Western Hemlock Dwarf Mistletoe

Western hemlock dwarf mistletoe is a hemi-parasite of primarily western hemlock trees. It absorbs water, nutrients, and carbohydrates from its host. Infected branches produce small structures called aerial shoots that have minor photosynthetic capabilities, but are primarily for pollination and seed production and require a certain amount of light to emerge from an infection site. Seeds are explosively discharged from aerial shoots when fully mature. Once landing on a susceptible host branch they being germination and mechanical penetration of the host branch.

Research Question

I wanted to explore the spatial patterns of dwarf mistletoe infections after a mixed severity fire and the roles post-fire, forest structure and fire refugia play.

How does the spatial pattern of post-fire stand structure and species composition affect spatial patterns of dwarf mistletoe distribution throughout the stand through physical barriers to susceptible hosts and seed dispersal?

Description of Dataset

I have a 2.2 ha rectangular study area (Wolf Rock), just northwest of the HJ Andrews Exp. Forest that was stem mapped in 1992. Each tree has an X and Y coordinates, as well as the tree species, height, diameter, and a variety of other tree inventory related data. I only have ages for a few western hemlocks, but many more for Douglas-fir. Only one western hemlock core was identified as being over 100 years old, at 170 years. I have a polygon layer for the fire refugia that are documented on the study area. For the western hemlocks I have a presence/absence of western hemlock dwarf mistletoe as well as a severity measure. However, I am unsure of the scale and ethod of rating, so I will not be using it. The infection presence and absence are from a single measurement season.

Hypotheses and Predictions

Western hemlock dwarf mistletoe spreads easily through moderate canopy densities, where distances between trees are close to the average dispersal distance of 2-4 meters. Disturbances that create patchy gaps increase likelihood of spread because of increased light reaching infected branches and increase the rate of infection abundance because of the lack of physical barriers such as high densities of branches and foliage.

Disturbances can also remove the disease from a forested stand by killing or knocking down infected western hemlocks. Post-disturbance regeneration can enable or inhibit the parasite’s reintroduction to the stand. Non-susceptible hosts such as Douglas-fir or western redcedar regenerate readily alongside western hemlock and will intercept seeds. Some gaps are only conducive to western hemlock regeneration which will be readily infected by any surviving infected western hemlocks post-disturbance.

The Wolf Rock stand experienced two fires that created a mosaic of ~110 year old regeneration and >110 year old remnant trees in fire refugia. Remnant, infected western hemlocks survived the most recent fire in those fire refugia. From this stand structure, I have several hypotheses and predictions:

1. Remnant, infected western hemlocks form the center of new infection centers post disturbance so infections in the regenerating susceptible hosts, will be clustered around these remnant trees.
2. Non-susceptible hosts regulate the rate of infection spread through physical barriers to dispersal, so infection cluster size will have an inverse relationship with non-host density.
3. Western hemlock infection spreads from a central remnant tree, so uninfected western hemlocks will have a dispersed spatial pattern from the remnant western hemlock, regardless of non-host density.
4. Post-fire regeneration with higher western hemlock composition will have more susceptible hosts and less physical barriers to spread so infection cluster size will have a positive relationship with western hemlock composition.

Analysis Approaches

For Exercise 1,  I wanted to know about the spatial pattern of western hemlock trees infected with western hemlock dwarf mistletoe. I used a hotspot analysis to determine where clusters of infected and uninfected trees were in my 2.2 ha study area. I discovered a hot spot and a cold spot, indicating two clusters, one of high values (infected) and one of low values (uninfected).

For Exercise 2, I wanted to know how the spatial pattern of these clustered infected and uninfected trees were related to the spatial pattern of fire refugia outlined in my study site. I used Geographically Weighted Regression to determine the significance of this relationship, however I did not find a significant relationship between a western hemlock, its intensity of clustering and infection status, and it’s distance to its nearest fire refugia polygon edge.

For Exercise 3, I wanted to know how the spatial patterns of remnant western hemlocks related to the spatial patterns of regeneration western hemlocks, uninfected western hemlocks, and non-western hemlock tree species. I used the Kcross and Jcross functions in spatstat in R and prepared the data in ArcMap to analyze spatial relationships between trees. I found clustering between regenerating western hemlocks and non-hosts to remnant western hemlocks but the uninfected western hemlock’s spatial pattern was independent of the remnant western hemlocks.

Results

I produced several maps that showed the spatial patterns in my study site which were helpful for understanding and investigating further relationships. I produced several charts from my exercise 3 analysis that were useful for visual representations of the relationships between trees. In Exercise 3 I also produced a map from raster addition that gave me the best visualization of where western hemlock and non-host trees were in the stand. Exercise 2 produced a map and statistical relationship but was not significant in explaining a hemlock’s infection and density status.

Significance

The biggest finding was that the fire refugia polygons are not significant for my analysis, the remnant infected hemlocks are more important explanatory variables in spatial patterns of infected trees. This supported hypothesis 1. Because refugia can be effectively defined using the “for what, from what” framework, western hemlock dwarf mistletoe refugia from fire could be delineated differently in the field focusing only on the remnant western hemlocks.

Data was not available to determine the rates of infection spread over time because I only had one season of measurements. I also could not evaluate the size of clusters because I did not have GPS points of infection center extent so I could not assess hypothesis 2 and 4 directly. However using Ripley’s K and the cross-variant I could see how the clusters changed over distance. I learned that infected, regenerating trees are going to be found closer to the remnant infected trees and that non-host trees may be blocking the spread of mistletoe into an uninfected patch because they were found clustered around remnant trees as well. This provides support for Hypothesis 1, 2, and 3.

Silvicultural prescriptions with the goal to preserve old growth forest structure, but that want to limit the amount of dwarf mistletoe in a forest can appropriately remove old infected hemlocks to preserve infection spread and extent. These prescriptions will also be able to predict future dwarf mistletoe spread. Forest operations that simulate disturbances that leave remnant hemlocks such as harvests, can incorporate spread predictions to limit regeneration being infected.

Learning From The Process

I learned a lot about the spatial analyst tools in ArcMap and how to produce easily interpretable maps and graphs. I also learned how to use several function in spatstat. I learned a lot about interpreting R outputs and spatial. Spatial autocorrelation can tell you a lot about what your data are doing but I thought it was most useful to be able to see on a map or chart what is specifically happening.