Modelling the geographical distribution of two bark ... - Jean-Pierre Rossi
May 11, 2012 - These factors are advantageously used to model species distribution ... Generalized Linear Model with logit link function (logistic regression).
Modelling the geographical distribution of two bark beetles Tomicus destruens and Tomicus piniperda in Europe and the Mediterranean region Jean-Pierre Rossi?, Agnès Horn, Francois Lieutier & Carole Kerdelhué? ?
CBGP - INRA & Univ. Orléans, France
7-03-14 IUFRO Group Symposium
“Entomological Research in Mediterranean Forest Ecosystems” May 11, 2012
Introduction
Climate and resource availability are major determinants of geographical distribution of species Climate strongly impacts survival, development and growth rate of insects These factors are advantageously used to model species distribution () Species Distribution Modelling, SDM) Invited conference by G. Gader (expert-GIZ, Tunisia) about vulnerability of forest ecosystems to climate change
Introduction
2
What is Species Distribution Modelling ? Quantitative predictive models of species-environment relationships I
Potential species distribution
Introduction
3
Tomicus destruens & T. piniperda Subject species Bark beetles of the genus Tomicus are considered among the more damaging insects of Eurasian pine forests Tomicus destruens has long been mistaken for T. piniperda due to high morphological similarity Molecular studies on mitochondrial and nuclear markers have shown they are distinct species (Kerdelhué et al., 2002 among others) Morphological analysis provided diagnostic morphological characters (Faccoli, 2006)
Species
4
Tomicus destruens & T. piniperda Questions Can we identify climate descriptors explaining species pattern ? Can we draw detailed maps of both species present distribution ?
Available information A set of valid species occurrences (molecular analyses) in Europe and the Mediterranean region (sampling + data from literature) Variables describing climate within the study area
Species
5
Tomicus destruens & T. piniperda Questions Can we identify climate descriptors explaining species pattern ? Can we draw detailed maps of both species present distribution ?
Available information A set of valid species occurrences (molecular analyses) in Europe and the Mediterranean region (sampling + data from literature) Variables describing climate within the study area
Species
5
Sampling design : Tomicus destruens
⌅ Presence (species sampled) ⇤ Absence (the species has been searched for and was absent)
Material and methods
6
Sampling design : Tomicus piniperda
⌅ Presence (species sampled) ⇤ Absence (the species has been searched for and was absent)
Material and methods
7
Climate data source : the worldclim data base Current conditions (interpolations of observed data, representative of 1950-2000) Spatial resolution of about 1 square kilometer Monthly mean, min & max temperatures and precipitations Example : Annual Mean Temperature (˚C)
Material and methods
8
Data modelling Model linking species occurrences to climate data Generalized Linear Model with logit link function (logistic regression) Links presence/absence data and quantitative data (climate) Model ! probabilities of presence of each species across geographical range Predicted presence-absences are derived from probabilities using a threshold
Predicting future geographical distributions The fitted model can be used with data depicting future climate We used the IPCC 4 climate scenario from CIAT (climate model CCCM + emission scenario a2a)
Data modelling
9
Results Significant predictors T. destruens Min Temperature of Coldest Month (BIO6) Mean Temperature of Warmest Quarter (BIO10) Mean Temperature of Coldest Quarter (BIO11) T. piniperda Min Temperature of Coldest Month (BIO6) Mean Temperature of Warmest Quarter (BIO10) Mean Temperature of Coldest Quarter (BIO11) Annual Mean Temperature (BIO1) Mean Temperature of Wettest Quarter (BIO8) Horn et al., 2012 Results : present distributions
10
Example of climate-species relationships
Results : present distributions
11
Predicted geographic range of T. destruens
(Horn et al., 2012)
⌅ Presence (species sampled) ⇤ Absence (the species has been searched for and was absent)
Results : present distributions
12
Predicted geographic range of T. piniperda
(Horn et al., 2012)
⌅ Presence (species sampled) ⇤ Absence (the species has been searched for and was absent)
Results : present distributions
13
Predicted zone of sympatry
(Horn et al., 2012)
⌅ T. piniperda sampled ⇤ T. destruens sampled
Results : present distributions
14
T. destruens - today
Results : future distributions
15
T. destruens - 2020
Results : future distributions
16
T. destruens - 2050
Results : future distributions
17
T. destruens - 2080
Results : future distributions
18
T. piniperda - today
Results : future distributions
19
T. piniperda - 2020
Results : future distributions
20
T. piniperda - 2050
Results : future distributions
21
T. piniperda - 2080
Results : future distributions
22
Sympatry - today
Results : future distributions
23
Sympatry - 2020
Results : future distributions
24
Sympatry - 2050
Results : future distributions
25
Sympatry - 2080
Results : future distributions
26
Conclusions 1/3 T. destruens and T. piniperda have contrasted ecological requirements Both species are linked to similar environmental factors (mostly temperature) but
T. destruens is associated to higher temperatures Reproduction and larval development occur in automn and winter
T. piniperda is associated to lower temperatures Overwintering under bark at base of tree during cold season Requires low temperatures in winter that trigger the obligate period of maturation
Conclusions
27
Conclusions 2/3 Distribution maps Maps are useful pictures of potential distributions Design additional samplings to build more accurate models Search for species where it is predicted while never been observed e.g . mountaineous regions of North Africa seem suitable for T. piniperda
Conclusions
28
Conclusions 3/3
Implications for pest management Preventive measures/monitoring can be adapted according to predicted maps ! earlier sampling in region with T. destruens because host attacks are expected months earlier Distribution forecasts give insight of possible future expansion/regression of both species ! helpful targeting of early monitoring and management practices
Conclusions
29
Conclusions 3/3 Implications for pest management Preventive measures/monitoring can be adapted according to predicted maps ! earlier in region with T. destruens because host attacks should be expected months earlier Distribution forecasts give insight of possible future expansion/regression of both species ! helpful targeting of early monitoring and management practices
destruens and Tomicus piniperda in Europe and the Mediterranean region. Agn `es Hornâ .... (i.e. the female can oviposit successively in different individual host trees). Because ..... man & Hall, CRC, Boca Raton, Florida. Fielding, A.H. & Bell ...
The common model for k production accounts for two major processes ..... matical foundation for a distributed drag force approach, Boundary-Layer Meteorol.
After accounting for the effects of phylogenetic relatedness and species prevalence, these four ... 2006, Pearson et al. 2006, ... October) to collect information about fish assemblages in a ...... Measuring the accuracy of diagnostic systems.
framework for specifying stochastic models for the dynamics of wild animal pop ... lutionary, conservation and management issues for wild populations (Pollock, 1991; ... Finally, Section 4 discusses important developments of capture-recapture ...
Legal daughters of the carnivalized nation in jeanpierre ronfards shakespearean adaptations lear and ... research electronically, as you are saving all that paper.
traction systems, two DC motors are connected in series in order to use only .... It then appears that it is sufficient to implement two usual vector control algorithms ..... X. Kestelyn, A. Bouscayrol, âSensitivity of a 5-phase Brushless DC machin
modifications in the integro-difference approach are more dramatic and include coexistence of the two-species among the possible results of the model. Finally ...
are depicted as 'invisible goods' by international economics theories. ... Knowledge Intensive Business Services are highly concentrated, and are becoming ...
3 http://www.electrosensible.org. 4 The concept ... the issue is in the process of normalization.5 Quite often, according to the doctrine founded on the principles of ...
larval development generally occurs through the winter. ... that they build on the tips of pine branches in winter and ... (2009) were the first to mention that ... distribution range of both species in Turkey for the first ... A 95% confidence stati
Apr 29, 2009 - Keywords: phylogeography; mitochondrial DNA; Europe; glacial refugia; Tomicus .... T. piniperda specific primers were used to amplify a.
The second approach, called Meta-Molecular Matching Pursuit. (M3P), is based on the a ... For the efficient representation of audio signals, it is hoped that.
The soil is sandy (sand ca. 80%) with low ..... nity in relation to soil properties and yield on a grass- ... worms and soil properties in an arable loess soil. Soil Bi-.
Pages 81-96 in: Evolutionary biology of host-parasite ... combination of local dynamics and regional movement, and (3) geographic scales, .... is aggravated by parasites, then the life-style of what are potentially the most Type 2 virulent ...
The enriched DNA was amplified by polymerase chain ... and 20 were used to design primers. ... primer 2, 0.07 µM of fluorescent-labelled primer 2, 0.8 mM.
J.-P. Rossi a,*, J.-C. Samalens a, D. Guyon b, I. van Halder a, H. Jactel a, P. Menassieu a, D. Piou a,c a INRA ..... Cambridge. University Press, Cambridge.
organisations (contracts between universities and industries) or with other enterprises (vertical ..... 1996). The theory suggesting that networks are more.
number of such zeros with ordinate less than T is proportional to T log T, most ... with respect to the critical line (s)=1/2 : for each non trivial zero s = Ï +it,.
Jun 1, 2004 - was performed using two complementary approaches: molecular ..... representation makes it possible to predict the shape, average value and ...