on the one hand, specialist vectors like Bi. olmeca act as efficient maintenance vectors among reservoirs; on the other, generalists like Lu. cruciata help Le. mexicana to increase its range of hosts. These properties acquire even more relevance under the current process of intense land conversion across the study area [70], which increases exposure of humans and domestic animals to Le. mexicana.

As is evident up to this point, there are several uncertainties in how Leishmania spp. parasites are maintained in the study area where we identify a large set of open questions relevant from a public health, ecological and environmental point of view: 1) lack of knowledge about specific competition between vector species; 2) inability to describe the response of the different sand fly species’ feeding preference to the abundance and availability of different bloodmeals; 3) the response of sand flies and rodent reservoirs to land use change and human population growth; 4) the relationship between sand fly species and domestic or synanthropic wild mammals; 5) the role of humans in the transmission dynamics, mainly the asymptomatic population with viable Leishmania spp. parasites; and 6) the effect that all of the previous limitations have on contact and transmission rates. Previous analyses show that leishmaniasis’ epidemiological parameters can be very variable, even among neighboring communities [45], which may be caused by some of the phenomena listed above and thus speak of the necessity to comprehensively characterize the eco-epidemiology of leishmaniasis

Conclusions

Our results highlight the importance of the feeding ecology of sand fly species on the Le. mexicana transmission dynamics in the Yucatán Peninsula. While Bi. olmeca is essential for maintaining the infection among reservoirs, Lu. cruciata and other generalist species drive zoonotic transmission and dilution among non-competent dead-end hosts, which could in turn expand the parasite’s range of reservoir hosts. By adapting to these new domestic hosts, the generalist species could act as ecological bridges, facilitating the emergence of novel transmission cycles and increasing the risk of infection in humans. The land conversion process currently accelerating across the study area increases the risk of emergence of non-sylvatic transmission cycles. Therefore, increasing our knowledge basis to understand the process of Le. mexicana transmission dynamics, to generate applicable solutions for the public health challenges the Yucatan Peninsula is facing is of utmost importance.

Supporting information

S1 Fig. Timelines of the three rodent species population densities used to estimate parameters rR, KR and mR.

https://doi.org/10.1371/journal.pntd.0013786.s001

(EPS)

S2 Fig. Model simulation for 2500 days to obtain prevalence at stability conditions.

https://doi.org/10.1371/journal.pntd.0013786.s002

(EPS)

S3 Fig. Sobol indices for the full 150 day simulation period.

Top panel shows First order indices obtained by keeping all parameters fixed except the target parameter. The bottom panel shows total sensitivity, obtained by simultaneous Monte Carlo simulation of all parameters.

https://doi.org/10.1371/journal.pntd.0013786.s003

(EPS)

References

  1. 1. Reithinger R, Dujardin J-C, Louzir H, Pirmez C, Alexander B, Brooker S. Cutaneous leishmaniasis. Lancet Infect Dis. 2007;7(9):581–96. pmid:17714672
  2. 2. Murray HW, Berman JD, Davies CR, Saravia NG. Advances in leishmaniasis. Lancet. 2005;366(9496):1561–77. pmid:16257344
  3. 3. Karimkhani C, Wanga V, Coffeng LE, Naghavi P, Dellavalle RP, Naghavi M. Global burden of cutaneous leishmaniasis: a cross-sectional analysis from the Global Burden of Disease Study 2013. Lancet Infect Dis. 2016;16(5):584–91. pmid:26879176
  4. 4. Grifferty G, Shirley H, McGloin J, Kahn J, Orriols A, Wamai R. Vulnerabilities to and the Socioeconomic and Psychosocial Impacts of the Leishmaniases: A Review. Res Rep Trop Med. 2021;12:135–51. pmid:34188584
  5. 5. Alvar J, Croft SL, Kaye P, Khamesipour A, Sundar S, Reed SG. Case study for a vaccine against leishmaniasis. Vaccine. 2013;31 Suppl 2:B244-9. pmid:23598489
  6. 6. Pan American Health Organization. Leishmaniases. Epidemiological Report of the Americas, December 2019. Leishmaniases Report;8. 2019. Available from: https://iris.paho.org/handle/10665.2/51734
    • 7. Ready PD. Biology of phlebotomine sand flies as vectors of disease agents. Annu Rev Entomol. 2013;58:227–50. pmid:23317043
    • 8. Instructions on how to make an Outbreak of American Cutaneous Leishmaniasis. J Trop Med Health. 2019;3(1).
    • 9. Ashford RW. The leishmaniases as model zoonoses. Ann Trop Med Parasitol. 1997;91(7):693–701.
    • 10. Saliba EK, Oumeish OY. Reservoir hosts of cutaneous leishmaniasis. Clin Dermatol. 1999;17(3):275–7. pmid:10384866
    • 11. Pan American Health Organization. Leishmaniasis: Epidemiological Report of the Americas, No. 10 (December 2021). Epidemiological Report of the Americas. 2021. Available from: https://iris.paho.org/handle/10665.2/55368
      • 12. Pan American Health Organization. Leishmaniasis: Epidemiological Report for the Americas. No.11 (December 2022). Epidemiological Report of the Americas. 2022. Available from: https://iris.paho.org/handle/10665.2/56831
        • 13. Pan American Health Organization. Leishmaniases. Epidemiological Report in the Americas, July 2016. Leishmaniases Report;4. 2016. Available from: https://iris.paho.org/handle/10665.2/51655
          • 14. Pan American Health Organization. Leishmaniases: Epidemiological Report in the Americas, No. 3 (July 2015). Leishmaniases Report. 2015. Available from: https://iris.paho.org/handle/10665.2/51747
            • 15. Pan American Health Organization. Leishmaniases: Epidemiological Report in the Americas, No. 2 (June 2014). Leishmaniases Report. 2014. Available from: https://iris.paho.org/handle/10665.2/51679
              • 16. Pan American Health Organization. Leishmaniases: Epidemiological Report in the Americas, No. 1 (April 2013). Leishmaniases Report. 2013. Available from: https://iris.paho.org/handle/10665.2/51680
                • 17. Pan American Health Organization. Epidemiological Report of the Americas. Leishmaniases, April 2017. Washington, DC: PAHO; 2017. Available from: https://iris.paho.org/handle/10665.2/34112
                  • 18. Pan American Health Organization. Leishmaniases. Epidemiological Report in the Americas, July 2016. Leishmaniases Report;4. 2016. Available from: https://iris.paho.org/handle/10665.2/51655
                    • 19. Pan American Health Organization. Leishmaniases. Epidemiological Report in the Americas, February 2018. Leishmaniases Report;6. 2018. Available from: https://iris.paho.org/handle/10665.2/51749
                      • 20. Canché-Pool EB, Panti-May JA, Ruiz-Piña HA, Torres-Castro M, Escobedo-Ortegón FJ, Tamay-Segovia P, et al. Cutaneous Leishmaniasis Emergence in Southeastern Mexico: The Case of the State of Yucatan. Trop Med Infect Dis. 2022;7(12):444. pmid:36548699
                      • 21. Mikery-Pacheco OF, Castillo-Vera A. Leishmaniasis. La frontera sur de México, ¿una salud en crisis? México, DF: Inter sistemas; 2018. p. 163–75.
                        • 22. Loría-Cervera EN, Sosa-Bibiano EI, Van Wynsberghe NR, Torres-Castro JR, Andrade-Narváez FJ. Preliminary epidemiological findings of Leishmania infection in the municipality of Tinum, Yucatan State, Mexico. Parasite Epidemiol Control. 2019;4:e00088. pmid:30705976
                        • 23. Cañeda-Guzmán IC, de Oca-Aguilar ACM, Miranda-Caballero CI, Grostieta E, Correa-Morales F, Romero-Pérez R, et al. Entomological Survey and Leishmania (Leishmania) mexicana Prevalence in Sand Fly Species during an Outbreak of Cutaneous Leishmaniasis in Quintana Roo State, Mexico. Trop Med Infect Dis. 2023;8(10):465. pmid:37888593
                        • 24. Beltran E. Cutaneous Leishmaniasis in Mexico. Scientific Monthly. 1944;59: 108–119. Available from: https://www.jstor.org/stable/18397
                        • 25. Andrade-Narvaez FJ, Van Wynsberghe NR, Sosa-Bibiano EI, Loria-Cervera EN. Eco-Epidemiological and Immunological Features of Localized Cutaneous Leishmaniasis in Southeastern Mexico: Thirty Years of Study. In: Claborn D, editor. The Epidemiology and Ecology of Leishmaniasis. InTech; 2017.
                        • 26. Andrade-Narváez FJ, Simmonds-Díaz E, Rico-Aguilar S, Andrade-Narvéez M, Palomo-Cetina A, Canto-Lara SB, et al. Incidence of localized cutaneous leishmaniasis (chiclero’s ulcer) in Mexico. Trans R Soc Trop Med Hyg. 1990;84(2):219–20. pmid:2389310
                        • 27. Secretaria de Salud P sectorial de S. Programa de Acción Específico: Prevención y Control de las Leishmaniasis 2013-2018. Ciudad de México: Secretaría de Salud; 2013. Available from: https://www.gob.mx/cms/uploads/attachment/file/38245/PAE_PrevencionControlLeishmaniasis2013_2018.pdf
                          • 28. Andrade-Narvaez FJ, Canto Lara SB, Van Wynsberghe NR, Rebollar-Tellez EA, Vargas-Gonzalez A, Albertos-Alpuche NE. Seasonal transmission of Leishmania (Leishmania) mexicana in the state of Campeche, Yucatan Peninsula, Mexico. Mem Inst Oswaldo Cruz. 2003;98(8):995–8. pmid:15049078
                          • 29. Merriam CH. Descriptions of eigth new ground squirrels of the genera Spermophilus and Tamias from California, Texas and Mexico. Proc Biol Soc Washington. 1893;8:129–38.
                          • 30. Allen JA, Chapman FM. On mammals from Yucatan, with descriptions of new species. American Museum of Natural History; 1897.
                            • 31. Chable-Santos JB, Van Wynsberghe NR, Canto-Lara SB, Andrade-Narvaez FJ. Isolation of Leishmania (L.) mexicana from wild rodents and their possible role in the transmission of localized cutaneous leishmaniasis in the state of Campeche, Mexico. Am J Trop Med Hyg. 1995;53(2):141–5. pmid:7677214
                            • 32. Sosa-Bibiano EI, Sánchez-Martínez LA, López-Ávila KB, Chablé-Santos JB, Torres-Castro JR, Fernández-Figueroa EA, et al. Leishmania (Leishmania) mexicana Infection in Wild Rodents from an Emergent Focus of Cutaneous Leishmaniasis in Yucatan, Mexico. J Trop Med. 2022;2022:8392005. pmid:35686208
                            • 33. Pech-May A, Escobedo-Ortegón FJ, Berzunza-Cruz M, Rebollar-Téllez EA. Incrimination of four sandfly species previously unrecognized as vectors of Leishmania parasites in Mexico. Med Vet Entomol. 2010;24(2):150–61. pmid:20604861
                            • 34. Pech-May A, Peraza-Herrera G, Moo-Llanes DA, Escobedo-Ortegón J, Berzunza-Cruz M, Becker-Fauser I, et al. Assessing the importance of four sandfly species (Diptera: Psychodidae) as vectors of Leishmania mexicana in Campeche, Mexico. Med Vet Entomol. 2016;30(3):310–20. pmid:27040367
                            • 35. Montes de Oca-Aguilar AC, Sosa-Bibiano E, López-Ávila KB, Torres-Castro JR, Loría-Cervera EN. Registro de flebotomíneos antropofílicos en el peridomicilio de un caso de leishmaniosis cutánea localizada en Yucatán, México. Rev Biomed. 2023;34(1):104–10.
                            • 36. Montes de Oca-Aguilar AC, Fernández-Figueroa EA, López-Ávila KB, Pavón-Méndez MI, Sosa-Bibiano EI, Rebollar-Téllez EA, et al. Abundance and Leishmania infection patterns of the sand fly Psathyromyia cratifer in Southern Mexico. PLoS Negl Trop Dis. 2024;18(9):e0012426. pmid:39255321
                            • 37. Vargas L, Diaz YA. Flebotomos comunes a Mexico y a los Estados Unidos. Instituto de Salubridad y Enfermedades Tropicales; 1953. p. 494–8.
                              • 38. Rebollar-Téllez EA, Ramírez-Fraire A, Andrade-Narvaez FJ. A two years study on vectors of cutaneous Leishmaniasis. Evidence for sylvatic transmission cycle in the state of Campeche, Mexico. Mem Inst Oswaldo Cruz. 1996;91(5):555–60. pmid:9137741
                              • 39. Montes de Oca-Aguilar AC, Euan-Canul RD, Sosa-Bibiano EI, López-Ávila KB, Rebollar-Téllez EA, Palacio-Vargas JA, et al. Phlebotomine sand flies in rural Mayan communities of Southern Mexico: The heterogeneity of the ruralscape increases the entomological risk. Acta Trop. 2024;249:107051. pmid:37875169
                              • 40. Montes de Oca-Aguilar AC, Sosa-Bibiano E, López-Ávila KB, Torres-Castro JR, Loría-Cervera EN. Registro de flebotomíneos antropofílicos en el peridomicilio de un caso de leishmaniosis cutánea localizada en Yucatán, México. Rev Biomed. 2023;34(1):104–10.
                              • 41. Sánchez-García L, Berzunza-Cruz M, Becker-Fauser I, Rebollar-Téllez EA. Sand flies naturally infected by Leishmania (L.) mexicana in the peri-urban area of Chetumal city, Quintana Roo, México. Trans R Soc Trop Med Hyg. 2010;104(6):406–11. pmid:20171709
                              • 42. de Oca-Aguilar ACM, Fernández-Figueroa EA, Pavón-Méndez MI, López-Ávila K, Sosa-Bibiano EI, Rebollar-Téllez EA, et al. First detection of Leishmania DNA in Lutzomyia longipalpis sensu lato (Diptera: Phlebotominae) in southem Mexico. J Vector Borne Dis. 2023;60(4):453–6. pmid:38174527
                              • 43. Montes de Oca-Aguilar AC, Pavón-Mendez MI, López-Ávila KB, Sosa-Bibiano EI, Rebollar-Téllez EA, Palacio-Vargas JA, et al. Biting rhythms and infection rates of anthropophilic sand fly species (Diptera: Phlebotominae) in sites with different land use in southern Mexico. Acta Trop. 2023;248:107014. pmid:37696485
                              • 44. Chaves LF, Hernandez M-J, Dobson AP, Pascual M. Sources and sinks: revisiting the criteria for identifying reservoirs for American cutaneous leishmaniasis. Trends Parasitol. 2007;23(7):311–6. pmid:17524806
                              • 45. Rabinovich JE, Feliciangeli MD. Parameters of Leishmania braziliensis transmission by indoor Lutzomyia ovallesi In Venezuela. Am J Trop Med Hyg. 2004;70(4):373–82.
                              • 46. Barradas I, Caja Rivera RM. Cutaneous leishmaniasis in Peru using a vector‐host model: Backward bifurcation and sensitivity analysis. Math Method App Sci. 2018;41(5):1908–24.
                              • 47. Restif O, Hayman DTS, Pulliam JRC, Plowright RK, George DB, Luis AD, et al. Model-guided fieldwork: practical guidelines for multidisciplinary research on wildlife ecological and epidemiological dynamics. Ecol Lett. 2012;15(10):1083–94. pmid:22809422
                              • 48. World Health Organization. Report of the tenth meeting of the WHO Strategic and Technical Advisory Group for Neglected Tropical Diseases. Geneva; 2017.
                                • 49. Castillo A, Serrano AK, Mikery OF, Pérez J. Life history of the sand fly vector Lutzomyia cruciata in laboratory conditions. Med Vet Entomol. 2015;29(4):393–402. pmid:26147368
                                • 50. Stauch A, Duerr H-P, Picado A, Ostyn B, Sundar S, Rijal S, et al. Model-based investigations of different vector-related intervention strategies to eliminate visceral leishmaniasis on the Indian subcontinent. PLoS Negl Trop Dis. 2014;8(4):e2810. pmid:24762676
                                • 51. Martinez-Niño MA, Camacho-Galván JR. Cutaneous Leishmaniasis in a Non-endemic Area in Mexico. 2023.
                                  • 52. Hernández-Betancourt SF, Cimé-Pool JA. Ecología poblacional de Heteromys gaumeri en la selva del sur de Yucatán, México. 2008.
                                    • 53. Hernández-Betancourt SF, Cimé Pool JA, Medina-Peralta S, González-Villanueva ML. Fluctuación poblacional de ototylomys phyllotis merriam, 1901 (rodentia: muridae) en una selva mediana subcaducifolia del Sur de Yucatán, México. AZM. 2008;24(2).
                                    • 54. Hernández-Betancourt SF, Cimé-Pool JA, Medina-Peralta S, Durán-Miranda CM. Parámetros poblacionales del ratón yucateco Peromyscus yucatanicus de una selva baja caducifolia del norte de Yucatán, México. Estudios sobre la Biología de Roedores Silvestres Mexicanos. Ciudad de México: UNAM; 2012.
                                      • 55. Berry RJ, Jakobson ME, Triggs GS. Survival in wild‐living mice*. Mammal Rev. 1973;3(2):46–57.
                                      • 56. Grenfell B, Keeling M. Dynamics of infectious disease. 3rd ed. In: May RM, McLean A, editors. Theoretical Ecology: Principles and Applications. 3rd ed. Chippenham: Oxford Univestity Press; 2007. p. 132–147.
                                        • 57. Keeling MJ, Rohani P. Modelling Infectious Diseases in Humans and Animals. 1st ed. New Jersey: Princeton University Press; 2007.
                                          • 58. Tatem AJ. WorldPop, open data for spatial demography. Sci Data. 2017;4:170004.
                                          • 59. Theers S, Weber F, Surmann D. ODEsensitivity:: Sensitivity Analysis of Ordinary Differential Equations. Available from: https://cran.r-project.org/package=ODEsensitivity
                                            • 60. Chaves LF, Hernandez M-J. Mathematical modelling of American cutaneous leishmaniasis: incidental hosts and threshold conditions for infection persistence. Acta Trop. 2004;92(3):245–52. pmid:15533294
                                            • 61. Pérez-Blas LG, Chiyean-Acosta ÁG, Canché-Pool EB, Tello-Martín R, Torres-Castro JR, Ruiz-Piña HA, et al. Molecular Detection of Leishmania (Leishmania) mexicana in Sandflies from the State of Yucatan, Mexico. Vector Borne Zoonotic Dis. 2022;22(12):589–95. pmid:36399687
                                            • 62. Seidelin H. LeishmaniasisandBabesiasisin Yucatan. Ann Trop Med Parasitol. 1912;6(2):295–300.
                                            • 63. Organización Panamericana de la Salud, Organización Panamericana de la Salud. Leishmaniasis: Informe epidemiológico de las Américas. Núm. 11 (Diciembre del 2022). Epidemiological Report of the Americas. 2022. Available from: https://iris.paho.org/handle/10665.2/56833
                                              • 64. López-Céspedes A, Longoni SS, Sauri-Arceo CH, Sánchez-Moreno M, Rodríguez-Vivas RI, Escobedo-Ortegón FJ, et al. Leishmania spp. epidemiology of canine leishmaniasis in the Yucatan Peninsula. Scientific World J. 2012;2012:945871. pmid:22927792
                                              • 65. Hernández-Rivera MP, Hernández-Montes O, Chiñas-Pérez A, Batiza-Avelar JM, Social A, Sánchez-Tejeda G, et al. Study of cutaneous leishmaniasis in the State of Campeche (Yucatan Peninsula), Mexico, over a period of two years. Salud Pública de México. 2015;57.
                                              • 66. Dantas-Torres F. The role of dogs as reservoirs of Leishmania parasites, with emphasis on Leishmania (Leishmania) infantum and Leishmania (Viannia) braziliensis. Vet Parasitol. 2007;149(3–4):139–46. pmid:17703890
                                              • 67. Dantas-Torres F. Canine leishmaniasis in the Americas: etiology, distribution, and clinical and zoonotic importance. Parasit Vectors. 2024;17(1):198. pmid:38689318
                                              • 68. Loría-Cervera EN, Sosa-Bibiano EI, López-Ávila KB, de Oca-Aguilar ACM, Moreno-Nava MS, Torres-Castro JR. Viable Leishmania parasites in the absence of an in vitro IFN-γ response in asymptomatic carriers. Rev Inst Med Trop Sao Paulo. 2024;66:e13. pmid:38381898
                                              • 69. Ostfeld RS, Keesing F. Biodiversity and Disease Risk: the Case of Lyme Disease. Conserv Biol. 2000;14(3):722–8.
                                              • 70. Ellis EA, Romero-Montero JA, Hernández-Gómez IU. Evaluación y mapeo de los determinantes de la deforestación en la Península Yucatán. Ciudad de México: Observatorio de la Selva Maya; 2015. Available from: https://sis.cnf.gob.mx/wp-content/plugins/conafor-files/2018/nacional/catalogo/biblioteca/154.pdf