Climate change is aggravating dengue and yellow fever transmission risk

Abstract

Dengue and yellow fever have complex cycles, involving urban and sylvatic mosquitoes, and non‐human primate hosts. To date, efforts to assess the effect of climate change on these diseases have neglected the combination of such crucial factors. Recent studies only considered urban vectors. This is the first study to include them together with sylvatic vectors and the distribution of primates to analyse the effect of climate change on these diseases. We used previously published models, based on machine learning algorithms and fuzzy logic, to identify areas where climatic favourability for the relevant transmission agents could change: 1) favourable areas for the circulation of the viruses due to the environment and to non‐human primate distributions; 2) the favourability for urban and sylvatic vectors. We obtained projections of future transmission risk for two future periods and for each disease, and implemented uncertainty analyses to test for predictions reliability. Areas currently favourable for both diseases could keep being climatically favourable, while global favourability could increase a 7% for yellow fever and a 10% increase for dengue. Areas likely to be more affected in the future for dengue include West Africa, South Asia, the Gulf of Mexico, Central America and the Amazon basin. A possible spread of dengue could take place into Europe, the Mediterranean basin, the UK and Portugal; and, in Asia, into northern China. For yellow fever, climate could become more favourable in Central and Southeast Africa; India; and in north and southeast South America, including Brazil, Paraguay, Bolivia, Peru, Colombia and Venezuela. In Brazil, favourability for yellow fever will probably increase in the south, the west and the east. Areas where the transmission risk spread is consistent to the dispersal of vectors are highlighted in respect of areas where the expected spread is directly attributable to environmental changes. Both scenarios could involve different prevention strategies.

Full text

www.ecography.org
ECOGRAPHY
Ecography
Page 1 of 9
is is an open access article under the terms of the Creative Commons
Attribution License, which permits use, distribution and reproduction in any
medium, provided the original work is properly cited.
Subject Editor: Timothée Poisot
Editor-in-Chief: Jean-François
Guégan Accepted: 21 May 2024
doi: 10.1111/ecog.06942
2024
1–9
2024: e06942
Disease Ecology Special Issue
Disease Ecology Special Issue
© 2024 e Authors. Ecography published by John Wiley & Sons Ltd on behalf of Nordic Society
Oikos
Dengue and yellow fever have complex cycles, involving urban and sylvatic mosquitoes,
and non-human primate hosts. To date, eorts to assess the eect of climate change on
these diseases have neglected the combination of such crucial factors. Recent studies
only considered urban vectors. is is the rst study to include them together with syl-
vatic vectors and the distribution of primates to analyse the eect of climate change on
these diseases. We used previously published models, based on machine learning algo-
rithms and fuzzy logic, to identify areas where climatic favourability for the relevant
transmission agents could change: 1) favourable areas for the circulation of the viruses
due to the environment and to non-human primate distributions; 2) the favourability
for urban and sylvatic vectors. We obtained projections of future transmission risk
for two future periods and for each disease, and implemented uncertainty analyses to
test for predictions reliability. Areas currently favourable for both diseases could keep
being climatically favourable, while global favourability could increase a 7% for yel-
low fever and a 10% increase for dengue. Areas likely to be more aected in the future
for dengue include West Africa, South Asia, the Gulf of Mexico, Central America
and the Amazon basin. A possible spread of dengue could take place into Europe, the
Mediterranean basin, the UK and Portugal; and, in Asia, into northern China. For
yellow fever, climate could become more favourable in Central and Southeast Africa;
India; and in north and southeast South America, including Brazil, Paraguay, Bolivia,
Peru, Colombia and Venezuela. In Brazil, favourability for yellow fever will probably
increase in the south, the west and the east. Areas where the transmission risk spread
is consistent to the dispersal of vectors are highlighted in respect of areas where the
expected spread is directly attributable to environmental changes. Both scenarios could
involve dierent prevention strategies.
Keywords: Biogeography, host–pathogen systems, pathogen spillover, vector-borne
disease ecology, zoonotic diseases
Climate change is aggravating dengue and yellow fever
transmission risk
Alisa Aliaga-Samanez ✉1, David Romero1, Kris Murray2,3, Marina Cobos-Mayo1, Marina Segura4,
Raimundo Real1,5 and Jesús Olivero1,5
1Grupo de Biogeografía, Diversidad y Conservación, Departamento de Biología Animal, Universidad de Málaga, Malaga, Spain
2Medical Research Council Unit the Gambia at London School of Hygiene and Tropical Medicine, Fajara, Gambia
3Centre on Climate Change and Planetary Health, London School of Hygiene & Tropical Medicine, London, UK
4Centro de Vacunación Internacional, Ministerio de Sanidad, Consumo y Bienestar Social, Estación Marítima, Malaga, Spain
5Instituto IBYDA, Centro de Experimentación Grice-Hutchinson, Malaga, Spain
Correspondence: Alisa Aliaga-Samanez (alisaliaga@uma.es)
Research article
9

Page 2 of 9
Disease Ecology Special Issue
Disease Ecology Special Issue
Introduction
Dengue and yellow fever are among the deadliest mosquito-
borne diseases worldwide (Kuno 2015, Colón-Gonzálezetal.
2021). ese diseases are responsible for approximately 390
million and 200 000 cases per year, respectively (CDC
2018, WHO 2021a), and they continue to spread, causing
outbreaks in areas from where they had disappeared and in
new areas, even if, as in the case of yellow fever, an eec-
tive vaccine already exists. is spread has several causes: cli-
mate change, forest loss, increased forest incursions, mining
and oil extraction, construction and land clearing for agri-
culture (Daszaketal. 2000, Ka-Wai Hui 2006, Kareshetal.
2012, Rohr et al. 2019). Climate change is exposing peo-
ple worldwide to increasing threats of vector-borne diseases
(Watts etal. 2019). For this reason, climate change is one
of the most daunting 21st century global health challenges
(Iwamuraetal. 2020), and is one of the key issues considered
by the Global Strategy to Eliminate Yellow Fever Epidemics
(EYE) 2017–2026, which is managed by WHO, Gavi and
UNICEF (WHO 2021b). It is dicult to predict how global
warming will aect dengue and yellow fever because they have
very complex cycles, involving dierent mosquito species,
both urban and sylvatic, and also non-human primate hosts
(Gaythorpeetal. 2020). Many studies on climate change and
mosquito-borne diseases do not take into account multiple
sources of uncertainty in their predictions, such as health,
environmental and socio-economic data, future global change
scenarios and model structure (Franklinos et al. 2019).
Franklinos and colleagues argue that an integrative approach
that takes into account interactions between socio-economic
and environmental systems is needed to better understand and
predict mosquito-borne disease risk. Future projections have
been made to assess the potential extent of dengue and yellow
fever. For dengue, the most recent studies only included Aedes
aegypti and Ae. albopictus as possible vectors. is is the case
of Messinaetal. (2019), who focused on changes in environ-
mental suitability, and of Colón-Gonzálezetal. (2021), who
quantied the extent to which climate change could inu-
ence the length of the transmission season. e most recent
research on how climate change will aect yellow fever is that
of Gaythorpe etal. (2020). In it, eects on the morbidity
of yellow fever in Africa were assessed considering only one
urban vector species, Ae. aegypti, in the set of covariates. On
the other hand, Hamletetal. (2018) tested whether seasonal
variations in climatic factors are associated with the seasonal-
ity of yellow fever reports. Again, only Ae. aegypti was con-
sidered as a vector in the analysis. Integrating all the agents
involved in the zoonotic cycles is important in order to get as
close as possible to having reliable future projections. In this
research, we aim to detect areas worldwide where changes in
the risk for dengue and yellow fever transmission are expected
to occur in the short and medium terms as a consequence of
climate change, on the basis of the dengue and yellow fever
transmission risk models published by Aliaga-Samanezetal.
(2021, 2022). In this way, projections are built taking into
account the most updated database of case reports to date,
and considering both urban and sylvatic mosquito vectors,
together with the distribution of non-human primates.
Material and methods
Methodological and temporal framework
Our forecasts consisted of projections to the future of the
dengue and yellow fever transmission models published by
Aliaga-Samanezetal. (2021, 2022), which were focused on
the distribution of transmission risk in the period 2001–
2017. ese models were based on the favourability function.
Favourability reects the degree to which the probability of
occurrence of the analysed entity diers from that expected
according to the initial prevalence (Realetal. 2006). So, in
contrast to probability, favourability values depend exclu-
sively on the eect of environmental conditions in the dis-
tribution area under analysis (Acevedo and Real 2012). In
Aliaga-Samanezetal. (2021, 2022), the degree of favourabil-
ity for the transmission of a disease (i.e. the level of transmis-
sion risk) was considered to be a result of combining a vector
model (which dened favourability values for the presence
of mosquito vectors) with a disease model (which dened
favourability values for the occurrence of disease cases in
humans). is combination was made using the fuzzy inter-
section operator (Zadeh 1965) which assigns, to each geo-
graphic unit, the lowest favourability value provided by each
model. is ensured the existence of suitable circumstances
for the presence of two agents that are geographically limit-
ing: 1) vectors, and 2) the environmental conditions needed
for a pathogen to cause disease (Romeroetal. 2016, Aliaga-
Samanezetal. 2021, 2022). e projection of a transmission
model to a future period was, thus, the result of combining
– also through the fuzzy intersection – future projections of
both a vector model and a disease model.
As projections to the future regarding the presence of
vectors, we used those developed by Aliaga-Samanezet al.
(2023) for dengue and yellow fever. ese models considered
urban vector species (Ae. aegypti and Ae. albopictus) together
with sylvatic vector species Ae. africanus, Ae. luteocephalus, Ae.
niveus, Ae. vittatus, Sabethes chloropterus, Haemagogus leucoce-
laenus and Hg. janthinomys (Supporting information), com-
bined according to the known set of sylvatic species involved
in the zoonotic cycles of each disease. Details of the projec-
tion to future of Aliaga-Samanezetal. (2021, 2022) disease
models are given below.
Projections were made considering two time periods:
2041–2060 (henceforth the ‘near future period’), and 2061–
2080 (henceforth the ‘far future period’). Future projections
were mapped to the same worldwide grid of 18 874 hexagons
of 7774 km2 (https://zenodo.org/records/10028166) used in
the original models by Aliaga-Samanezetal. (2021, 2022).
Disease model projection to the future
We used dierent climate change scenarios (i.e. representative
CO2 concentration pathways, RCPs) and atmosphere–ocean

Page 3 of 9
Disease Ecology Special Issue
Disease Ecology Special Issue
general circulation models (GCMs) to account for uncertain-
ties in our forecasts, based on a range of variation in climate
predictions according to the Intergovernmental Panel on
Climate Change (IPCC). Using as information source the free
high-resolution climate data from CHELSA (Knutti etal.
2013, Kargeretal. 2017, 2020, 2021), two RCP scenarios
were chosen: the RCP 4.5 stabilized emissions scenario,
and the RCP 8.5 high emissions scenario. Five GCMs with
the lowest detected biases with respect to the actual climate
(McSweeneyet al. 2015, Sandersonet al. 2015), and with
data available for the chosen RCPs, were selected: CESM1-
CAM5, CNRM-CM5, FIO-ESM, GFDL-CM3 and MPI-
ESM-LR. So, for every disease model and future period, we
got 10 projections, corresponding to 10 RCP-GCM com-
binations. In Aliaga-Samanezetal. (2023), the same RCP-
GCM combinations and the same future periods were used
to make future projections for dengue and yellow fever vector
distributions.
To obtain the future forecasting for each disease, a disease
model was projected according to the following favourability
(F) function (Realetal. 2006):
Fynn y







ex
pe
xp10
So forecasts are based on models built on information
regarding the early 21st century, in this equation: n1 and n0
are the number of presences and absences of the modelled
entity (in this case, of dengue or of yellow fever case reports)
in such a time period. As explained by Aliaga-Samanezetal.
(2021, 2022), this information was obtained, in the case of
dengue, from Messinaetal. (2019) and completed with data
from various sources such as WHO, ECDC, Promedmail,
Gideon and scientic articles (see Aliaga-Samanez et al.
(2021) for data and source details; and Supporting informa-
tion for geo-referenced information on case location). In the
case of yellow fever, information came from Sheareretal.
(2018) and supplemented with the same sources mentioned
above (Aliaga-Samanez et al. (2022) for data and source
details; and Supporting information for geo-referenced
information on case location). In the above F function, y
is a linear combination of predictor variables (Supporting
information). Taking into account the y equation of the
disease models in Aliaga-Samanezetal. (2021, 2022), we
calculated future favourability values replacing the values of
climate variables according to the future scenarios described.
Although it is a strong statement that could involve inter-
pretative limitations, we assumed that the values of non-
climate variables (i.e. human concentration, infrastructures,
livestock, topography, agriculture or ecosystem types) will
not change in the future period considered (Supporting
information). e y equations also include variables refer-
ring to the biogeography of non-primate hosts participating
in the sylvatic zoonotic cycles (Aliaga-Samanezetal. 2021,
2022; see Supporting information). ese variables were
represented by primate chorotypes, i.e. types of distribu-
tions (Oliveroetal. 2011, 2017). Because primate distribu-
tions could be subject to changes due to global warming,
these chorotypes were also projected into the future. For
this purpose, a reformulation of chorotypes was needed, for
which we made the following steps:
1) In total, 14 primate species belonging to the chorotypes
that formed part of the models in Aliaga-Samanezetal.
(2021, 2022) were considered (Table 1).
2) Favourability models were made for the current distribu-
tion of each primate species. For this purpose, range maps
of the African, Asian and American primate species were
obtained from the IUCN (IUCN 2021).
3) Chorotype variables were then recalculated through the
accumulated favourability value (Faet al. 2014), also
named ‘fuzzy species richness’ (Estrada et al. 2008);
that is, in every geographic unit (i.e. in each hexagon),
favourability values for each primate species forming
part of the chorotype were summed up (Supporting
information).
4) We ensured that the original chorotypes (based on species
richness) was consistent with the reformulated chorotypes
(based on fuzzy species richness) by testing for signicance
of Spearman correlations.
Table 1. Chorotypes considered for disease models projection to the future. AS: Asia; SA: South America; AF: Africa.
Diseases Chorotypes Genera
Dengue AS19 Carlito
AS8 Hylobates, Trachypithecus, Nomascus, Pygathrix
AF2 Arctocebus, Lophocebus, Sciurocheirus, Gorilla, Euoticus, Cercopithecus, Colobus, Mandrillus, Miopithecus
SA2 Alouatta, Ateles, Plecturocebus, Chiropotes, Mico
AS15 Hylobates, Trachypithecus, Presbytis, Nycticebus
SA4 Alouatta, Sapajus, Brachyteles, Callithrix, Callicebus, Leontophitecus
SA5 Aotus, Cebus, Ateles, Saguinus
Yellow fever SA7 Plecturocebus, Pithecia, Leontocebus
SA1 Callimico, Cebus, Cebuella, Saguinus, Leontocebus, Saimiri, Sapajus
SA6 Alouatta, Aotus, Ateles, Cheracebuss, Saguinus, Saimiri, Leontocebus
SA12 Alouatta, Plecturocebus, Mico, Saguinus
AF9 Cercopithecus, Chlorocebus, Galago, Otolemur, Papio
Both diseases SA8 Alouatta, Cebus, Chiropotes, Saguinus
SA14 Leontocebus, Plecturocebus, Aotus, Lagothrix

Page 4 of 9
Disease Ecology Special Issue
Disease Ecology Special Issue
Once we had favourability models for all primate species,
the chorotype projection to the future consisted of summing
up the projected favourability values for each primate species
in the chorotype. is model projection followed the same
guidelines explained above.
e replacement, in the original disease model, of choro-
types based on species richness with chorotypes based on
fuzzy species richness made it necessary to make a model
recalibration. at is, as the mathematical expression of the
chorotype variable was varied, the coecients of all vari-
ables in the disease models could also change. So, the disease
model was run again xing the same set of predictor variables
as in Aliaga-Samanezetal. (2021, 2022), only replacing the
chorotypes as redened (see the Supporting information for
mathematical models). e resulting recalibrated model was
the base for all future projections.
Finally, we obtained the average of the 10 projections
obtained for each period as a consensus of the nal forecast.
e level of uncertainty of predictions, in each hexagon, was
estimated through the standard deviation shown by the 10
projected favourability values.
Measuring the rate of future change in the models
In order to quantify, in global terms, to what extent the
original favourability values (F0) could be modied accord-
ing to the future favourability forecasts (Ff), we calculated
increment and maintenance rates according to the equations
(Romeroetal. 2014):
IcF cF
cF andM cF F
cF
ff








0
0
0
0
,
where c(Fx) is the cardinality of either the initial model (F0)
or of the future projection (Ff); that is, the sum of all the
hexagons’ favourability values according to a model (Zadeh
1965). e intersection between future (Ff) and present (F0)
favourability values was calculated as follows:
FFMinFF
ff


00
,.
Positive increment values (I) indicate a global net increase,
or gain of favourability, in the future, with respect to those
of the present; whereas negative values mean a net loss in
favourability. Maintenance values (M) reect to what degree
the level of favourability for the presence of disease will keep
the same status in the future. Finally, local changes in favour-
ability were analysed by mapping, in each hexagon, the
dierence between future forecasts and the original favour-
ability values.
Results
Our model projections show that currently favourable areas
for dengue transmission are likely to maintain this status in the
near (2041–2060) and distant (2061–2080) future (M > 0.98).
In the near future, the global level of favourability for dengue
transmission could increase very slightly (I = 0.004), specically
in Central Africa, the southern limits of the Himalayas, eastern
China, Mediterranean Europe, the Amazon basin, western and
northern Brazil, southern Venezuela and the Guianan shield
(Fig. 1). Nevertheless, forecasts for these areas are subject to
certain levels of uncertainty (SD = 0.04 to 0.06, only reaching
0.1 in the Himalayas; note that favourability values range from
0 to 1). In contrast, in the distant future, an increased global
degree of favourability for dengue transmission of almost 10%
(I = 0.099) is expected. Risk levels could thus increase notably
in the above-mentioned areas, and also in Mexico, southern
USA, India and Southeast Asia, involving countries such as
ailand, Laos, Myanmar, Cambodia, Malaysia and Indonesia
(Fig. 1). Forecasts for the distant future period are remarkably
consistent, SD reaching values over 0.05 only in some areas of
the Amazon basin and in Mediterranean Europe.
Predictions for yellow fever show a dierent picture com-
pared to that of dengue. Currently favourable areas for trans-
mission will probably keep this status (M > 0.95), but the
increase in global favourability is predicted to be faster. In the
near future, this increase could be > 5% (I = 0.051), aecting
the Amazon basin (Peru, Colombia, Venezuela, and western
and southern Brazil), the eastern areas of Central Africa, scat-
tered regions in southern and south-eastern Africa, and India
(Fig. 2). Uncertainty values ranging between SD = 0.08 and
0.2 are seen in the forecasts for the Amazon basin and for
Central Africa, but these are negligible in south-eastern Africa
and in India. In the distant period, the global increase with
respect to the present could reach 7% (I = 0.071), aecting the
above-mentioned areas together with northern Central Africa
(Fig. 2). Uncertainty in the forecasts for this period rarely goes
beyond SD = 0.12, but it increases in south-eastern Africa and
in India compared to that for the near future forecasts.
Increasing and decreasing future distribution of
primate chorotypes
Only three South American (SA1, SA6 and SA12) and one
Asian (AS19) primate chorotypes are expected to increase
their distribution area in the near and distant periods, com-
pared to the present (Supporting information). e non-
human primate genera forming part of these chorotypes are:
Alouatta, Aotus, Ateles, Saguinus, Saimiri, Callimico, Cebuella,
Cheracebus, Sapajus, Cebus, Leontocebus, Plecturocebus and
Mico in South America, and Carlito in Asia. Nevertheless, all
these primate chorotypes are likely to maintain their status in
both future periods (M > 0.8), except SA8 in the near future
(M = 0.424) and AF2 in the distant future (M = 0.436)
(Supporting information).
Discussion
is study is the rst to analyse possible changes in the geo-
graphical distribution of dengue and yellow fever transmission

Page 5 of 9
Disease Ecology Special Issue
Disease Ecology Special Issue
Figure1. Dengue transmission model projections into the future for the periods 2041–2060 and 2061–2080. Transmission model for the
period 2001–2017, average model projections into the future for the periods 2041–2060 and 2061–2080, areas where favourability
increases and decreases in the future relative to the present. Dierence between the future projection and the current model. I: increment
rate; M: maintenance rate. Positive values of I indicate a net increase in favourability, that is, a gain in favourable areas, whereas negative
values of I mean a net loss of favourable areas. M indicates the degree to which the favourable areas in the current model overlap with the
favourable forecasted areas. Uncertainty of the vector model in the period 2041–2060 and 2061–2080. SD: standard deviation.
Figure2. Yellow fever transmission model projections into the future for the periods 2041–2060 and 2061–2080. Transmission model for
the period 2001–2017, average model projections into the future for the periods 2041–2060 and 2061–2080, areas where favourability
increases and decreases in the future relative to the present. Dierence between the future projection and the current model. I: increment
rate; M: maintenance rate. Positive values of I indicate a net increase in favourability (i.e. a gain in favourable areas), whereas negative values
of I mean a net loss of favourable areas. M indicates the degree to which the favourable areas in the current model overlap with the favour-
able forecasted areas. Uncertainty of the vector model in the period 2041–2060 and 2061–2080. SD: standard deviation.

Page 6 of 9
Disease Ecology Special Issue
Disease Ecology Special Issue
areas taking into account both urban and sylvatic vectors, as
well as the biogeography of non-human primate hosts. Our
model projections detect that, for both diseases, the risk of
transmission could expand to several regions of the Amazon
basin, Central Africa, Asia and Europe.
Where transmission risk could increase
Despite the knowledge gained from a macro-scale analysis
(our spatial resolution was 7700 km²), there are inherent
limitations in its ability to capture the specic factors that
may inuence virus distribution. Micro-scale environmen-
tal, socio-economic and political factors can play a critical
role in disease dynamics through their inuence on local vec-
tors and hosts. It is therefore essential to recognise that the
ndings presented here provide an overview, but may not
comprehensively address the complexities at the local level.
In many regions worldwide, climate change could increase
the risk of dengue and yellow fever transmission in the
future. e consequences of such a forecast have been dis-
cussed in several papers (Dhimaletal. 2015, Messinaetal.
2019, Gaythorpeetal. 2020, Mordecaietal. 2020, Colón-
González et al. 2021, Sadeghiehet al. 2021). Our projec-
tions do not show large global changes, but they do show
responses at the regional scale. In the case of dengue, our dis-
tant-future projections show an increase of transmission risk
in regions of Central and West Africa, South Asia, the Gulf
of Mexico, Central America, the Amazon basin and Europe.
ese regions, with the exception of West Africa and South
Asia, coincide with the areas where, according to Colón-
Gonzálezetal. (2021), the length of the transmission season
within a year will probably increase, due to changes in rainfall
and humidity generated by global warming. Furthermore, in
the projections published by Messinaetal. (2019), these same
regions are predicted as areas of possible dengue expansion.
However, only our projections show a possible expansion
of dengue on the European continent. e Mediterranean
basin, and some locations along the coasts of the UK and
Portugal, show an increased risk of dengue transmission and
with low uncertainty in these last two places. Mordecaietal.
(2020) report that, in Africa, the eects of global warming are
likely to promote greater environmental suitability for arbo-
viruses transmitted by Ae. aegypti (for example, dengue), and
to reduce suitability for pathogens transmitted by Anopheles
gambiae (for example, malaria). Other studies focused in
Asia, specically in China, mention that there has already
been an expanding trend for dengue infections from south
to north, in line with warming temperatures (Yietal. 2019).
Our models predict with low uncertainty that the south will
remain favourable and that there will be an increase in risk
towards the north of China (Fig. 1). In Nepal, Dhimaletal.
(2015) conclude climate change could intensify the risk of
dengue epidemics in the mountain regions of the country.
is is consistent with our results, which predict an increase
not only in Nepal, but also in the whole Himalayan moun-
tain range, comprising several countries such as India and
Bhutan (Fig. 1).
In the case of yellow fever, Gaythorpeetal. (2020) sug-
gest that transmission may change heterogeneously across
Africa. Our projections predict a favourability increase in the
near future, with high uncertainty, in many areas to the east
of the continent. In the distant future, favourability could
increase, with low uncertainty, also in the northern and west-
ern regions of Central Africa. is is in agreement with the
results of Gaythorpeet al. (2020), according to which the
Central African Republic is one of the countries most likely
to see an increase in transmission risk. In Asia, although yel-
low fever infections do not occur, our model projections
point to a risk increase in India, in both the near and the
distant future periods (Fig. 2). Bicca-Marquesetal. (2022)
predicted that southern India and parts of Southeast Asia
could be considered to be suitable for the presence of yellow
fever virus, which could potentially threaten primate species.
e global strategy to EYE 2017–2026 [7] warns that larger
outbreaks will take place in Asia, including the possibility
of outbreaks in countries such as India and China, which
harbour Aedes mosquitoes and are home to 2 billion people
who lack immunity to yellow fever. If the virus were ever
introduced, there could be a risk of major urban outbreaks
because of the high population density of non-immune
humans and of Aedes mosquito species. For America, our
model projections predict an increase in dierent regions
of South America, including Brazil, Paraguay, Bolivia, Peru,
Colombia and Venezuela (Fig. 2). is result contrasts with
that of Sadeghiehetal. (2021), who predict that the intensity
of outbreaks may reduce in Brazil as temperatures increase.
According to our forecasts for the distant future, favour-
ability for yellow fever transmission could increase, with low
uncertainty, in south and eastern Brazil, following the trend
observed in the last decade (Mir etal. 2017). During the
period 2016–2019, Brazil faced one of the largest outbreaks
of yellow fever in recent decades. In the city of São Paulo, the
virus was detected in Alouatta in an area of Mata Atlantica
and Callithrix mainly in urban areas of the city (Cunhaetal.
2020). Our projections suggest that the favourability of non-
human primate distributions could increase in the future,
aecting these primate genera (Supporting information).
e areas where we predict an increased risk of transmission
coincide with those where mosquito favourability also increases
(Aliaga-Samanezetal. 2023), often with that for sylvatic mos-
quitoes. For example, in the case of dengue, in Europe these
areas of increased risk coincide with increased favourable con-
ditions for Ae. aegypti in Portugal, Spain and Italy (although
this species has not yet established populations in the conti-
nent). Aedes aegypti could probably survive all year round and
re-establish itself in these regions under climate change sce-
narios (Krameretal. 2020). In countries such as those of the
UK, the increase should be attributable to the environment,
as no signicant changes regarding vectors is predicted in this
area by Aliaga-Samanezet al. (2023). In the case of dengue
in Asia, the increase in areas such as the southern Himalayas
coincides with an increase in future favourable conditions for
transmitting mosquitoes such as Ae. aegypti, Ae. vittatus and Ae.
niveus (Aliaga-Samanezetal. 2023). Aedes aegypti mosquitoes

Page 7 of 9
Disease Ecology Special Issue
Disease Ecology Special Issue
sampled in Nepal were able to survive the low temperatures
for a short period (Krameretal. 2020). With rising tempera-
tures due to climate change, A. aegypti will be able to spread to
higher altitudes (Dhimaletal. 2021). In contrast, as in the UK,
the projected increase in dengue transmission risk in northern
China should be attributable to the environment.
e areas of increased favourability for future transmission
of both dengue and yellow fever are likely to coincide with
increased favourable conditions for Ae. vittatus, Ae. luteoceph-
alus and Ae. africanus in the northern and western regions of
Central Africa such as Cameroon, Central African Republic
and northern Democratic Republic of Congo (Aliaga-
Samanezet al. 2023). ese predictions are of concern, as
sylvatic cycles of dengue persist in Africa and continue to spill
over to humans (Cardosaet al. 2009, Dieng et al. 2023).
In South America, the most favourable areas for future yel-
low fever transmission coincide with a more suitable pres-
ence of S. chloropterus in Peru, Brazil and Bolivia, and also
of Ae. albopictus in the Amazon basin. In countries such as
Colombia and Paraguay, the increase could be due to the
environment instead (Aliaga-Samanezetal. 2023).
Conclusion
According to our results, surveillance strategies could be
applied taking into account two scenarios. On the one side,
vector surveillance eorts should be prioritised in those areas
where the increased dengue or yellow fever transmission risk
is seen to coincide with forecasted changes in the distribu-
tion of vectors. e early detection of arrival of new mosquito
species could give place to further measures in prevention.
Once the vector is present in the region, if risk is predicted
to be higher following environmental changes, we suggest
an exhaustive surveillance for the appearance of outbreaks in
new areas, and the launching of deep information campaigns
focused on people travelling to and from endemic areas. is
is really challenging, as each country has its own surveillance
systems; however, integration of surveillance systems has been
shown to improve surveillance performance (Georgeet al.
2020). Without a globally standardised system, it is dicult
to prevent new outbreaks. According to our study, dengue
could spread to the European continent, specically to the
Mediterranean basin, the UK and Portugal; and, in Asia, to
northern China. In the case of yellow fever, climatic favour-
ability could increase in central and southeast Africa, India,
South America (Brazil, Paraguay, Bolivia, Peru, Colombia
and Venezuela). Specically, in Brazil, yellow fever favour-
ability could increase in the south, west and east. In order
to get detailed forecasts, we suggest that ne-scale pathogeo-
graphic analyses be made in the regions of concern.
Funding – is study was supported by Project PID2021-
124063OB-I00 of the Spanish Ministry of Science and Innovation
and European Regional Development Fund (ERDF). DR is
supported by the incorporation Doctor program of the University
of Malaga, UMA-2022/REGSED-64576. AA-S was supported
by a postdoctoral contract of the Plan Propio de Investigación,
Transferencia y Divulgación Cientíca of the University of Malaga.
Author contributions
Alisa Aliaga-Samanez: Conceptualization (lead), Data cura-
tion (lead), Formal analysis (lead), Funding acquisition (lead),
Investigation (lead), Methodology (lead), Visualization (lead),
Writing – original draft (lead), Writing – review and editing
(lead). David Romero: Investigation (supporting), Supervision
(supporting), Writing – original draft (supporting). Kris
Murray: Investigation (supporting), Supervision (supporting).
Marina Cobos-Mayo: Investigation (supporting), Writing
– original draft (supporting). Marina Segura: Investigation
(supporting), Supervision (supporting). Raimundo Real:
Conceptualization (supporting), Investigation (supporting),
Supervision (supporting). Jesús Olivero: Conceptualization
(lead), Formal analysis (lead), Funding acquisition (lead),
Investigation (lead), Methodology (lead), Supervision (lead),
Validation (lead), Writing – original draft (lead).
Transparent peer review
e peer review history for this article is available at https://
www.webofscience.com/api/gateway/wos/peer-review/
ecog.06942.
Data availability statement
Data are available from the Figshare: https://doi.org/10.6084/
m9.gshare.25897024.v1 (Aliaga Samanezetal. 2024).
Supporting information
e Supporting information associated with this article is
available with the online version.
References
Acevedo, P. and Real, R. 2012. Favourability: concept, distinctive
characteristics and potential usefulness. – Naturwissenschaften
99: 515–522.
Aliaga-Samanez, A., Cobos-Mayo, M., Real, R., Segura, M.,
Romero, D., Fa, J. E. and Olivero, J. 2021. Worldwide dynamic
biogeography of zoonotic and anthroponotic dengue. – PLoS
Negl. Trop. Dis. 15: e0009496.
Aliaga-Samanez, A., Real, R., Segura, M., Marl-Daza, C. and Oli-
vero, J. 2022. Yellow fever surveillance suggests zoonotic and
anthroponotic emergent potential. – Commun. Biol. 5: 530.
Aliaga-Samanez, A., Romero, D., Murray, K., Segura, M., Real, R.
and Olivero, J. 2023. Potential climate change eects on the
distribution of urban and sylvatic dengue and yellow fever vec-
tors. – bioRxiv 2023.11.06.565841.
Aliaga-Samanez, A., Romero, D., Murray, K., Cobos-Mayo, M.,
Segura, M., Real, R. and Olivero, J. 2024. Presence and absence
of records of dengue cases, yellow fever cases, and vector occur-
rences on a worldwide 7,774-km² hexagonal grid. – Figshare
Dataset, https://doi.org/10.6084/m9.gshare.25897024.v1.

Page 8 of 9
Disease Ecology Special Issue
Disease Ecology Special Issue
Bicca-Marques, J. C., Rabelo, R. M., de Almeida, M. A. B. and
Sales, L. P. 2022. e risks of yellow fever to Asian primates.
– Int. J. Primatol. 43: 74–91.
Cardosa, J., Ooi, M. H., Tio, P. H., Perera, D., Holmes, E. C.,
Bibi, K. and Abdul Manap, Z. 2009. Dengue virus Serotype 2
from a sylvatic lineage isolated from a patient with dengue hem-
orrhagic fever. – PLoS Negl. Trop. Dis. 3: e423.
CDC 2018. Global health: yellow fever. – CDC, https://www.cdc.
gov/globalhealth/newsroom/topics/y.
Colón-González, F. J., Sewe, M. O., Tompkins, A. M., Sjödin, H.,
Casallas, A., Rocklöv, J., Caminade, C. and Lowe, R. 2021.
Projecting the risk of mosquito-borne diseases in a warmer and
more populated world: a multi-model, multi-scenario intercom-
parison modelling study. – Lancet Planet. Health 5: e404–e414.
Cunha, M. S. etal. 2020. Possible non-sylvatic transmission of
yellow fever between non-human primates in São Paulo city,
Brazil, 2017–2018. – Sci. Rep. 10: 1–81.
Daszak, P., Cunningham, A. A. and Hyatt, A. D. 2000. Emerging
infectious diseases of wildlife – threats to biodiversity and
human health. – Science 287: 443–449.
Dhimal, M., Ahrens, B. and Kuch, U. 2015. Climate change and
spatiotemporal distributions of vector-borne diseases in Nepal
– a systematic synthesis of literature. – PLoS One 10: e0129869.
Dhimal, M., Kramer, I. M., Phuyal, P., Budhathoki, S. S., Hartke,
J., Ahrens, B., Kuch, U., Groneberg, D. A., Nepal, S., Liu, Q.
Y., Huang, C. R., CissÉ, G., Ebi, K. L., Klingelhöfer, D. and
Müller, R. 2021. Climate change and its association with the
expansion of vectors and vector-borne diseases in the Hindu
Kush Himalayan region: a systematic synthesis of the literature.
– Adv. Clim. Change Res. 12: 421–429.
Dieng, I.etal. 2023. Re-emergence of Sylvatic dengue 2 during an
outbreak in southeastern Senegal, Kedougou 2020. – medRxiv:
11.22.23298824.
Estrada, A., Real, R. and Vargas, J. M. 2008. Using crisp and fuzzy
modelling to identify favourability hotspots useful to perform
gap analysis. – Biodivers. Conserv. 17: 857–871.
Fa, J. E., Olivero, J., Farfán, M. Á., Márquez, A. L., Vargas, J. M.,
Real, R. and Nasi, R. 2014. Integrating sustainable hunting in
biodiversity protection in central Africa: hot spots, weak spots,
and strong spots. – PLoS One 9: e112367.
Franklinos, L. H. V., Jones, K. E., Redding, D. W. and Abubakar,
I. 2019. e eect of global change on mosquito-borne disease.
– Lancet Infect. Dis. 19: e302–e312.
Gaythorpe, K. A. M., Hamlet, A., Cibrelus, L., Garske, T. and
Ferguson, N. M. 2020. e eect of climate change on yellow
fever disease burden in Africa. – eLife 9: 1–27.
George, J., Häsler, B., Mremi, I., Sindato, C., Mboera, L., Rwey-
emamu, M. and Mlangwa, J. 2020. A systematic review on
integration mechanisms in human and animal health surveil-
lance systems with a view to addressing global health security
threats. – One Heal. Outlook 2: 1–15.
Hamlet, A., Jean, K., Perea, W., Yactayo, S., Biey, J., Van Kerkhove,
M., Ferguson, N. and Garske, T. 2018. e seasonal inuence
of climate and environment on yellow fever transmission across
Africa. – PLoS Negl. Trop. Dis. 12: e0006284.
IUCN 2021. IUCN Red List of threatened species. Ver. 2021–2.
– www.iucnredlist.org/.
Iwamura, T., Guzman-Holst, A. and Murray, K. A. 2020. Acceler-
ating invasion potential of disease vector Aedes aegypti under
climate change. – Nat. Commun. 11: 1–10.
Ka-Wai Hui, E. 2006. Reasons for the increase in emerging and re-
emerging viral infectious diseases. – Microbes Infect. 8: 905–916.
Karesh, W. B., Dobson, A., Lloyd-Smith, J. O., Lubroth, J., Dixon,
M. A., Bennett, M., Aldrich, S., Harrington, T., Formenty, P.,
Loh, E. H., Machalaba, C. C., omas, M. J. and Heymann,
D. L. 2012. Ecology of zoonoses: natural and unnatural histo-
ries. – Lancet 380: 1936–1945.
Karger, D. N., Conrad, O., Böhner, J., Kawohl, T., Kreft, H., Soria-
Auza, R. W., Zimmermann, N. E., Linder, H. P. and Kessler,
M. 2017. Climatologies at high resolution for the earth’s land
surface areas. – Sci. Data 4: 170122.
Karger, D. N., Schmatz, D. R., Dettling, G. and Zimmermann, N.
E. 2020. High-resolution monthly precipitation and tempera-
ture time series from 2006 to 2100. – Sci. Data 7: 248.
Karger, D. N., Wilson, A. M., Mahony, C., Zimmermann, N. E.
and Jetz, W. 2021. Global daily 1 km land surface precipitation
based on cloud cover-informed downscaling. – Sci. Data 8: 307.
Knutti, R., Masson, D. and Gettelman, A. 2013. Climate model
genealogy: generation CMIP5 and how we got there. – Geo-
phys. Res. Lett. 40: 1194–1199.
Kramer, I. M., Kreß, A., Klingelhöfer, D., Scherer, C., Phuyal, P.,
Kuch, U., Ahrens, B., Groneberg, D. A., Dhimal, M. and Mül-
ler, R. 2020. Does winter cold really limit the dengue vector
Aedes aegypti in Europe? – Parasit. Vectors 13: 178.
Kuno, G. 2015. A re-examination of the history of etiologic confu-
sion between dengue and chikungunya. – PLoS Negl. Trop.
Dis. 9: 1–11.
McSweeney, C. F., Jones, R. G., Lee, R. W. and Rowell, D. P. 2015.
Selecting CMIP5 GCMs for downscaling over multiple regions.
– Clim. Dyn. 44: 3237–3260.
Messina, J. P.etal. 2019. e current and future global distribution
and population at risk of dengue. – Nat. Microbiol. 4:
1508–1515.
Mir, D., Delatorre, E., Bonaldo, M., Lourenço-de-Oliveira, R.,
Vicente, A. C. and Bello, G. 2017. Phylodynamics of yellow
fever virus in the Americas: new insights into the origin of the
2017 Brazilian outbreak. – Sci. Rep. 7: 7385.
Mordecai, E. A., Ryan, S. J., Caldwell, J. M., Shah, M. M. and
LaBeaud, A. D. 2020. Climate change could shift disease bur-
den from malaria to arboviruses in Africa. – Lancet Planet.
Health 4: e416–e423.
Olivero, J., Real, R. and Márquez, A. L. 2011. Fuzzy chorotypes as
a conceptual tool to improve insight into biogeographic pat-
terns. – Syst. Biol. 60: 645–660.
Olivero, J., Fa, J. E., Real, R., Farfán, M. Á., Márquez, A. L., Var-
gas, J. M., Gonzalez, J. P., Cunningham, A. A. and Nasi, R.
2017. Mammalian biogeography and the Ebola virus in Africa.
– Mamm. Rev. 47: 24–37.
Real, R., Barbosa, A. M. and Vargas, J. M. 2006. Obtaining envi-
ronmental favourability functions from logistic regression. –
Environ. Ecol. Stat. 13: 237–245.
Rohr, J. R., Barrett, C. B., Civitello, D. J., Craft, M. E., Delius,
B., DeLeo, G. A., Hudson, P. J., Jouanard, N., Nguyen, K.
H., Ostfeld, R. S., Remais, J. V., Riveau, G., Sokolow, S. H.
and Tilman, D. 2019. Emerging human infectious diseases
and the links to global food production. – Nat. Sustain. 2:
445–456.
Romero, D., Olivero, J., Márquez, A. L., Báez, J. C. and Real, R.
2014. Uncertainty in distribution forecasts caused by taxonomic
ambiguity under climate change scenarios: a case study with two
newt species in mainland Spain. – J. Biogeogr. 41: 111–121.
Romero, D., Olivero, J., Brito, J. C. and Real, R. 2016. Compari-
son of approaches to combine species distribution models based
on dierent sets of predictors. – Ecography 39: 561–571.

Page 9 of 9
Disease Ecology Special Issue
Disease Ecology Special Issue
Sadeghieh, T., Sargeant, J. M., Greer, A. L., Berke, O., Dueymes,
G., Gachon, P., Ogden, N. H. and Ng, V. 2021. Yellow fever
virus outbreak in Brazil under current and future climate. –
Infect. Dis. Modell. 6: 664–677.
Sanderson, B. M., Knutti, R. and Caldwell, P. 2015. A representa-
tive democracy to reduce interdependency in a multimodel
ensemble. – J. Clim. 28: 5171–5194.
Shearer, F. M., Longbottom, J., Browne, A. J., Pigott, D. M.,
Brady, O. J., Kraemer, M. U. G., Marinho, F., Yactayo, S.,
de Araújo, V. E. M., da Nóbrega, A. A., Fullman, N., Ray,
S. E., Mosser, J. F., Stanaway, J. D., Lim, S. S., Reiner, R.
C., Moyes, C. L., Hay, S. I. and Golding, N. 2018.
Existing and potential infection risk zones of yellow fever
worldwide: a modelling analysis. – Lancet Glob. Health 6:
e270–e278.
Watts, N.etal. 2019. e 2019 report of the Lancet Countdown
on health and climate change: ensuring that the health of a child
born today is not dened by a changing climate. – Lancet 394:
1836–1878.
WHO 2021a. Dengue and severe dengue. – WHO, http://www.
who.int/mediacentre/factsheets/fs117/en.
WHO 2021b. Eliminate yellow fever epidemics (EYE) strategy
2017–2026. – WHO, https://www.who.int/publications/i/
item/9789241513.
Yi, L., Xu, X., Ge, W., Xue, H., Li, J., Li, D., Wang, C., Wu, H.,
Liu, X., Zheng, D., Chen, Z., Liu, Q., Bi, P. and Li, J. 2019.
e impact of climate variability on infectious disease transmis-
sion in China: current knowledge and further directions. –
Environ. Res. 173: 255–261.
Zadeh, L. A. 1965. Fuzzy sets. – Inf. Control 8: 338–353.