During the recent outbreak in western Canada, mountain pine beetle (MPB) Dendroctonus ponderosae Hopkins (Coleoptera: Curculionidae) and its associated fungi expanded their geographic and host ranges to include jack pine, Pinus banksiana Linnaeus (Pinaceae), in the boreal forest. While previous research indicates that MPB and its symbiotic fungi can infest and induce chemical host defences in jack pine, it is uncertain how jack pine defences will impact MPB-associated fungi, particularly under novel climatic conditions. We investigated the growth response of the three known MPB fungal symbionts (Grosmannia clavigera (Robinson-Jeffrey and Davidson), Ophiostoma montium (Rumford), and Leptographium longiclavatum (Lee, Kim, and Breuril) in media amended with constitutive and fungal symbiont-induced jack pine monoterpene profiles from jack pine at three temperatures (18 °C, 23 °C, and 30 °C). While the impact of host monoterpene defence profile on fungal growth only marginally varied between temperatures, there were significant effects of temperature on the optimal growth of fungal species. Fungal growth was facilitated by host monoterpene profiles induced by other species of MPB-associated fungi, suggesting indirect host-mediated interspecific interactions between MPB fungal symbionts that may assist in maintaining redundant symbioses. Since the species forming the community of symbiotic fungi have different requirements, regarding both optimal temperature ranges and host monoterpene composition, the ongoing climate change may alter the suitability of jack pine as a host, potentially impacting the establishment of MPB populations in these forests.

The experiment was conducted at the Northern Forestry Centre in Edmonton, Alberta, Canada, in January and February 2024. Using a bioassay where media was amended with monoterpenes, we evaluated the impact of four host defence treatments on fungal growth, including a control without monoterpenes and three jack pine monoterpene profiles from Cale et al. (2019): (1) constitutive monoterpene concentrations, (2) induced monoterpene concentrations in RZs after inoculation by L. longiclavatum, and (3) O. montium (Table 1). The monoterpene profiles induced by L. longiclavatum and O. montium were selected as host defence treatments because they are most dissimilar, with the G. clavigera-induced profile representing an intermediate (Cale et al., 2019). To maintain manageable sample sizes, the monoterpene profile induced by G. clavigera was not included as a host defence treatment. We also evaluated fungal growth at three temperatures (18 °C, 23 °C, and 30 °C) that reflect a range of conditions at which growth was previously observed in these species (Rice et al., 2008), and climatic conditions that MPB-associated fungi may experience in the extended boreal range under climate change (Wotherspoon et al., 2023).

Malt extract agar (2 % MEA; 20 g malt extract and 15 g L-1 agar) was prepared, autoclaved, and cooled before directly adding all monoterpenes into agar using established methods (Ullah et al., 2021; Wang et al., 2020). Synthesized monoterpene amendments contained a mixture of six compounds that make up 95 % of the total concentration of jack pine monoterpene profiles: +/- α-pinene (98% chemical purity), (-)-β-pinene (99 %), (+)-β-pinene (98.5 %), myrcene (> 90 %), (-)-limonene (> 99 %), (+)-limonene (> 97 %), terpinolene (> 85 %), and 3-carene (> 90 %) (Sigma-Aldrich, St. Louis, MO, USA) (Cale et al., 2019). Monoterpene concentrations used were from the jack pine reaction zone tissue after inoculation by L. longiclavatum and O. montium and from un-infected phloem in Alberta (Cale et al., 2019). We poured 20 mL MEA into plastic Petri dishes (100 mm diam. x 15 mm ht.), cooled for 24 hours at room temperature, and inoculated each plate with one 4-6 d old actively growing fungal culture. All Petri dishes were sealed with parafilm. Fungal cultures included two strains each of G. clavigera (strains 2894 and 2896), L. longiclavatum (strains 2892 and 2955), and O. montium (strains 2889 and 2950) (Northern Forestry Centre Culture Collection) to account for some variation between and within fungal species. To inoculate each plate, a single plug (8 mm dia.) was taken from the edge of an actively growing culture and placed on top of the MEA in the centre of the plate.

The six fungal strains, three temperatures, and four host defence treatments were crossed in a full factorial design for a total of 72 treatment groups (n = 6-9; N = 583). Petri dishes within each species and host defence treatment group were randomly placed in an incubator set at 18 °C, 23 °C, or 30 °C. The growth of each culture was traced daily for 4 d. After the growing period, we scanned the plates and measured the daily and total area of growth using ImageJ software (Schneider et al., 2012). Since several fungal cultures had grown to the edge of the Petri dish by the fourth day, fungal growth measured on the third day was used as the dependent variable in all analyses. To maintain manageable batch sizes, we conducted the experiment over two date ranges (blocks), two weeks apart, each containing half of each treatment combination.

All statistical analyses were performed using R (ver. 4.4.1), using α = 0.05. Post-hoc tests were performed using the emmeans package (ver. 1.10.6-090001, Lenth 2024). To evaluate if the impact of temperature on fungal growth varies with species, we ran generalized linear mixed-effects models (GLMMs) for each defense treatment using the glmmTMB package (ver. 1.1.10, Brooks et al., 2017), with species, temperature, their interaction, and block as fixed effects, and fungal growth as the response variable. Strain was included as a random effect. Models for each defense treatment were run using a Tweedie family distribution with a log link (Jørgensen, 1987). This distribution was a better fit than Gaussian, negative binomial and Poisson distributions, based on AIC values and residual plots. Differences between species within each temperature were evaluated with Tukey’s post hoc test.

The facilitative or inhibitory effects of host defences on fungal growth were further explored by calculating the percent difference in growth between each host defence treatment and the average of their corresponding control for each species and temperature. We compared percent difference in growth from control using the same GLMM model structure and post-hoc tests as growth response variables.

To determine whether jack pine host monoterpene profiles exert inhibitory, neutral, or facilitative effects on fungi, and if these effects are temperature-dependent for each fungal species, including an assessment of potential indirect interactions between fungal species, the data were subjected to an ANCOVA using the gls function (nlme package, ver. 3.1-166, Pinheiro & Bates, 2000). The response variable was fungal growth with a square-root transformation applied for all three fungal species. The predictor variables were temperature, host defence treatment, their interaction, block, and strain. Non-significant interaction terms were removed from the final model. A separate model was run for each species. The models were visually evaluated for normality and homogeneity of variance with Q-Q and residual plots. To account for heterogeneous variances in the models for G. clavigera and L. longiclavatum, we included a variance structure defined by the combination of temperature and host defence treatment. When the temperature, treatment, or their interaction were significant, differences were evaluated using a Tukey’s post-hoc test. When the interaction term was significant, we tested the differences between each combination of temperature and treatment.