0 500 1000 1500 2000 2500 3000 3500 4000 1 8 1522 2936 43 50 57 6471 7885 9299 106 113 120 127 134 Days M (t ) 10/0 8/2002 24/0 8/2002 07/0 9/2002 21/0 9/2002 05/1 0/2002 19/1 0/2002 02 /11/2002 16/11/2002 30 /11/2002 14 /12/2002 28/1 2/2002 11 /01/20 03 25/0 1/2003 S1S6S 11S 16S 21S 26S31S36S 41S46S 51S56S 61S 66S71S76S81S 86S91S 96 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 15 /06/2002 29 /06/2002 13 /07/2002 27/0 7/2002 10/0 8/2002 24 /08/2002 07 /09/2002 21 /09/2002 05/1 0/2002 19/1 0/2002 02/1 1/2002 16 /11/2002 30 /11/2002 14 /12/2002 28 /12/2002 S1 S9 S17 S25 S33 S41 S49 S57 S65 S73 S81 S89 S97 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 20000 40000 60000 80000 100000 120000 140000 16111621263136414651566166717681869196101 106 111 116121126131136141 Days Po n d a re a (m 2)
Contribution of a hydrologic pond model to predict spatial and
temporal mosquito population dynamics in Northern Senegal
Soti V.
a,b
, Tran A.
b,c
, Mondet B.
d
, Fontenille D.
d
, Chevalier V.
b,
Lancelot R.
b
,
Thiongane Y.
e
, Degenne P.
b,c
, Lo Seen D.
b,c
, Guegan J-F.
d
a - SAS Nevantropic, 16 bis av. du 14 juillet, 97300 Cayenne, French Guiana.
b - CIRAD (French Agricultural Research Center for International Development), Baillarguet Campus, 34398 Montpellier Cedex.
c - UMR TETIS
(
Territories, Environment, Remote Sensing and Spatial Information Joint Research Unit), Maison de la Télédétection, 500 rue J.-F. Breton, 34093 Montpellier Cedex 5, France.
d - IRD (Institut de Recherche pour le Développement), 34394 Montpellier Cedex 05, France.
e - ISRA/LNERV Route du front de terre, BP 2057, Dakar-Hann, Sénégal.
Summary
We developed a spatial and temporal model to predict the population
dynamics of the two main mosquito vectors (Aedes vexans and Culex
poicilipes) involved in the Rift Valley fever virus transmission.
Covering an area of 11x10 km around the village of Barkedji, and
located in the Ferlo Valley (Northern Senegal), the study area is
characterized by a complex and dense network of ponds that are filled
during the rainy season (from July to mid-October). These ponds are
the main mosquito breeding sites in the area. The vector population
dynamics model combines a spatial and temporal hydrological model
with a mosquito population model. The hydrological model uses daily
rainfall as main input to predict spatial and temporal changes in pond
surfaces. Output is then fed into the mosquito population model to
predict vector population dynamic. This approach will allow predicting
how mosquito abundance varies in space and in time.
References
General model description
Ae. vexans population model
Cx. poicilipes population model
Input
Mosquito population
abundance
model
Daily water-rings variations (surface: m2 )were simulated from a hydrologic pond model [soti et al, submitted paper] and calibrated and validated with field data (rainfall and water level data recorded during the rainy seasons 2002 and 2003).
Results
Conclusion and Perspectives
)
(
)
,
(
)
(
)
(
t
k
T
t
T
M
t
T
M
dt
dM
=
−
α
+
σ
χ
−
σ
(T)=vectorial production capacity
x(t,T
)
= breeding habitat capacity
2 ) ( ) ( ) , ( m S T t S T t t S T t = − − χ ) ( ) ( S T L T mφ
λ
β
σ
=The Rift Valley fever in Senegal
DATA
During the rainy season in 2002 and 2003, mosquitoes have been collected
every 20 days using sheep-baited traps (Photo1: Sheep-baited trap), in
Barkedji area (Ferlo, 14.87 W, 15.28 N). Three temporary ponds, Niaka,
Barkedji and Furdu, have been chosen according to ecological and
structural criteria (Fig 1). A trap was laid a few meters from the pond bank.
Trapping sessions were carried out between 6 pm and 6 am on three
consecutive days to account for daily fluctuations in mosquitoes abundance.
Figure 2 shows the total numbers of trapped mosquitoes. For this study, we
have used only the trap closed to the pond to measure the capacity of each
pond to produce mosquito.
Study area
Covering an area of 11x10 km around the village of Barkedji, the study area
is characterized by a complex and dense network of ponds filling during the
rainy season (from July to mid-October). Located in the Ferlo region, the
climate is Sahelian with 3 seasons:
- a dry, cold season from November to March,
- a dry, hot season from April to June,
- a rainy season from June to November with annual rainfall ranging
from 100 to 500 mm.
Ae. vexans 86,4 % 6,4 % Ae. ochraceus 7,2 % Cx. poicilipes2002
77,9 % 1,8 % Ae. ochraceus 20,2 % Cx. poicilipes Ae. vexans3093 female mosquitoes captured, including 2344 RVF vectors.
Fig 2 : female mosquitoes captured
2003
6290 female mosquitoes captured, including 4855 RVF vectors.
Rift Valley Fever (RVF) is an arbovirosis caused by a Phlebovirus (Bunyaviridae) Main RVF vectors in Senegal are Ae. vexans and Cx. poicilipes genera mosquitoes
[Fontenille et al., 1998; Diallo et al., 2000]
Main hosts : ruminants (sheep, goats and cattle)
Temporary ponds are favorable areas for RVF transmission
Ae. vexans species oviposit their eggs on the soil above the water level depressions. Aedes eggs required consecutive drying (7 day minimum) and immersions periods for hatching. Eggs are resistant to dessication for several weeks, months or years.
Cx. poicilipes poicilipes lay eggs directly on water surfaces [Gjullin et al, 1950] . Eggs could not resist to dessication requiring ~ 10 days immersion for hatching.
adult
Larva Egg
Pupa
Fig.4: Mosquito life cycle Ae. vexans breeding site
Cx. poicilipes breeding site
Water
Fig 3: Breeding sites for Aedes and Culex species
Fig 1: Study Area
β: Imago emergence probability/day
λ: number of eggsper female/day
Φ: Pre-imago survival probability Sm : maximum pond surface (m2)
L: Gonotrophic cycle
Sm: maximum pond surface(m2)
S(t-T): oviposition surface (m2)
S(t/t-T): effective emerging surface (m2) T: developement period
EPIDEMICS
2-4 December 2009
Athens, Greece
Output
Number of Cx. poicilipes
(log.) in space and in time
8-July-2002 22-Sept-2002
1) For the calibration phase, the parameters values have been chosen from the literature.
2) The results have been validated for the year 2002 and 2003 with field data.
Model validation for 2002 and 2003
We have validated the model with the field data (Cx. poicilipes and Ae. vexans) collected during the rainy season in 2002 and 2003:
Daily number of female mosquitoes (Cx. poicilipes & Ae. vexans) for each pond represented by a different color.
0 50 100 150 0 .0 0 .2 0 .4 0 .6 0 .8 1 .0 0 50 100 150 0 .0 0 .2 0 .4 0 .6 0 .8 1 .0
Simulated mosquito density was heterogeneous in time and in space.
At the beginning of the rainy seasons, mosquito populations were small and well distributed in space.
At the end of the rainy season, the highest density of Cx. poicilipes was located in the main stream of the Ferlo Valley where ponds are largest. Here, we could observe a highest Cx. poicilipes density at the end of the rainy season 2002. [Mondet et al 2005]. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 15/0 6/20 02 15/0 7/20 02 15/0 8/20 02 15/0 9/20 02 15/1 0/20 02 15/1 1/20 02 15/1 2/20 02 15/0 1/20 03 15/0 2/20 03 15/0 3/20 03 15/0 4/20 03 15/0 5/20 03 15/0 6/20 03 15/0 7/20 03 15/0 8/20 03 15/0 9/20 03 15/1 0/20 03 15/1 1/20 03
Simulation results showed two different population dynamics patterns:
- For the rainy season 2002, Ae. vexans were more abundant at the beginning of the rainy season, and Cx. poicilipes were dominant at the end of the rainy season [Mondet et al 2005].
- The rainy season 2003 (moderate and regular rainfall), showed a relatively stable population growth rate throughout the rainy season, with a Cx. poicilipes density peak at the end of the rainy season.
Number of mosquito in time (2002)
Here, we could observe a highest Cx. poicilipes density at the end of the rainy season 2002. [Chevalier et al 2005].
Each pond is represented by a different color
Simulation results were similar to Cx. poicilipes and Ae. vexans observations in time and space. Models were able to simulate the importance of rainfall patterns on
mosquito populations. Long pauses in rainfall were favourable for Ae. vexans and conversely, frequent rainfall are more appropriate for the development of a high Cx.
poicilipes population.
In subsequent steps, these models will be coupled with a mosquito diffusion model to test the importance of different parameters (vegetation density, host compound
location, wind) for mosquito spread in space and in time around their breeding sites.
Results of this work will be useful to help designing better prevention and control programmes for vector-borne diseases.
Cx. poicilipes Ae. vexans
The graphic shows a population dynamics which starts early with the first effective rains. This dynamic is observed throughout the rainy season with a peak of density in the middle season followed by a quick decline [Fontenille et al, 1998, Chevalier et al, 2004] .
Diallo, M., Lochouarn, L., Ba, K., Sall, A. A., Mondo, M., Girault, L., Mathiot, C.- 2000: First isolation of the Rift Valley fever virus
from Culex poicilipes (Diptera : Culicidae) in nature, American Journal of Tropical Medicine and Hygiene, 62, 702-704, 2000.
Gjullin C.M., Yates W.W., Stage H.H., 1950 - Studies on Aedes Vexanx and Aedes Sticticus, flood-water mosquitoes, i the lower
columbia river valley – Annals entomological Society of America, Vol43, 262-276.
Chevalier V, Mondet B.,Diaitte A., Lancelot R., Fall A.G., Ponçon N, 2004 – Exposure of sheep to mosquito bites: possible
consequences for the transmission risk of Rift Valley Fever in Senegal.- Medical and Veterinary Entomology (18) , 247–255
Chevalier V., Renaud L., Yaya T., Sall B., Diaité A Mondet B. , 2005 – Rift Valley Fever in Small Ruminant, Senegal
2003-Emerging Infectious Diseases • Vol. 11, No. 11.
Fontenille D, Traore-Lanmizana M., Diallo M., Thonnon J., Digoutte JP., HG Zeller HG., 1998 – New Vectors of Rift Valley
Fever i West Africa- Emerging Infectious Diseases, Vol.4, n°2.
Mondet, B., Diaïté, A., Ndione, J., Fall, A., Chevalier, V., Lancelot, R., Ndiaye, M. & Ponçon, N. 2005. Rainfall patterns and population dynamics of Aedes (Aedimorphus) vexans arabiensis, Patton 1905 (Diptera: Culicidae), a potential vector of Rift Valley fever virus in Senegal. Journal of Vector Ecology, 30: 102-106.
Porphyre T., Bicout D.J., Sabatier P., 2005 - Modelling the abundance of mosquito vectors versus flooding dynamics – Ecological
modelling, 183, 173–181.
Soti V., Puech C., Lo Seen D., A Bertran, Dessaye N., Tran A. - Remote sensing and spatial modelling for the monitoring of the
dynamics of temporary ponds in arid areas- Hydrology and Earth System Sciences (submitted).
)
(
).
(
.
.
)
(
t
k
T
1E
t
T
1M
dt
dM
=
−
α
+
β
Φ
−
M(t): nbr of adult females mosquitoes at time step t a: mortality rate
k: sex-ratio
b: Imago emergence probability
Φ1: Pre-imago survival probability
T1: Development period
E(t): number of hatching eggs at time step t
> − − − ∆ − + − − − − ∆ − =
∑
∑
= = otherwise T t j S T t S if T t Eggs r T t S T t j S T t S r t E i j R p e i j R p 0 0 ) , ( ) ( ) ( . ) ( ) , ( ) ( ) 1 ( ) ( 1 2 2 2 2 1 2 2 r: Leach rate T2: Desiccation periodSp(t): Water surface of the pond at time step t
∆Sp(t): Pond surface variations between t and t-1
− − > ∆ > − − ∆ ∆ − − = > ∆ − > ∆ =
∑
∑
− = − = otherwise t i S t S if otherwise t j S t S if t S t j S t i S otherwise t S t S if t S R p i j R p i j p R R p p R ) 1 , 1 ( 0 ) ( 0 0 ) 1 , ( ) ( ) ( ) 1 , ( ) , 1 ( ) ( 0 ) ( 0 ) , 1 ( 1 1 1 1SR(i,t): Dry ring surface with eggs in desiccation phase
since i days (1≤i≤T2) at time step t
Se(t): Dry surface of the pond with eggs in desiccation phase and ready to hatch, at time step t Eggs(t): Number of eggs in desiccation phase at time step t
∑
= − − = 2 1 ) , ( ) ( ) ( T i R et Sm Spt S it S ) ( . ) ( ) ( ) ( ) ( ). ( 2 2 2 2 Eggst T t S T t S t M L t Eggs T dt dEggs e p − − ∆ − + Φ =λ
Sm: Maximum pond surface
Φ2: Egg survival probability during the
desiccation phase
λ: Number of eggs per female/day
L: Duration of the gonotrophic cycle (days)
Barkedji surface pond area during the rainy season 2002
Number of Cx. poicilipes, Furdu 2003
Number of Ae. vexans, Furdu 2003
Day Day A b u n d a n c e (r e la ti ve v a lu e ) A b u n d a n c e (r e la ti ve v a lu e ) Pond s clas sified by size Ponds cla ssifi ed b y size A b u n d a n e in r e la ti ve v a lu e A b u n d a n e in r e la ti ve v a lu e A b u n d a n e in r e la ti ve v a lu e 5 km Pond s Pond s Hatch
Yellow lines correspond to limit of pond area at different times
Photo1: Sheep baiting trap
@V. Chevalier
In red, number of mosquito In blue, pond area
In black, number of mosquito captured on the field (relative value)
Date
Day
Day M(t): nbr of adult
females at time step t k: sex ratio a: mortality rate
Adapted from [Porphyre, 2005]
Ae. vexans Cx. poicilipes
Barkedji pond 2002 - 2003 Elevation (m)
Value High : 101.565
Low : 98.5421
- peaks of mosquito density were well simulated for both Cx. poicilipes and Ae. vexans.
- for Aedes spp. , the peak following the first rainfall was also well simulated.
- for Cx. poicilipes, an important peak of simulated density was observed at the end of the rainy season, as described in the literature.