0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0 1 2 3 4 5 6 10 Km
Populations of ~ 30 ind. in the same locality Sparsed individuals
Sampling of individuals and populations among a global collection of isolates
Australia N. Caledonia Fidji Malaysia Philippines Vanuatu Tonga Samoa Cook Islands Futuna PNG Indonesia Viet Nam China Panama Costa Rica Jamaica Mexico Ecuador Florida Cuba Honduras Colombia Venezuela Dom. Rep. C.Ivoire Comoros Cameroun Nigeria Uganda Mayotte Gahna Congo Tanzania Rwanda Burundi
AREAS of INTRODUCTION AREA of ORIGIN
1978 Gabon 1973 ? Zambie 1972 Honduras 1963 Fidji
- Infected plant material - Ascospores (sexual cycle) - Conidia (asexual cycle)
- Existence of sexual reproduction - Stable and panmictic populations
- Historical data about global spread available
A convenient biological model for population genetics and phylogeographic studies:
Dispersal modes of M. fijiensis:
Life cycle of Mycosphaerella fijiensis From Agrios, George N. 2005.
COM UGA CAM CAM CIV GAB NGA COL CR HND HND MEX PAN VEN JAM RD
IND MYS PHL PNG FIJ NCL AUS Fst = 0.57 Fst = 0.40 Fst = 0.20 He = 0.35 He = 0.22 He = 0.65 He = 0.32 Fst = 0.21
S. ROBERT*, A. RIEUX*, F. HALKETT†, M-F. ZAPATER*, L. DE LAPEYRE DE BELLAIRE‡, C. ABADIE*, V. RAVIGNE* and J. CARLIER*
Explorative population genetics analysis
Molecular phylogenetic analysis
23 populations (~700 individuals) 21 microsatellite markers
(Zapater
et al
, 2008 Mol. Ecol. Res.; Robertet al
, 2010 Am. J. Bot.)120
M. fijiensis
individuals8 sequence-based nuclear markers
Phylogenetic tree reconstructed by the maximum likelihood method
How plant pathogenic fungi spread is the first question to consider for understanding the emergence of diseases caused by such organisms.
The ascomycete fungus
Mycosphaerella fijiensis,
causing the black leaf streak disease of banana, is an exampleof a recent pandemic in agriculture and a good model to address this question in the case of an aerial plant
pathogen. The pandemic started around 1960 from the South-East Asia and
M. fijiensis
has spread to almost allbanana producing areas for the last 40 years.
Perspective: test different scenarii and estimate parameters using ABC (Approximate Bayesian Computation) method
CONTEXT & OBJECTIVES
PHYLOGEOGRAPHY STUDY AT GLOBAL SCALE
SPATIAL GENETIC ANALYSIS AT LOCAL SCALE
Dispersal processes underlying the recent pandemic caused by the plant
pathogenic fungus
Mycosphaerella fijiensis
Previous studies of genetic structure of populations have highlighted a higher genetic diversity in South-East Asia, and high levels of genetic differentiation between and within continents, suggesting
that founder events have accompanied the pandemic (Carlier
et al
, 1996; Rivaset al
, 2004 Mol. Ecol.).A further investigation of the dispersal processes underlying the pandemic is now required.
- A pattern of genetic admixture in AMERICA, possibly due to multiple introductions - An only successful introduction in AFRICA, probably associated with a bottleneck, explaining emergence of an homogeneous group comparatively to the other continents
These different analysis were congruent and suggested strong hypotheses:
- Asian origins for all these introduction events certainly from transport of infected plant materials
Low level of intraspecific polymorphism
Markers of old events of introductions High level of polymorphism Markers of fine-scale dispersal processes
Objectives:
To infer historical and demographic events (bottlenecks, admixture) undergone by invading populations
To reconstruct global routes of dissemination and clarify dispersal modes of the pathogen
Diversity and differentiation indices, Clustering analysis (STRUCTURE)
NJ tree Cavalli-Sforza distances Nigeria Cameroon Gabon 1978 1980 Avr. 1983 1986
Moungo producing area
Disease mentions 0 0,2 0,4 0,6 0,8 1 0.0 0.5 1.0 He = 0.43 Fst= 0.1 FST / (1 -FST ) Log distance (km)
P=0.13 (all the samples) P<0.05 (Samples 1 to 5)
The Cameroon situation,
an example of a recently
(
1980) colonized area
10 Populations (~ 290 individuals) 19 microsatellite markers
Diversity and differentiation indices, Clustering analysis (STRUCTURE)
Isolation by distance analysis (discrete populations)
F(1-F) = f (Ln(d)) y = 0,004x + 0,0097 -0,4 -0,3 -0,2 -0,1 0 0,1 0,2 0,3 0,4 0,5 -5 -4,5 -4 -3,5 -3 -2,5 -2 -1,5 -1 Ln (d) F/(1-F) Distance géographique Diff ér enc iat ion g én éti que 5 Forest 6
-A genetic discontinuity was detected at the middle of the transect (between sites 5 and 6)
- An Isolation by distance was detected before the above discontinuity suggesting gradual dispersal through ascospores
2D sampling (105 sites, 570 individuals) 17 microsatellite markers
1D sampling on a 30-km-long transect (90 sites, 321 individuals)
17 microsatellite markers
Isolation by distance analysis (continuous populations)
Spatial clustering analysis (GENELAND)
- The genetic discontinuity was sharp delimiting 2 populations. No landscape
features matched this discontinuity colonization trace (e.g. through gene surfing)
- No isolation by distance high effective population size in
M.fijiensis
issuspected
Perspective: estimate dispersal from allelic clines detected
between the 2 populations
* CIRAD, UMR BGPI, Campus international de Baillarguet, TA A-54K, F-34398 Montpellier Cedex 5, France. Correspondence: jean.carlier@cirad.fr † Present adress; INRA, Nancy-Université, UMR 1136 Interactions Arbre-Microorganismes, F-54280 Champenoux, France.
‡ CIRAD, Persyst, UPR Syst. Banan. Ananas, TA B-26 / PS4, Blvd. de la Lironde, 34398 Montpellier Cedex 5, France.
Along a 300-km-long transect
(Halkett
et al
, 2010 Mol. Ecol.)
On a 50x50 km area
(Rieux
et al
, 2011 Mol. Ecol.)7 7 5 7 53 6 0 9 4 9 8 7 7 8 2 9 2 Africa America Asia Pacific