Context-dependent disturbance synergies: Subcortical competitors may constrain bark beetle irruption following wildfires
Data files
Mar 11, 2025 version files 154.44 KB
-
Mitchell_et_al_2025.rmd
24.21 KB
-
Plot_Level_All_Data.csv
6.51 KB
-
README.md
15.39 KB
-
Tree_Level_Performance.csv
14.30 KB
-
Tree_Level_Preference.csv
94.04 KB
Abstract
Wildfires and bark beetles have historically interacted to create complex and resilient forests. However, recent record-breaking wildfires in western North America raise concerns that the large areas of injured and dead trees could facilitate increases in insect populations that respond to resource pulses. Populations of Douglas-fir beetle (Dendroctonus pseudotsugae), the primary mortality agent of Douglas-fir (Pseudotsugae menziesii) often irrupt following fires due to the resultant ephemeral pulses of defensively compromised hosts. Other subcortical phloeophagous insects are also attracted to fire (e.g., woodboring Coleoptera: Buprestidae, Cerambycidae), and similarly colonize damaged trees. Although Douglas-fir beetle and woodboring beetle species are known to colonize the phloem of injured trees, the potential for interactions among them following fire is relatively unknown. Rapid colonization by woodborers of the bark beetle niche following fires could constrain bark beetle population growth, potentially suppressing population irruptions through subcortical competition. To evaluate this hypothesis, we studied three wildfire complexes in mature Douglas-fir forests that burned in British Columbia in 2017. We found Douglas-fir beetle preferentially colonized mature stands containing large-diameter trees with moderate fire injury, and that these trees were frequently co-colonized by woodborers. In the absence of woodborers, we found that potential rates of increase of Douglas-fir beetle populations (i.e., offspring per female) were sufficient to lead to a local population irruption. Conversely, when woodborers were common (>50% of trees infested per stand), potential rates of increase of Douglas-fir beetle populations fell below replacement. These findings suggest that competition by woodboring beetles may suppress irruptions of Douglas-fir beetle in fire-injured forests. Our results reveal complex, context-dependent interactions among disturbance agents, and indicate that population irruptions by resource pulse-driven bark beetles following fire may depend upon the response of local subcortical competitors. Forest management practices that enhance the diversity and abundance of non-irruptive phloeophagous insects such as many woodboring beetle species may limit the potential for wildfires to contribute to subsequent bark beetle outbreaks.
https://doi.org/10.5061/dryad.dv41ns262
Description of the data and file structure
These data were derived from field sampling conducted in the interior Douglas-fir forests of British Columbia by Mitchell et al., in 2019. Additional data (not shared here) was derived from the same field season and published in Kitchens* et al., 2022 (https://doi.org/10.1016/j.foreco.2022.120203).
*Note: Kate Mitchell has previously published under her maiden name Kitchens, but this is the same researcher across both studies
Four files have been provided and attached.
Three excel .csv files containing raw field data are named as follows, and apply to the relevant models/hypotheses which were tested in the manuscript:
Tree Level Preference.csv (for the testing of the tree level preference model)
Tree Level Performance.csv (for the testing of the tree level performance model)
Plot Level All Data.csv (for the testing of both the performance and preference plot level models)
Some information across these .csv files is repeated, as all data was derived from the same field sampling season and protocols outlined in Mitchell et al., 2025 (see Methods). For information about individual variables and how they were collected, see Methods. For descriptions of variables, see tables below. Blank cells represent missing data.
One R Markdown (.rmd) file containing cleaned and summarized code pertaining to model creation and evaluation is also attached as:
Mitchell et al 2025.rmd
There is no attached readme file for the code, as all directions and relevant information are structured into the markdown file as needed. As stated in the code, if there are any questions please reach out to Kate Mitchell (kate.a.mitchell@gov.bc.ca) and Allan Carroll (allan.carroll@ubc.ca).
Sharing/Access information
The authors request that any additional work on these data (including but not limited to follow-up publications) be done with the permission of the Forest Insect Disturbance Ecology Lab (FIDEL) PI Allan Carroll. For permission, additional information or questions about these data, please contact allan.carroll@ubc.ca.
Variable Descriptions
Plot Level All Data.csv
Variables | Units | Description |
---|---|---|
complex | categorical | Fire complex names, EH = Elephant Hill, CTA = Hanceville Riske Creek, GL = Prouton Lakes |
plot | categorical | Plot name/identifiers |
dbhallspp | cm | Diameter at breast height (dbh), calculated as average for all species |
dbhdf | cm | Diameter at breast height (dbh), plot level average for Douglas-fir (df) only |
ba | m^2/ha | basal area |
htallspp | m | plot-level average of tree height for all species |
htdf | m | plot-level average of tree height for Douglas-fir (df) only |
charallspp | proportion | plot-level average of proportion of bole char (m) as a function of total tree height (m), calculated for all species |
chardf | proportion | plot-level average of proportion of bole char (m) as a function of total tree height (m), calculated for Douglas-fir (df) only |
dfbsum | number | sum of Douglas-fir beetle (dfb) infested trees per plot |
wbsum | number | sum of woodborer (wb) infested trees per plot |
conifersum | number | sum of conifer trees per plot (all species) |
dfhostsum | number | sum of Douglas-fir (df) trees per plot |
dfprop | proportion | proportion of Douglas-fir (df) trees per plot |
dfbpropallspp | proportion | proportion of Douglas-fir beetle (dfb) attacked trees as a function of all conifer tree species per plot |
dfbprop | proportion | proportion of Douglas-fir beetle (dfb) attacked trees as a function of all Douglas-fir trees (df) per plot |
wbprop | proportion | proportion of woodborer (wb) attacked trees as a function of all conifer tree species per plot |
r | offspring/female | number of successfully emerging Douglas-fir beetle offspring per plot / number of initiating maternal gallery starts per plot. For more information see Methods. |
dfbdistkm | km | distance to existing Douglas-fir beetle outbreaks |
fd250mprop | proportion | proportion of Douglas-fir trees within the surrounding 250m areas of a given plot, as derived by VRI; for more information see Methods. |
fd250marea | proportion | total area of Douglas-fir trees within the surrounding 250m areas of a given plot, as derived by VRI; for more information see Methods. |
densitystemsha | stems/ha | plot-level stand density, as a function of stems per ha |
cm = centimeters, m = meters, km = kilometers, ha = hectares
Tree Level Preference.csv
Variables | Units | Description |
---|---|---|
complex | categorical | Fire complex names, EH = Elephant Hill, CTA = Hanceville Riske Creek, GL = Prouton Lakes |
plot | categorical | Plot name/identifiers |
Species | categorical | tree species identifier |
DBH | cm | Diameter at breast height |
avg char | proportion | proportion fire injury (m) / tree height (m) |
Tree Height | m | total tree height (m) |
wb | 0/1 | presence/absence of woodborer (wb) infestation per tree |
dfb-18 | 0/1 | presence/absence of Douglas-fir beetle (dfb) infestation per tree |
plot-ba | m^2/ha | basal area per plot |
dfprop | proportion | proportion of Douglas-fir (df) trees per plot |
wbprop | proportion | proportion of woodborer infested trees per plot |
fireweightavg | numerical | weighted average fire injury per plot |
fireunweightvar | numerical | unweighted average variation in fire injury per plot |
dfbdistkm | km | distance to existing Douglas-fir beetle outbreaks |
fd250mprop | proportion | proportion of Douglas-fir trees within the surrounding 250m areas of a given plot, as derived by VRI; for more information see Methods. |
fd250marea | proportion | total area of Douglas-fir trees within the surrounding 250m areas of a given plot, as derived by VRI; for more information see Methods. |
densitystemsha | stems per ha | plot-level stand density, as a function of stems per ha |
Tree Level Performance.csv
Variables | Units | Description |
---|---|---|
complex | categorical | Fire complex names, EH = Elephant Hill, CTA = Hanceville Riske Creek, GL = Prouton Lakes |
plot | categorical | Plot name/identifiers |
treeid | categorical | tree name/identifier |
ba | m^2/ha | plot basal area |
mat_gall_length | cm | length of initiating maternal galleries |
imagej_dfb_perc | % | Perc of r-survey samples (ideally x2 per tree) occupied by Douglas-fir beetle (dfb) galleries |
wb_pa | 0/1 | presence/absence of woodborer (wb) infestation per tree |
dfb_pa_18 | 0/1 | presence/absence of Douglas-fir beetle (dfb) infestation per tree |
offspring_total | number | number of successful emerging Douglas-fir beetle offspring per tree |
offspring_gallery | proportion | offspring / gallery length (cm) |
rbai1 | numerical | relative basal area increment for 1 year pre-fire, see Methods for calculation. |
rbai5 | numerical | relative basal area increment for 5 years pre-fire, see Methods for calculation. |
dbh | cm | Diameter at breast height per tree |
height | m | Height per tree |
char_perc | % | Perc of bole char (m) / tree height per tree |
dfprop | proportion | proportion of Douglas-fir (df) trees per plot |
wbprop | proportion | proportion of woodborer infested trees per plot |
fireweightavg | numerical | weighted average fire injury per plot |
fireunweightvar | numerical | unweighted average variation in fire injury per plot |
dfbdistkm | km | distance to existing Douglas-fir beetle outbreaks |
fd250mprop | proportion | proportion of Douglas-fir trees within the surrounding 250m areas of a given plot, as derived by VRI; for more information see Methods. |
fd250marea | proportion | total area of Douglas-fir trees within the surrounding 250m areas of a given plot, as derived by VRI; for more information see Methods. |
densitystemsha | stems per ha | plot-level stand density, as a function of stems per ha |