Impact of a Mountain Pine Beetle Outbreak on Young Lodgepole Pine Stands in Central British Columbia

The current mountain pine beetle (MPB) (Dendroctonous ponderosae Hopkins) epidemic has severely affected pine forests of Western Canada and killed millions of hectares of lodgepole pine (Pinus contorta Dougl. ex Loud. var. latifolia Engelm.) forest. Generally, MPB attack larger and older (diameter > 20 cm or >60 years of age) trees, but the current epidemic extends this limit with attacks on even younger and smaller trees. The study’s aim was to investigate the extent of MPB attack in young pine stands and its possible impact on stand dynamics. Although MPB attacks were observed in trees as small as 7.5 cm diameter at breast height (DBH) and as young as 13 years old, the degree of MPB attack (percent stems ha) increased with increasing tree diameter and age class (13–20, 21–40, 41–60, and 61–80 years old) (6.4%, 49.4%, 62.6%, and 69.5% attack, respectively, by age class) which is greater than that reported from previous epidemics for stands of this age. The mean density of surviving residual structure varied widely among age classes and ecological subzones. Depending on age class, 65% to 77% of the attacked stands could contribute to mid-term timber supply. The surviving residual structure of young stands offers an opportunity to mitigate the effects of MPB-attack on future timber supply, increase age class diversity, and enhance ecological resilience in younger stands. OPEN ACCESS Forests 2015, 6 3484


Introduction
Lodgepole pine (Pinus contorta Dougl.ex Loud.var.latifolia Engelm.) and the mountain pine beetle (MPB) (Dendroctonus ponderosae Hopkins) have co-existed as a natural part of the ecosystem in western North American pine forests for millennia [1].Endemic MPB populations periodically surge during a natural cycle and induce periods of considerable forest mortality as well as playing an important functional role in directing ecological processes and maintaining the biological diversity of forest ecosystems [2,3].However, the current patterns of MPB outbreak appears to be more severe, and impacts have been documented across at least 18.5 million hectares of lodgepole pine forest killing approximately 723 million m 3 or 53% of the merchantable mature pine volume in British Columbia (BC) [4].Active forest management by fire suppression increased the abundance of old pine in forests while climate change, such as mild winters from 1993 onwards, and drier summers resulted in a beetle population expanding exponentially and spreading across the BC landscape [3,5].In central BC, there "is approximately 1.96 million ha of young lodgepole pine stands (<60 years) and more than 357,000 hectares of it had some level of attack during the current MPB outbreak [6].
Based on past MBP outbreaks, early predictions assumed that only mature stands older than 80 years would be significantly attacked by MPB, and stands 61 to 80 years old would have attack rates of about 50%, while stands 41 to 60 years old would be attacked occasionally and stands less than 40 years old would be very rarely attacked [7][8][9].However, due to the tremendous amplification and expansion of the MPB population during the early 2000, young stands were attacked at increasingly greater rates [10][11][12].According to MacLauchlan [10], in central BC 49.2% of young planted stands (stands aged ≤ 60 years old) contained some level of MPB-attack, as well as populations of secondary bark beetles.However detailed studies in young lodgepole pine stands are scarce.Therefore, this investigation provides a unique opportunity to understand the extent and impact of MPB in young lodgepole pine stands.Timber supply analyses undertaken by both government and industry for various forest management units predicted a significant mid-term shortfall as a result of the current MPB epidemic [13] (mid-term = portion of the timber inventory that would be available for harvest within the middle of a normal management cycle i.e., ~30-70 years in central BC [14].It was estimated the allowable annual cut (AAC) would drop by approximately 12.6 million m 3 from the pre-MPB AAC [15].At the provincial level, this shortfall in future harvests is likely to begin in approximately 2016 and last for up to 50 years [16].Generally, post MPB epidemic estimated AAC is based on five assumptions: (1) oldest attacked stands are harvested first; (2) only mature stands are affected with attack rates of about 80 percent in age class 5 and older (>81 years old), and attack rates of about 50 percent in age class 4 (61-80 years of age); (3) no attack in age classes 1 to 3 (≤60 years old); (4) regeneration delay of 15 years in areas not logged; and (5) a shelf life of 10 years with five years for sawlogs [8].The observed MPB-induced mortality in immature lodgepole pine stands invalidates two of the timber supply assumptions, numbers two and three, and is not reflected in current timber supply analysis for central BC [16].Immature mortality caused by the MPB will result in an extension, and possibly a deepening, of the mid-term timber supply fall down period beyond that which has been predicted in any provincial timber supply analysis to date [16].
Further understanding of the role of MPB in immature/young stands (≤60 years) is critical to develop management strategies for these stands.In a previous investigation, Hawkins et al. [17] described the abundance of residual secondary stand structure (SSS) (seedlings, saplings, sub-canopy, and canopy trees that survived a beetle attack) in a wide range of age classes in central BC.Despite promising results, their report did not detail the role of MPB in young lodgepole pine stands nor young stand dynamics after MPB attack.Therefore, further investigation/analysis was carried out to fill this knowledge gap.The main objectives of this study were to describe (i) the extent of mountain pine beetle attack and (ii) post beetle stand dynamics in 13 to 80 years old lodgepole pine leading (dominated) stands in central BC.

Stand Selection
The sampled area was located in the sub-boreal spruce (SBS) biogeoclimatic/ecological zone with three broad subzones dry, mesic, and moist of central British Columbia [18].All investigated stands were lodgepole pine-dominated, unsalvaged stands.Before field reconnaissance, stands for possible sampling were identified on forest cover maps.Seventy percent (70%) of the sampled stands were selected randomly and the remaining thirty percent (30%) were targeted to ensure complete coverage of ecological zones and different age classes.Prior to temporary sample plot (TSP) establishment and sampling, reconnaissance was carried out to determine if stands met the following criteria: (a) lodgepole pine-leading (dominated); (b) within the SBS biogeoclimatic/ecological zone [18]; and (c) based on forest cover maps, age class 1 or AC1 (13-20 years), 2 or AC2 (21-40 years), 3 or AC3 (41-60 years) and 4 or AC4 (61-80 years) (Table 1).Large scale harvesting and silvicultural practices have created a legacy of managed age class 1 and 2 stands in central BC.These stands were established by clear cut logging and planting and generally no further management activities.Prior to the industrial logging era (1960) wildfires, followed by natural regeneration, was the primary disturbance agent and regeneration method for age class 3 and 4 stands.Age class 4 stands are transitional from young to mature stands and are thought to be more susceptible to MPB than younger stands [7].They are included for comparison in our analysis as young stands.
In total, 309 lodgepole pine-dominated stands and 1527 TSP were sampled in age class 1 to 4 stands (Table 1).Approximately half of the samples were in age class 2 (21-40 years old) stands because this is the age class most crucial for forest management decisions regarding midterm timber supply.

Sampling Protocol
The entire survey was completed over the course of three years in the three different ecological subzones (dry, moist, and mesic).Initial stand assessment began in July 2005 in the dry (southwest), followed by the moist (southeast) in 2006, and finishing with the mesic (central) in 2007.Sampling commenced at least 50 m from stand boundaries, roads, trails or forest cover boundaries.At each sample locale, two different types of temporary sample plot (TSP) (5.64 m radius (100 m 2 ) for all residual mature trees (diameter at breast height; DBH ≥ 7.5 cm) and 3.99 m radius (50 m 2 ) for advanced regeneration (diameter at breast height (DBH) <7.5 cm and >4 m height) were established every 50 m along the transect line.When atypical plot locations were encountered, such as areas with excessive wind-throw or a water body, the plot was moved 25 or 50 m along the transect line.All TSPs had a minimum buffer zone of 50 m in all directions.Depending on the size and shape of the stand, 3-10 TSPs were established.For each TSP, site series (soil moisture and nutrient regime), site index (determined in the lab from a site tree core at DBH), and macro aspect (direction relative to noon sun) were recorded to identify and characterize the plot.

Surviving Residual Secondary Structure Data Collection
In this study, secondary stand structure was defined according to the 2008 BC Ministry of Forests and Range Forest Practices and Planning Regulation (FPPR) amendments [19].However, to improve our assessment of residual secondary structure after MPB attack, live secondary structure was divided into two categories: (a) advanced regeneration (>4 m height but <7.5 cm DBH) and (b) residual mature trees (DBH > 7.5 cm).Any regeneration <4 m in height was not considered in this investigation to follow the FPPR amendments (for details, see Hawkins et al. [17]).
For each TSP, species, DBH, stage of MPB-attack, relative crown position (dominant, co-dominant, intermediate, and suppressed) were collected for residual mature trees (DBH ≥ 7.5 cm).The stage of MPB attack was classified according to current MPB attack status as: green attack (entrance holes visible, with or without boring dust, but crown still green), red attack (trees attacked the previous season, foliage red in color but still attached to branches), grey attack (trees killed by MPB two or more years ago, and no longer had any foliage) and no attack (no sign of the beetle).All attacked trees had evidence of MPB entrance holes and when bark was removed from grey trees there seldom was evidence of secondary beetle attack.Young green and red attacked trees were not examined for secondary bark beetles, though they may have been present.Therefore, secondary bark beetles may have a limited influence in the overall assessment.Based on the number of lodgepole pine trees attacked, each age class was divided into one of four lodgepole pine attack categories: A ≤25; B >25 to ≤50; C >50 to ≤75; and D >75% attack.
One tree core at DBH from the largest diameter, defect-free, dominant or co-dominant tree (site tree) was taken in each plot to verify stand age and calculate height over age site index [20] for mature trees and growth intercept site index for secondary structure (SI50 = site index is based on top height and a reference age of 50 years (breast height) in BC.Density and species composition of advanced regeneration (stems > 4 m in height with DBH < 7.5 cm) was also determined for each age class and ecological subzone.

Analysis
All reported MPB attack values are based solely on lodgepole pine.Simple linear regression was carried out to find the relationship between mean MPB attack (percent stems ha −1 ) and mean pine DBH and overall stand density for each age class using SYSTAT version 12 ® (Systat Software, Inc., San Jose, CA, USA).The Kruskal Wallis [21] non-parametric test was conducted when attack categories (A, B, C and D, as the dependent variable) were compared against age classes (independent variables).A chi-squared test was applied to determine the relationship between DBH class and MPB attack.Further analysis was conducted to determine the effects of ecological subzones (dry, moist and mesic) and age class on density of secondary stand structure and MPB-attack percentages, where plots were averaged for each stand and stand average was used in the analysis.During analysis general linear model (GLM)-based analysis of variance (ANOVA) [21] was used instead of mixed-effect model [22] as the attack rates were similar among plots within a stand.Therefore, GLM-based ANOVA can easily cope with most of the variability after averaging the plots within the stands.In addition, pairwise comparisons were conducted using Tukey's multiple comparison test (α = 0.05) to determine differences among age classes and ecological subzones [21].In order to satisfy the ANOVA model assumptions variance, normality and homogeneity were checked before each analysis, and no transformations were required [21].Analyses were conducted using the SYSTAT and programming language R (version 3.1.2)[23].

MPB Attack
Simple linear regression of mean MPB-attack in lodgepole pine as a function of stand mean lodgepole pine DBH showed a significant relationship for all age classes.When regression analysis was carried out based on stand level mean MPB attack as a function of mean stand density only age classes 3 (41-60 years) and 4 (61-80 years) showed a significant relationship (Table 2).There was also a significant difference among age classes (p < 0.001; AC1 = 6.4%,AC2 = 49.4%,AC3 = 62.6%, AC4 = 69.5%)and ecological subzones (p = 0.038; moist = 50.6,Dry = 43.4,and mesic = 40.6%)based on the percentages of MPB-attack.Mean MPB-attack increased with increased age class and the greatest mean MPB attack was observed in age class 4 (69.5%),followed by age classes 3 (62.6%)and 2 (AC2: 21-40 year) (49.4%) (Figure 1).For ecological subzone, the moist subzone showed the greatest rates of attack and was followed by the dry and mesic subzones (Figure 1).Tukey's test for multiple comparisons (α = 0.05) showed age class 1 (AC1: 13-20 years) was significantly (p < 0.001) different from the other age classes and AC2 was different from AC4 (Figure 1).The Kruskal Wallis non-parametric test indicated that only MPB attack category "A" was significantly (H = 16.60,df = 3, p < 0.001) different among age classes, whereas the other categories (B, C, D) were not different from each other.The percentage of stands in the lowest attack category "A" decreased with increasing age class while the opposite trend was observed for attack categories "C" and "D" (Table 3).When testing the differences in distribution of MPB-caused mortality percent across lodgepole pine DBH classes a significant (p < 0.001) relationship was observed.Overall beetle attacks increased with increased DBH class and attack rates exceeded 50% at a DBH of 13.0 cm, 11.0 cm, and 12.5 cm for age classes 2, 3 and 4, respectively (Figure 2).Attack rates were also examined for broader DBH classes: ≤15 cm, >15 to ≤20 cm, and >20 cm to determine MPB-attack preferences with respect to basal area (m 2 ha −1 ) and density (stems ha −1 ).The results indicated the only significant relationship was observed in the smallest DBH class (≤15 cm) as it lost greater basal area (38.7%) than density (stems ha −1 ) (30.5%), whereas no detectable difference was observed for the two larger DBH classes (basal area and density for >15 to ≤20 cm = 67.9% and 67.9% and for >20 cm = 78.0%and 78.1% ) (Figure 3).

Surviving Residual Structure of Immature Stands
GLM (ANOVA) indicated that the density of residual mature trees (stem > 7.5 cm DBH) in post MPB attacked stands was significantly (p < 0.001) different among age classes but not ecological subzones (p = 0.133) (Table 4).The greatest number of residual trees was observed in age class 1 (13-20 years) followed by age classes 3 (41-60 years), and 2 (21-40 years) (Table 5).Tukey's test for multiple comparisons (α = 0.05) revealed a significant difference among age classes except age class 2 and 4 were similar (p = 0.941) (Table 5).Considering the density of advanced regeneration (stem >4 m in height and <7.5 cm DBH) age classes (p < 0.001) and ecological subzones (p = 0.018) were significantly different following MPB-attack (Table 4).The density of advanced regeneration, however, varied considerably across age class and ecological subzone.The greatest number (1485 stems ha −1 ) of advanced regeneration was found in age class 3, whereas age class 4 had the least (555 stems ha −1 ) ( different from the other age classes while the mesic subzone was significantly different from dry and moist subzones (Table 5).
The study also suggests that species composition of secondary stand structure (stem > 4 m height) generally varied from stand to stand.After MPB-attack of lodgepole pine, secondary stand structure in age classes 1 to 3 was dominated by broadleaf species with minor components of conifer species.The most common species for these age classes were aspen (Populus tremuloides Michx.), paper birch (Betula papyrifera Marsh.), black cottonwood (Populus trichocarpa Torr.& A. Gray), lodgepole pine, hybrid spruce (Picea glauca (Moench) Voss × Picea engelmannii Parry), and sub-alpine fir (Abies lasiocarpa (Hook.)Nutt).Conifer species lodgepole pine, hybrid spruce, and sub-alpine fir dominated age class 4 secondary stand structure.
Depending on the age class, 65%-77% of the stands were satisfactorily stocked (2008 FPPR amendments) with ≥900 stems ha −1 (stem height > 4 m) of secondary structure (Figure 4).However when we consider residual mature trees (DBH > 7.5 cm), the results reveal that all stands within the sampling area met minimum stocking levels for MPB attacked stands (minimum residual tree stocking 600 stems ha −1 based on BC Ministry of Forests and Range [24] protocols).If we consider only advanced regeneration (stem >4 m height but <7.5 cm DBH) 22%-46% of stands had ≥1000 stems ha −1 .

MPB Attack
This study indicated the MPB-attack rate in immature or young pine leading stands was a function of tree size, age, and to a lesser extent, stand density.MPB attack percentage increased in young lodgepole pine leading stands with increasing mean stand DBH and age.According to Safranyik and Carroll [1], mountain pine beetle attacks and brood production are directly related to tree age and DBH.Larger trees have thicker bark and phloem, offering more protection for the insects and larvae from natural enemies, extreme temperatures, and sapwood drying.Resistance to MPB-attack is also a function of carbon allocation to resin duct production rather than radial growth [25,26] The average level of MPB-attack for age class 1 stands was very low (6.4%); as a result, these stands will only be brought into the discussion when pertinent.For the other age classes, the percentages of lodgepole pine attack were much higher than the previously reported [7,27,28].According to studies from previous MPB outbreaks, younger stands are less susceptible to MPB attack and have lower brood production due to a thinner phloem and higher growing densities.Other than for age class 3, our regression analysis supports this observation (Table 2).However pre-MPB stand density may not a good predictor of MPB-attack because the mean densities of age class 3 were approximately 40% greater than that of age class 1 and almost double that of age class 4 (Table 5).This suggests less density-dependent mortality in this age class as MPB-induced mortality was similar between age classes 3 and 4.
From historical observations it was reported that more than 50% of lodgepole pine trees with a DBH > 25 cm and a small proportion of trees with a DBH between 10.5-25 cm were attacked by MPB whereas no attack was observed for younger and smaller trees (DBH < 10 cm) [7,9,27].However depending on age class and DBH our study suggests that MPB killed >95% of lodgepole pine trees with a DBH > 23 cm in age class 3 and 4 and 85% with a DBH >22.5 cm in age class 2 (Figure 2).The difference in observations is likely due to the immensity of the current MPB outbreak in central BC compared to previous outbreaks.A similar observation was reported by the BC Ministry of Forests in its 2008 aerial overview survey where around 95% of trees > 22.5 cm DBH were killed by the MPB [6].Based on trees with a DBH ≤15 cm, 35% to 45% of the stems and 45% to 58% of the basal area was killed for age classes 2 to 4 and such mortality levels have previously only been reported for trees with DBH larger than 25 cm [1,7,29].According to Amman et al. [7] MPB-caused lodgepole pine mortality is 1% for trees with a DBH ≤ 10 cm, 20% for a DBH ≤ 20 cm, and 55% for a DBH ≤ 30 cm, whereas Safranyik and Carroll [1] and Björklund et al. [29] observed 10% mortality for trees with a DBH ≤ 10 cm, 40% for trees with a DBH ≤ 20 cm and 70% for trees with a DBH ≤ 30 cm in central BC.Progar et al. [30] observed mortality rates in the Rocky Mountain states of <40% for trees with a DBH below 23 cm and >80% for trees larger than 33 cm.The data from our study demonstrates that small lodgepole pine were attacked by MPB at much greater rates (exceeding 50% at DBH's of 13.0 cm, 11.0 cm and 12.5 cm for age classes 2, 3 and 4 respectively) than previously reported.This indicates that although MPB preferentially attacked the larger lodgepole pine they likely indiscriminately select hosts when their population densities are extreme [11] and no other host choices exist.

Post Beetle Stand Dynamics of Immature Stands
Extensive attacks in young stands demonstrate the need for inventory data collected at layers below the canopy level.Highly variable attacks at stand levels also demonstrate the need to collect inventory data on a stand-by-stand basis.For age class 1 stands, except in a few instances, there are no management concerns about their growth and future productivity, unless there is another MPB outbreak within the next 20 years (Figure 1).
Considering residual mature trees (stem ≥ 7.5 cm DBH), most of the MPB-attacked stands met minimum stocking levels of 600 stems ha −1 for all age classes and ecological subzones which is comparable to Hawkins et al. [17] study.According to the BC Ministry of Forests and Range to contribute to mid-term timber supply, [30] the minimum stocking level for residual mature trees in MPB-attacked stands is 600 stems ha −1 .Like other studies, density of advanced regeneration (stem > 4 m height but <7.5 cm DBH) was found to be variable across age classes and ecological subzones [17,[31][32][33].
Depending on stand age, 65%-77% of stands can be considered stocked with advanced regeneration and live residual trees (secondary stand structure) after MPB attack.A stocked stand is expected to provide at least 150 m 3 ha −1 of timber at harvest [19].Based on a study by Coates et al. [31], around 25% to 57% of stands had enough secondary structure after MPB-attack in the Cariboo-Chilcotin region of BC to be considered stocked.This study also suggested that about 23%-35% stands within these age classes (age classes 2 to 4) required further attention to achieve the mid-term timber supply goal.This is comparable to a study by Coates et al. [31] in central BC where they suggested a single blanket management approach may not be possible for MPB affected stands.Severely impacted stands will have to be surveyed on a stand-by-stand basis.Only those stands which do not meet the mid-term timber supply minimum volume will need rehabilitation for timber management.Moreover, these attacked stands have also started to release within three to five years after MPB attack [20] and grow at rates proportional to trees that have never been suppressed [34,35].Therefore, retention of stands with suitable secondary structure can shorten rotation age greatly [31] compared to starting a new plantation after logging.Additionally, by retaining stands with suitable secondary stand structure, structural and species diversity will be promoted [20,[35][36][37] and future stand dynamics and age class distribution will be controlled [38].Moreover, the forest landscape will convert to heterogeneous structures where mosaics of even-aged and uneven-aged patches intermingle in space and time [39,40].This heterogeneous natural system is more resilient than homogeneous systems, as a significant portion of the biological legacies of that particular ecosystem remain intact [41][42][43] and this allows it to "remember" genetically, compositionally, and structurally pre-disturbance conditions of the stand to build a new complex ecosystem [43,44] which could be more adaptable to a changing climate [41,45].Therefore, stands that have arisen after MPB-attack may have enhanced resilience to both biotic and abiotic stressors, as well as could significantly reduce susceptibility to future MPB infestation [43,46].
When managing MPB-attacked stands it is important to set priorities with respect to which stands should be visited first.This can be done by considering two important stand characteristics: geographic location and stand age.Geographic location is important because some regions are more affected by MPB than others [47].For example, in this study, the moist subzone was more severely impacted by MPB than the other subzones.Stand age is also important because the level of MPB attack and tree diameter generally increase with stand age [7] as was demonstrated in the present study.Older stands generally have larger diameter pine trees and are more susceptible to MPB attack than younger stands.

Conclusions
About 5% of trees less than 10 cm in DBH were attacked and attack rates greater than 80% were observed in trees with a DBH ≥ 20 cm.Beetle attack rates significantly relate to stand densities of older age classes (2 and 3) while the moist subzone is more severely attacked than the other subzones.Clearly there will be some immature stands that are best logged for biomass and planted today to achieve future timber supply.However it appears a majority of the stands may be best left alone after MPB-attack, allowing them to make contributions to the mid-term timber supply albeit, in some cases, at reduced yields, as well as maintain long-term ecological benefits and services for human well-being [17,36,48,49].Additionally, the occurrence of a MPB outbreaks can result in more structurally and compositionally diverse stands [36,43,[48][49][50][51] which should significantly contribute to forest resilience against predicted climate change-related disturbances (i.e., future MPB attack).In addition, minimizing management intervention in these stands would allow the money allocated to stand establishment activities to be utilized for other needed management activities to increase quality or volume.Moreover, these stands are not only important for ecological restoration, they also provide a unique opportunity to reshape our knowledge regarding regeneration and stand dynamics under complex conditions following a MPB outbreak [41,39].management.We are grateful to five anonymous reviewers for their constructive feedback on the manuscript.

Table 1 .
Number of stands sampled and temporary sample plots (TSP) established by age class and biogeoclimatic/ecological subzone.

Table 2 .
Linear regression of mean MPB-attack in lodgepole pine as a function of stand mean pine DBH (=X) and MPB attack as a function of stand mean stems ha −1 (=X) for each age class.

Table 3 .
Mean percentage of lodgepole pine attacked by MPB and mean lodgepole pine DBH by age class and attack category.

Table 4 .
ANOVA results for the density of residual mature trees and advanced regeneration following MPB-attack based on age class and biogeoclimatic/ecological subzone.

Table 5 )
. Dry and moist ecological subzones of age class 3 (41-60 years) and the moist subzone of age class 2 (21-40 years) had >1000 stems ha −1 .Tukey's test for multiple comparisons showed age class 3 and 4 were significantly