Pest-generic risk mapping based on spatial and temporal distribution of production

Transcription

1 Pest-generic risk mapping based on spatial and temporal distribution of production Juha Tuomola 1, Hanna Huitu 2, Denys Yemshanov 3 and Salla Hannunen 1 1 Finnish Food Safety Authority Evira 2 Natural Resources Institute Finland 3 Natural Resources Canada, Canadian Forest Service Finnish Risk Assessment in Focus 19 th May 2017

2 Aim Develop a simple methodology that ranks production sites of a crop species according to the relative probability of pest invasion based on the spatial distribution of the crop species during the past 5-10 years enables more efficient targeting of the plant health surveys

3 Framework Not pest specific crop specific! Modelling is performed on a km grid and the results are presented as cell-specific risk indices. Indices do not represent actual probabilities, but they allow comparison of the relative probabilities.

4 Relative probability of invasion Assessed to each cell as follows 1. Propagule pressure 2. Spread 3. Survival Relative probability of invasion

5 Relative probability of invasion Assessed to each cell as follows 1. Propagule pressure 2. Spread 3. Survival Relative probability of invasion

6 PP Volume of imported propagation material PP 1) Relative propagule pressure to a cell is assumed to depend linearly on the production area of the studied crop plant in that cell Volume of imported propagation material Production area Production area

7 1. Propagule pressure

8 Relative probability of invasion Assessed to each cell as follows 1. Propagule pressure 2. Spread 3. Survival Relative probability of invasion

9 Probability 2) Relative probability of spread to the surrounding cells is assumed to be distance-dependent Modeled with a Cauchy dispersal kernel 1,0 0,8 k(x, y) = ǁx yǁ2 σ 0,6 0,4 0,2 0,0 0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5 5,5 6 Distance (km)

10 1. Propagule pressure 2. Spread

11 Relative probability of invasion Assessed to each cell as follows 1. Propagule pressure 2. Spread 3. Survival Relative probability of invasion

12 Probability 3) Relative probability of survival in a cell is assumed to depend on the presence of the studied crop species in that cell If host plants are cultivated in the cell, the index remains unchanged If host plants are not cultivated in the cell, the index decreases according to a predetermined proportion 1,0 0,8 0,6 0,4 0,2 0, Year t Year t+1 Year t+2 Year t+3 Year t+4

13 2. Spread 3. Survival

14 Relative probability of invasion Assessed to each cell as follows 1. Propagule pressure 2. Spread 3. Survival Relative probability of invasion

15 Accumulation of the probability over the years 1. year 2. year 3. year 1. PP 1. PP 1. PP 2. Spread 2. Spread 2. Spread 3. Surviv. 3. Surviv. 3. Surviv. Probability of invasion after year 1 Probability of invasion after year 2 Probability of invasion after year 3

16 Accounting pest specific traits Pest generic approach! fit for all purpose parameters of spread and survival cannot be set Probability distributions of the spread behavior parameter (dispersal) and survival parameter values are defined Monte Carlo simulation is used to obtain a distribution of risk index values for each cell

17 Ranking of the cells Results need to be presented in a simple, user friendly form Ranking of the cell-specific probability distributions Probability distributions need to be ranked in a way that takes into account the whole distributions Distributions are ranked with a pairwise stochastic dominance rule Hypervolume indicator denotes the quantitative position of each rank

18 Potato as an example ( )

19 Summary Spatial and temporal distribution of places of production is used to map the invasion probability of pests with risk indices Indices do not represent actual probabilities, but they allow comparison of the relative probabilities Simple methodology for a practical problem, i.e. targeting phytosanitary surveys Results are published in an easy-to-use web mapping service for risk managers in Finland

20 20