Topographic effects on dispersal patterns of Phytophthora cinnamomi at a stand scale in a Spanish heathland

Abstract

Phytophthora cinnamomi is one of the most important plant pathogens in the world, causing root rot in more than a thousand plant species. This observational study was carried out on a P. cinnamomi infected heathland of Erica umbellata used as goat pasture. The patterns and shapes of disease foci and their distribution were described in a spatial and temporal context using an aerial photograph record. A set of topographic traits was selected on the basis of a disease dynamic hypothesis and their effects on observed spatial disease patterns were analyzed. Incipient infections situated in flat terrain expanded as compact circular front patterns with a low growth rate. On slopes, disease patches developed more rapidly down slope, forming parabolic shapes. The axis direction of the parabolas was highly correlated with terrain aspect, while the parabolic amplitude was associated with land curvature and slope. New secondary foci appeared over the years producing an accelerated increase of the affected surface. These new foci were observed in sites where disease density was higher or near sites more frequently visited by animals such as the stable or the forage crop. In contrast, a smaller number of disease foci occur in areas which animals are reluctant to visit, such as where they have a short range of vision. Our results suggest that 1) the growth of existing P. cinnamomi foci is controlled by a combination of root-to-root contact and water flows, 2) the increase in the diseased area arises mainly from the multiplication of patches, 3) the formation of new foci is mediated by long-distance transport due to the movement of animals and humans along certain preferential pathways, and 4) geomorphology and topography traits are associated with the epidemiology of this soil-borne pathogen.

Fig 1. Patches at initial stage of Phytophthora cinnamomi infection on a heathland of Erica umbellata.

Top: small focus with a circular shape on a flat area. Bottom: a focus spreading downhill on a steeper slope. Colonization of the bare soil by Cistus ladanifer started inside the disease front. White bar at the top left is approx. 50 cm length. Images taken by the authors.

Fig 2. Panchromatic aerial image of the study area photographed in 1981.

Triangular patches of pale grey colour are disease foci originated by Phytophthora cinnamomi infection of Erica Umbellata heathland (dark grey background). Foci can be seen emerging from the farm access road (a), tractor tire tracks, see the upper left magnification (5x) window (b), main road at the north edge (c), or a few dispersed points in the south (d). A cultivated rye paddock (e), and a goat stable, marked with a white arrow (f) are identifiable too. Purple line shows study area limits. Printed under a CC BY license, with permission from Junta de Extremadura CICTEX, original copyright: Orthoimagery CC-BY 4.0 CICTEX, Junta de Extremadura.

Fig 3. Colour aerial image taken in 2000 showing Phytophthora cinnamomi infection patches on Erica umbellata heathland in Villuercas Sierra (Spain).

Following traits are identifiable: needle shapes of foci at their initial stage (a), bending effect due to changes in terrain aspect (b), a wide patch in a zone with divergent curvature (c), well developed focus with asymmetric branches following slope aspect (d). All of the disease fronts have parabolic shapes ending where drainage is reached. Contour lines were superimposed to show terrain morphology. Printed under a CC BY license, with permission from Junta de Extremadura CICTEX, original copyright: Orthoimagery CC-BY 4.0 CICTEX, Junta de Extremadura.

Fig 4. Descriptive sketch showing measurements taken on foci of disease caused by Phytophthora cinnamomi in Erica umbellata heathland.

The focus axis along the main direction of spread (red dashed line) was divided in 5m segments (green crosses) to measure focus width and axis azimuth.

Fig 5. Aerial images showing proliferation of disease patches on a heathland (Erica umbellata) caused by Phytophthora cinnamomi.

Year of image acquisition is shown in the upper-left corner. The wide patches on the left side of the 1981–1997 images are ploughed plots. Later, in 2000, a firebreak (the diagonal white line) was built. No disease foci emerging from these features were observed. Printed under a CC BY license, with permission from Junta de Extremadura CICTEX, original copyright: Orthoimagery CC-BY 4.0 CICTEX, Junta de Extremadura.

Fig 6. Map of Phytophthora cinnamomi inoculum introduction sites (IIS).

Data from imagery of a heathland stand of Erica umbellata from 1981 to 1997. Legend is as follows. IIS: black dots. Double solid black lines: roads; green square: goat stable and water supply; discontinuous blue lines: streams; brown continuous lines: contour lines; and finally, the shadowed blue background show the IIS kernel point density, the darker the shadow, the higher the density. Purple line shows study area limits. The map has been created in QGIS by the authors using their own data.

Fig 7. Univariate effects of risk factors analyzed.

Fitted univariate logistic regression models for topographic factors with a significant effect on Phytophthora cinnamomi disease probability. The histograms represent the observed data, in the top are the number of cases (diseased) and in the bottom the number of controls (healthy). Frequencies of both can be read from the left Y axis. The red thick line is the predicted probability for the onset of a new P. cinnamomi focus. The thin red line corresponds to the average probability without the effect of any risk factor. Both probabilities can be read from the right Y axis. Positive openness and plan curvature are dimensionless numbers without units.

Fig 8. Risk map for the occurrence of new Phytophthora cinnamomi foci in the study area.

The map shows probabilities ranks estimated by a logistic model with four risk factors: topographic openness index (visibility), plan curvature of terrain and distances to forage crop and goat stable. The map has been created in QGIS by the authors using their own data.