Predicting the dispersal and invasion dynamics of ambrosia beetles through demographic reconstruction and process-explicit modeling

Abstract

Evaluating potential routes of invasion of pathogens and vectors of sanitary importance is essential for planning and decision-making at multiple scales. An effective tool are process-explicit models that allow coupling environmental, demographic and dispersal information to evaluate population growth and range dynamics as a function of the abiotic conditions in a region. In this work we simulate multiple dispersal/invasion routes in Mexico that could be taken by ambrosia beetles and a specific symbiont, Harringtonia lauricola, responsible for a severe epiphytic of Lauraceae in North America. We used Xyleborus bispinatus Eichhoff 1868 as a study subject and estimated its demography in the laboratory in a temperature gradient (17, 20, 26, 29, 35 °C), which we then used to parameterize a process-based model to estimate its metapopulation dynamics. The maximum intrinsic growth rate of X. bispinatus is 0.13 with a thermal optimum of 26.2 °C. The models suggest important regions for the establishment and dispersal the states of Veracruz, Chiapas and Oaxaca (high host and secondary vectors diversity), the Isthmus of Tehuantepec (connectivity region), and Michoacán and Jalisco (important avocado plantations). The use of hybrid process-based models is a promising tool to refine the predictions applied to the study of biological invasions and species distributions.

Figure 1

Mathematical function of growth rate of X. bispinatus at different temperatures considering a 95% confidence interval.

Figure 2

(A) Map with the suitable sites estimated from the ecological niche modeling, the records of X. bispinatus presence, and the starting coordinates used for the spatiotemporal simulations. (B) Projection in geography of X. bispinatus growth estimated from the intrinsic growth rate function (r_i map) used for the construction of the process-explicit models. This figure was generated in ArcGIS ver. 10.4 (https://www.arcgis.com/features/index.html).

Figure 3

Predicted invasion dynamics and abundance of the ambrosial complex X. bispinatus–H. lauricola from the port of Salina Cruz and Veracruz at different time periods. Maps were generated with the raster package ver. 3.6 and wesanderson color palette (https://github.com/karthik/wesanderson) in R.

Figure 4

Predicted invasion dynamics and abundance of the ambrosial complex X. bispinatus–H. lauricola from the port of Manzanillo and Altamira at different time periods. Maps were generated with the raster package ver. 3.6 and wesanderson color palette (https://github.com/karthik/wesanderson) in R.

Figure 5

Predicted invasion dynamics and abundance of the ambrosial complex X. bispinatus–H. lauricola from the port of Nogales and Texas at different time periods. Maps were generated with the raster package ver. 3.6 and wesanderson color palette (https://github.com/karthik/wesanderson) in R.

Figure 6

Workflow diagram used for the construction of the process-explicit model simulations. The figure was assembled with the growth curve depicted in Fig. 1, along with maps created using the raster package version 3.6 in R and ArcGIS version 10.4. (https://www.arcgis.com/features/index.html).