Transcript Slide 1

Sharp increases in UK Fasciola hepatica
abundance: driven by climate change?
Jan van Dijk
Cyril Caminade
Diana Williams
Matthew Baylis
DELIVER
01/07/2010
The life cycle of Fasciola hepatica
Fasciola as a zoonosis
Worldwide 2.5 million people
infected (mainly southern
America, France/Portugal and
Egypt)
180 million people thought to be
at risk of infection
Especially women and children
affected
Infection associated with
consumption of raw
vegetables/salad leaves grown
on irrigated pasture
Severe fluke infection has been
described in children in the
Lake District (wild watercress)
Fasciola in livestock
Chronic wasting/ production
losses in cattle and sheep
Acute death also observed
in sheep
Worldwide, economic losses
estimated at $US 3 billion
Evidence of infection in 70%
of UK dairy herds
Losses in UK cattle alone
total £40 million p.a.
Trends: Chronic GB fasciolosis (VIDA
database 1977-2008)
2.5
2
1.5
Cattle
1
rs = 0.829, p < 0.001
0.5
0
1975
1980
1985
1990
1995
2000
2005
2010
4
3.5
3
Sheep
2.5
2
1.5
rs = 0.771, p < 0.001
1
0.5
0
1975
1980
1985
1990
1995
2000
2005
2010
NB: Other parasites appear to do well too…
Teladorsagiosis/
Trichostrongylosis
Round
worms
infecting
sheep
Haemonchosis
Nematodirosis
All trends rs≥0.540, p≤0.004
Veterinary Surveillance (VIDA) dataset
Causal web
Stocking
density
Motivation
farmer
(Lamb price)
Laboratory
submissions
Diagnoses
Tests
available
Anthelmintic
resistance
Climate
change
Analysis of confounders: only climate (change), and perhaps anthelmintic
resistance, likely to significantly influence diagnostic rate (van Dijk et al. 2008)
Recent UK trends (past 5-10 years)
•
•
•
•
•
•
Sharp overall UK increase in fasciolosis incidence
Many reports on increase coming from Scotland
Emergence in (south) east Scotland
Emergence in East Anglia (2002)
Acute fasciolosis in calves and adult pigs
Emergence of the rumen fluke Paramphistomum
Hypotheses:
“It must be climate change”
Increase in animal movements
Resistance to wormers
Back to basics - Summer infection
Ollerenshaw (1959)
Disease
Metacercariae (herbage)
Multiplication in snail
Start
Egg development
Egg output
January
June
e.g. Acute disease in autumn mainly resulting from eggs which developed
during April-June and cercariae developed (in snail) during July-September
December
Back to basics - Winter infection
Disease
Metacercariae (herbage) - year 2
Multiplication in snail - year 2
Multiplication in snail - year 1
Egg development
Egg output
January
June
Ollerenshaw (1959)
December
Chronic fasciolosis (1977-2007)
0.25
0.2
0.15
0.1
Cattle
0.05
0
Jan
Spearman:
Feb
Mar
Apr
- - -
May
Jun
Jul
+
Aug
Sep
Oct
Nov
Dec
Increased over-winter survival snails?
0.25
Decrease in over-winter survival metacercariae?
(NB: van Dijk et al. 2008 reported the same for
nematodes)
0.2
0.15
0.1
Sheep
0.05
0
Jan
Spearman:
Feb
- -
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Acute fasciolosis (1977-2007) -Sheep
2
1.8
1.6
1.4
1.2
1
rs = 0.616, p < 0.001
0.8
0.6
0.4
First significant: 2001
0.2
0
1975
1980
1985
1990
1995
2000
2005
2010
Earlier start to the
summer infection?
0.35
0.3
0.25
Increased overwinter survival
snails?
0.2
0.15
0.1
0.05
0
Jan
Spearman:
Feb
Mar
Apr
May
Jun
+
Jul
Aug
Sep
+ +
Oct
Nov
Dec
The timing of development of various stages
Egg development
Results:
In-snail development
(release of first cercaria)
SW- Scotland 76 +/- 28 days
SW-England
37 +/- 10 days
April 5th - June 25th
June 11th – July 28th
88 +/- 33 days
37 +/- 8 days
March 25th - June 28th
June 8th – July 23rd
Egg development mainly April-June; But June also appears to be an important
month for the development of cercaria. The rate of development of cercaria is
relatively independent of temperature.
Validation timing development:
add in-host development and predict outbreaks
0.45
Frequency distribution of VIDArecorded first outbreaks of
summer infection-related acute
fasciolosis, Scotland, 1977-2007
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
August
September
October
Predicted:
95% CI
Mean
95% CI
November
December
Building a simple regression model (1)
• For the 4 most sheep-dense regions, calculate
mean minimum and maximum monthly
temperature, total rainfall and rainydays (>1mm
rain) from surrounding weather stations and
correlate with disease abundance data, 19772007
• Correlate summed April-June (egg development),
July-September (cercarial development) and JanOctober periods (all development)
Building a simple regression model (2)
• In all 4 regions peak disease incidence always
significantly correlated to minimum temperature in
April-June (rs ≥ 0.495, p ≤ 0.01)
• Rainydays (April-June) stronger correlated than total
rainfall
• In Southwest/Wales stronger correlations with
rainydays than in North/Scotland
• In Scotland and North stronger correlations with
temperature
A simple regression model (y = ax + b)
• mean monthly minimum temperature April-June
and total rainydays in April-June explain approx.
50% of the annual variation in diagnostic rate in
all 4 regions
• Run this model on 40 year’s worth of UK ‘GIS
data’ (state-of-the-art Ensembles models,
Geography department, University of Liverpool ):
UK and
Ireland
divided into
25 km
squares
Each square
expressed as
a ~ 40 year
anomaly of
itself
(red: ‘more
fluke’, blue:
‘less fluke’)
But if fluke is rain-dependent,
why is there so much more of it?
λ(T) PM1(T) PSI1(R,S) PSS1 PSM MS(T,R) PMC PEH β H q
F0 =
(μC(T) + βH) μA
F0 = basic reproduction quotient for fluke = the predicted number of
adult offspring of a single fluke being present in a non-immune host
for one year
Model output
Blue = predicted success
Red = measured (disease)
validation
sensitivity analysis
(temperature)
Fasciola: trade- off in the effects of temperature and rainfall
Fasciola offspring →
Scotland
Horizontal (dotted) line: current
situation
Sloped lines: 1°C increase
Southwest
Loss of rainy days (per month) →
Fasciola: trade- off in the effects of temperature and rainfall
In the UK,
temperature
strongly limits
developmental
success
‘parasite success’
increases on a
logarithmic scale
with increasing
temperatures,
almost regardless
of rainfall
Fasciolosis- Conclusions
• Together, temperature and rainfall can explain all
observed UK changes in parasite abundance
• Climate change will alter seasonality and spatial
distribution in ways which may be counter-intuitive
• In the UK, the effect of increasing temperature is likely
to outweigh the effect of decreasing rainfall
• It appears that prevalence of acute autumn fasciolosis
can be predicted in early summer
questions?
answers?