https://orcid.org/0000-0001-8715-7825
Abstract
Weeding (commonly referred to as release) is a common practice in major timber-producing regions of the United States, yet the effects have not been well examined in recently established plantations in northern Idaho. This study tested the effects of selective postplanting forb control with clopyralid herbicide applied 1 year after planting on the growth and survival of Douglas-fir (Pseudotsuga menziesii var. glauca) and western larch (Larix occidentalis) for 5 years across a site productivity gradient in northern Idaho. Forb cover was reduced in Douglas-fir plots but not western larch plots. The result was an increase in diameter, height, and stem volume of Douglas-fir seedlings 5 years after treatment at low and high productivity sites, but no effect on western larch. Survival of both species was unaffected by the treatment and remained high, likely because of the generally high productivity of all sites. Results suggest that postplanting forb control with clopyralid may be best suited to Douglas-fir plantations in the region. The early gains in size are likely to persist into the future given the divergent growth trajectories observed, potentially shortening rotations and increasing final stand volume production.
Study Implications: Clopyralid is an effective tool for selectively controlling forbs after planting in recently established plantations in northern Idaho. The response was species-specific. Results showed Douglas-fir had positive gains in size 5 years after treatment, whereas western larch growth was unaffected. Results can help guide operational decisions regarding whether and in which situations clopyralid may be applied to accelerate stand growth by reducing postplanting competition.
Silvicultural practices shortly after planting seedlings are critical aspects of the reforestation pipeline, yet few techniques are widely applied across the United States besides precommercial thinning (Fargione et al. 2021). Treatment options for vegetation management are limited to mechanical (e.g., cutting aboveground biomass and root excavation) or chemical competition control (e.g., directed or broadcast application), broadly referred to as weeding or release (Ashton and Kelty 2018). The primary objective of weeding is to shift resource capture from competing vegetation to desired tree regeneration with an expected outcome of increasing survival and growth (Walstad and Kuch 1987). Chemical weeding is widely practiced in operational conifer plantations in the Pacific Northwest and Southeast United States, where treatments can reduce rotation lengths by accelerating tree growth (Newton and Cole 2008; Rose, Rosner, and Ketchum 2006; South et al. 2006). Chemical weeding of individual stem treatments is also used in hardwood systems in the United States. Comparatively, less is known regarding whether similar benefits are realized in the conifer-dominated forests of northern Idaho, where sites are less productive and rotations are longer than in the Pacific Northwest.
Highly fecund nontree vegetation can rapidly establish in open-growing conditions of young forest plantations (Dinger and Rose 2010). The result is often competition for water, nutrients, and light that can reduce growth and survival of tree seedlings (Nambiar and Sands 1993). Thus, numerous studies have shown that weeding improves the growth and survival of young seedlings (Stewart, Gross, and Honkala1984), often with long-term gains in yield (Wagner et al. 2006). The COMP study in the southeastern United States found substantial increases in the growth of loblolly pine (Pinus taeda) stands after 15 years in response to control of woody and herbaceous vegetation for the first 3 to 5 years (Miller et al. 2003). In the Garden of Eden study in California, Powers and Ferrell (1996) found that repeated weeding in young plantations of ponderosa pine (Pinus ponderosa var. ponderosa) increased growth within the first 6 years. An experiment in western Oregon examined repeated and delayed weeding treatments in Douglas-fir (Pseudotsuga menziesii var. menziesii) plantations over a 5-year period and found delaying treatment by just 1 year resulted in a 15% reduction in potential growth compared with repeated weeding (Maguire et al. 2009). Although weeding often increases growth, there can be trade-offs with mortality. For example, Simard and Vyse (2006) found that long-term gains in growth following weeding in British Columbia were offset by increased mortality due to the loss of ameliorative effects of nontree vegetation.
Different plant lifeforms have unique strategies for resource capture, resulting in varying ecological niches on recently disturbed forest sites (Radosevich, Holt, and Ghersa 2007). These different competitive strategies present numerous challenges to young plantations as the type and intensity of competition varies depending on the stage of stand development (Balandier et al. 2006). Grasses often start growing earlier in the growing season than trees or other woody vegetation, allowing grasses to capture soil moisture and nutrients before other plants and develop thick root mats near the soil surface (McDonald 1986). Shrubs compete for both moisture and nutrients, and tall shrubs also compete for light. Fast growing annual and biannual forbs can compete with tree seedlings for soil water and nutrients. Fine tuning silvicultural treatments to target specific lifeforms or species that are most competitive is beneficial from an integrated forest vegetation management perspective, especially with advances in selective chemicals (Wagner 1994). This may help minimize effects on less competitive vegetation and targeting control of nonnative species while minimizing damage to native vegetation (Miller and Miller 2004).
Western larch (Larix occidentalis) is a common plantation species in the Inland Northwest but is particularly sensitive to most forestry herbicides. The lack of a waxy cuticle on their needles coupled with a tendency to flush new foliage in early spring makes western larch very sensitive to herbicides compared with other species, such as Douglas-fir (Parent, Mahoney, and Barkley 2010). Clopyralid is one of the few available chemicals that western larch can tolerate. In addition to physical traits that may confer tolerance, it is possible that physiological traits also assist with clopyralid tolerance in western larch. Herbicide tolerance can be influenced by rates of chemical uptake and translocation, metabolism, or ability to detoxify the chemical (Warwick 1991). Needle curling of conifer foliage formed shortly after application that disappears over time further suggests physiological tolerance (Dixon, Clay, and Willoughby 2005). Clopyralid narrowly targets some forbs, including species in the Asteraceae family that commonly establish postplanting, many of which are nonnative, and often noxious, forbs. Examples include Canada thistle (Cirsium arvense), bull thistle (Cirsium vulgare), oxeye daisy (Leucanthemum vulgare), prickly lettuce (Lactuca serriola), and hawkweed species (Hieracium spp.) (Kelpsas and Landgren 2023). Clopyralid also targets legumes and other broadleaf genera (Shaner 2014). Comparatively, grasses are very tolerant of clopyralid (Shaner 2014). Currently, little research exists in the Inland Northwest on the effects of postplanting clopyralid application on growth and survival of tree seedlings across a site productivity gradient.
This study examines the effects of a postplanting clopyralid application at the beginning of the second growing season in plantations of western larch and interior Douglas-fir (Pseudotsuga menziesii var. glauca) across a range of site productivity classes in northern Idaho. The primary objectives were to examine (1) change in vegetation cover by life form following treatment and (2) the effects of treatment and site productivity on the survival and growth of western larch and Douglas-fir seedlings through the fifth year after treatment.
Methods
Site Description
The study occurred in north-central Idaho on forest land owned and managed by PotlatchDeltic Corp. The climate has cool, wet winters and a defined period of warming and dry conditions that typically spans July through September (Abatzoglou, Rupp, and Mote 2014). Forest soils in northern Idaho are commonly overlain with a volcanic ash mantle, resulting primarily from the eruption of Mount Mazama approximately 7,600 years ago. The high-water holding capacity of these ashy soils and the common occurrence of windblown loess deposits contribute to moisture retention in forest soils during the growing season with limited precipitation (McDaniel and Wilson 2007).
Eleven sites were used for this study (figure 1); elevation ranged from 877 m to 1116 m, and slopes ranged from 11° to 30° (Table 1). Average annual precipitation ranged from 77.4 cm to 111.8 cm. Habitat series of the sites that define productivity of forests in the region (Cooper, Neiman, and Roberts 1991) included grand fir (Abies grandis), western redcedar (Thuja plicata), and western hemlock (Tsuga heterophylla). All sites had an ash mantle or soil was overlain with loess.
Description of the eleven stands by productivity class (low, moderate, and high). Parent material was obtained from Web Soil Survey (Soil Survey Staff 2023). Precipitation (Precip), mean annual temperature (Tmean), and maximum vapor pressure deficit during the annual summer drought between June and September (VPD) were obtained from Prism Climate Group 30-year norms (1991–2020) at 800 m resolution (Prism Climate Group 2023).
| Stand | Planted species | Parent material | Elevation (m) | Slope (degree) | Precip (cm) | Tmean (°C) | VPD (hPa) |
|---|---|---|---|---|---|---|---|
| High 1 | Douglas-fir Western larch | Volcanic ash over loess over colluvium derived from basalt | 1008 | 16 | 95.8 | 7.9 | 23.4 |
| High 2 | Western larch | Volcanic ash over loess | 912 | 19 | 102.2 | 6.9 | 23.9 |
| High 3 | Douglas-fir | Volcanic ash over colluvium derived from gneiss and/or quartzite | 961 | 20 | 102.5 | 6.5 | 24.6 |
| High 4 | Douglas-fir Western larch | Volcanic ash over colluvium derived from schist | 1116 | 16 | 111.8 | 6.7 | 23.0 |
| Moderate 1 | Douglas-fir Western larch | Volcanic ash over colluvium derived from granite and/or gneiss | 996 | 30 | 110.5 | 6.5 | 23.4 |
| Moderate 2 | Douglas-fir Western larch | Volcanic ash over loess | 918 | 10 | 89.3 | 7.1 | 24.9 |
| Moderate 3 | Douglas-fir Western larch | Volcanic ash over loess (Fragipan 56–101 cm) | 877 | 11 | 105.4 | 11.9 | 21.9 |
| Low 1 | Douglas-fir Western larch | Volcanic ash over loess (Fragipan 79–117 cm) | 1029 | 25 | 90.6 | 7.4 | 24.6 |
| Low 2 | Douglas-fir Western larch | Volcanic ash and loess over colluvium derived from basalt (Fragipan 61–99 cm) | 949 | 20 | 92.3 | 6.8 | 21.9 |
| Low 3 | Western larch | Volcanic ash over loess (Fragipan 61–99 cm) | 906 | 13 | 77.4 | 7.9 | 25.2 |
| Low 4 | Douglas-fir | Volcanic ash over colluvium derived from gneiss and/or quartzite | 1035 | 21 | 93.9 | 7.3 | 25.1 |
| Stand | Planted species | Parent material | Elevation (m) | Slope (degree) | Precip (cm) | Tmean (°C) | VPD (hPa) |
|---|---|---|---|---|---|---|---|
| High 1 | Douglas-fir Western larch | Volcanic ash over loess over colluvium derived from basalt | 1008 | 16 | 95.8 | 7.9 | 23.4 |
| High 2 | Western larch | Volcanic ash over loess | 912 | 19 | 102.2 | 6.9 | 23.9 |
| High 3 | Douglas-fir | Volcanic ash over colluvium derived from gneiss and/or quartzite | 961 | 20 | 102.5 | 6.5 | 24.6 |
| High 4 | Douglas-fir Western larch | Volcanic ash over colluvium derived from schist | 1116 | 16 | 111.8 | 6.7 | 23.0 |
| Moderate 1 | Douglas-fir Western larch | Volcanic ash over colluvium derived from granite and/or gneiss | 996 | 30 | 110.5 | 6.5 | 23.4 |
| Moderate 2 | Douglas-fir Western larch | Volcanic ash over loess | 918 | 10 | 89.3 | 7.1 | 24.9 |
| Moderate 3 | Douglas-fir Western larch | Volcanic ash over loess (Fragipan 56–101 cm) | 877 | 11 | 105.4 | 11.9 | 21.9 |
| Low 1 | Douglas-fir Western larch | Volcanic ash over loess (Fragipan 79–117 cm) | 1029 | 25 | 90.6 | 7.4 | 24.6 |
| Low 2 | Douglas-fir Western larch | Volcanic ash and loess over colluvium derived from basalt (Fragipan 61–99 cm) | 949 | 20 | 92.3 | 6.8 | 21.9 |
| Low 3 | Western larch | Volcanic ash over loess (Fragipan 61–99 cm) | 906 | 13 | 77.4 | 7.9 | 25.2 |
| Low 4 | Douglas-fir | Volcanic ash over colluvium derived from gneiss and/or quartzite | 1035 | 21 | 93.9 | 7.3 | 25.1 |
Description of the eleven stands by productivity class (low, moderate, and high). Parent material was obtained from Web Soil Survey (Soil Survey Staff 2023). Precipitation (Precip), mean annual temperature (Tmean), and maximum vapor pressure deficit during the annual summer drought between June and September (VPD) were obtained from Prism Climate Group 30-year norms (1991–2020) at 800 m resolution (Prism Climate Group 2023).
| Stand | Planted species | Parent material | Elevation (m) | Slope (degree) | Precip (cm) | Tmean (°C) | VPD (hPa) |
|---|---|---|---|---|---|---|---|
| High 1 | Douglas-fir Western larch | Volcanic ash over loess over colluvium derived from basalt | 1008 | 16 | 95.8 | 7.9 | 23.4 |
| High 2 | Western larch | Volcanic ash over loess | 912 | 19 | 102.2 | 6.9 | 23.9 |
| High 3 | Douglas-fir | Volcanic ash over colluvium derived from gneiss and/or quartzite | 961 | 20 | 102.5 | 6.5 | 24.6 |
| High 4 | Douglas-fir Western larch | Volcanic ash over colluvium derived from schist | 1116 | 16 | 111.8 | 6.7 | 23.0 |
| Moderate 1 | Douglas-fir Western larch | Volcanic ash over colluvium derived from granite and/or gneiss | 996 | 30 | 110.5 | 6.5 | 23.4 |
| Moderate 2 | Douglas-fir Western larch | Volcanic ash over loess | 918 | 10 | 89.3 | 7.1 | 24.9 |
| Moderate 3 | Douglas-fir Western larch | Volcanic ash over loess (Fragipan 56–101 cm) | 877 | 11 | 105.4 | 11.9 | 21.9 |
| Low 1 | Douglas-fir Western larch | Volcanic ash over loess (Fragipan 79–117 cm) | 1029 | 25 | 90.6 | 7.4 | 24.6 |
| Low 2 | Douglas-fir Western larch | Volcanic ash and loess over colluvium derived from basalt (Fragipan 61–99 cm) | 949 | 20 | 92.3 | 6.8 | 21.9 |
| Low 3 | Western larch | Volcanic ash over loess (Fragipan 61–99 cm) | 906 | 13 | 77.4 | 7.9 | 25.2 |
| Low 4 | Douglas-fir | Volcanic ash over colluvium derived from gneiss and/or quartzite | 1035 | 21 | 93.9 | 7.3 | 25.1 |
| Stand | Planted species | Parent material | Elevation (m) | Slope (degree) | Precip (cm) | Tmean (°C) | VPD (hPa) |
|---|---|---|---|---|---|---|---|
| High 1 | Douglas-fir Western larch | Volcanic ash over loess over colluvium derived from basalt | 1008 | 16 | 95.8 | 7.9 | 23.4 |
| High 2 | Western larch | Volcanic ash over loess | 912 | 19 | 102.2 | 6.9 | 23.9 |
| High 3 | Douglas-fir | Volcanic ash over colluvium derived from gneiss and/or quartzite | 961 | 20 | 102.5 | 6.5 | 24.6 |
| High 4 | Douglas-fir Western larch | Volcanic ash over colluvium derived from schist | 1116 | 16 | 111.8 | 6.7 | 23.0 |
| Moderate 1 | Douglas-fir Western larch | Volcanic ash over colluvium derived from granite and/or gneiss | 996 | 30 | 110.5 | 6.5 | 23.4 |
| Moderate 2 | Douglas-fir Western larch | Volcanic ash over loess | 918 | 10 | 89.3 | 7.1 | 24.9 |
| Moderate 3 | Douglas-fir Western larch | Volcanic ash over loess (Fragipan 56–101 cm) | 877 | 11 | 105.4 | 11.9 | 21.9 |
| Low 1 | Douglas-fir Western larch | Volcanic ash over loess (Fragipan 79–117 cm) | 1029 | 25 | 90.6 | 7.4 | 24.6 |
| Low 2 | Douglas-fir Western larch | Volcanic ash and loess over colluvium derived from basalt (Fragipan 61–99 cm) | 949 | 20 | 92.3 | 6.8 | 21.9 |
| Low 3 | Western larch | Volcanic ash over loess (Fragipan 61–99 cm) | 906 | 13 | 77.4 | 7.9 | 25.2 |
| Low 4 | Douglas-fir | Volcanic ash over colluvium derived from gneiss and/or quartzite | 1035 | 21 | 93.9 | 7.3 | 25.1 |
Locations of the eleven study sites in the Northern Rockies of north-central Idaho.
Sites were selected across a productivity gradient using modeled Douglas-fir site index defined by soil variables and climate (Kimsey, Moore, and McDaniel 2008). Productivity classes were defined by calculating site index for all PotlatchDeltic’s northern Idaho land base. Stands were then stratified into high-, moderate-, and low-productivity stands, where site index for the high productivity stands were greater than the third quartile of site index, the moderate productivity stands fell between the first and third quartile, and low productivity stands were less than the first quartile.
All stands that were planted by PotlatchDeltic in 2018 with western larch and Douglas-fir were identified in each productivity class. Stands were then randomly ordered and visited in the summer of 2018. The first three stands that fell between 914 and 1128 m and did not occur in areas prone to frost damage, ungulate browsing, or cow trampling on north or east aspect sites were selected for the study. Preference was given to stands planted with both species in single-species blocks, although some stands contained only one species (Table 1).
Experimental Design
All stands were treated with chemical site-preparation in summer or fall of 2017 or spring 2018 following complete clearcut harvesting. In the spring of 2018, stands were planted with 1-year-old 130 mL containerized seedlings (415C Styroblock, Beaver Plastics, Acheson, Canada) in single-species blocks on an approximate 3.05 m × 3.05 m spacing (1077 trees/ha).
Each stand had six plots per species, with each plot containing 36 seedling planting locations (6 × 6 seedlings) for a total of 216 seedling locations per species per stand. Half the plots were randomly assigned for postplanting selective forb removal with clopyralid, whereas the remaining three plots remained untreated. The treatment consisted of a broadcast application of Transline (clopyralid; Corteva Agriscience) at a rate of 1.12 L ha-1 (403 g acid equivalent clopyralid ha-1) targeting 93.5 L ha-1 total spray volume applied using the waving wand technique with backpack sprayers in early May 2019. The experimental design was a randomized complete block design with each stand within a productivity class serving as a replicate block.
Measurements
A one-row buffer of seedlings surrounded each measurement plot to avoid edge effect. Due to variability in planting location caused by stumps, rocks, and other unplantable microsites, measurement plots contained between sixteen and twenty-six seedlings. Seedling survival was recorded at the end of each growing season. Distinctions were made between dead and missing seedlings, where dead seedlings were recorded only if the seedling could be found. Because the cause of mortality could not be identified for missing seedlings, they were removed from the analysis. Signs of animal browse or other forms of damage were also recorded, and damaged seedlings were removed prior to analysis. Seedling height and diameter was measured shortly after planting and at the end of the growing season in late September through early November four times (1 year before treatment and 1, 2, and 5 years after treatment). Height was measured from the ground line to the base of the terminal bud to the nearest tenth of a centimeter. Stem diameter was measured with calipers approximately 5 cm above the soil surface to the nearest hundredth of a millimeter.
Competing vegetation was measured around each seedling using a square PVC frame (0.5 m2 before and 1 year after treatment and 1 m2 2 years after treatment to account for lateral crown and root spread). The percent cover of each vegetation life form was measured to the nearest 10% within the quadrat in July of each year. Some species were measured individually and then combined into groups based on life form (e.g., forb, grass, or shrub). Thistle species that are susceptible to clopyralid were measured by individual species, including bull thistle and Canada thistle. These two species were combined as total thistle. Based on this methodology, it was possible for total vegetation cover to exceed 100% when vegetation was vertically layered.
Analysis
Change in vegetation cover over time was analyzed separately by thistle and lifeform with linear mixed-effect analysis of variance (ANOVA). Fixed effects in the models were productivity class, treatment, year, and their interactions, whereas stand and plot within stand were random effects. Analyses were performed separately for the western larch and Douglas-fir plots.
Repeated measures ANOVA was used to examine the effects of productivity, treatment and time on survival, height, diameter, and stem volume index (SVI) of seedlings. The SVI was calculated assuming a parabolic stem using the equationt presented in Harrington (2006):
where diameter is in millimeters and height is in centimeters.
Linear mixed effects models were fit using the nlme package version 3.1–149 (Pinheiro et al. 2023) in R (R Core Team 2024). Post hoc contrasts of estimated marginal means were used to examine the effects of the main factors and their interaction using the emmeans package (Lenth 2024). Analysis of competition cover and seedling size were analyzed at the population level by averaging values for all observations within a plot. Seedling survival was calculated as the proportion of seedlings alive at the time of observation within a plot. All models were checked for normality and heterogenous variance. Variance was not heterogeneous, so natural logarithm transformations were applied to cover and seedling size dependent variables. Survival did not require transformation. All plots in the moderate 2 stand were accidentally treated with clopyralid, so the site was removed from analysis. In addition, the western larch plots in the moderate 1 stand were heavily damaged by late season frost at the start of the second growing season and were excluded. This only left one moderate productivity stand for western larch, so only the low and high productivity stands were included in the western larch analyses. Significance in all models was assessed at the p ≤ .05 level.
Results
Change in Vegetation Cover
Prior to treatment, vegetation cover was similar between the treated and untreated plots across the three site productivity classes, including thistle (p ≥ .991), total forbs (p ≥ .434), grasses (p ≥ .830), and shrubs (p ≥ .993) (figures 2 and 3). Thistle cover in the western larch plots was lower in the treated plots than in the untreated plots except for the second year after treatment at the low productivity sites. Similarly, thistle cover was significantly lower in the treated than the untreated Douglas-fir plots across all productivity classes through the second year after treatment. The first year after treatment, forb cover was 18.4% (p = .024) and 16.9% (p = .033) lower in the treated than the untreated plots in the high and low productivity Douglas-fir plots, respectively. Although forb cover was 13.1% lower on average in the western larch treated plots than the untreated plots, differences were not significant (p ≥ .523). Grass and shrub cover remained similar between the treated and untreated Douglas-fir plots for all years, except grass cover was greater in the treated plots in the moderate productivity class. Grass and shrub cover did not differ between treated and untreated western larch plots through the second year posttreatment.
Cover of thistles (bull thistle and Canada thistle), total forbs, grasses, and shrubs in the Douglas-fir plots across a site productivity gradient (low, moderate, and high) prior to treatment (1-Yr Pre Trt) and for 2 years after treatment (1 year post treatment (1-Yr Post Trt) and 2 years post treatment (2-Yr Post Trt)) with selective chemical control of forbs the first year after planting. Asterisks indicate significant differences in cover between untreated and treated plots within a given year at p ≤ .05.
Cover of thistles (bull thistle and Canada thistle), total forbs, grasses, and shrubs in the western larch plots across a site productivity gradient (low, moderate, and high) prior to treatment (1-Yr Pre Trt) and for 2 years after treatment (1 year posttreatment (1-Yr Post-Trt) and 2 years posttreatment (2-Yr Post-Trt)) with selective chemical control of forbs 1 year after planting. Asterisks indicate significant differences in cover between untreated and treated plots within a given year at p ≤ .05.
Seedling Survival and Size
Survival averaged 75% or greater for both species across all three productivity classes and across all the years of measurements (figure 4). Survival did not differ among productivity classes or between treatments (p ≥ .479).
Population survival of Douglas-fir and western larch seedlings through 2 years following postplanting herbicide treatment in the low, moderate, and high productivity classes. The bars represent the mean values whereas the error bars represent one standard error. No differences were detected at p ≤ .05.
Douglas-fir seedling size did not differ between treatments for the first 2 years following treatment, but differences were observed 5 years after treatment in the low and high productivity classes (figure 5). Larger base diameter in response to treatment was only observed in the high productivity class, where diameter was 3.4 mm greater 2 years after treatment and 7.8 mm greater 5 years after treatment. Height was greater with treatment in the low and high productivity classes 5 years after treatment. Height was 0.22 m greater in the low productivity class and 0.32 m in the high productivity class. Stem volume was 0.17 dm3 and 0.61 dm3 greater in the treated plots than the untreated plots in the low and high productivity classes, respectively, by 5 years after treatment.
Change in Douglas-fir stem base diameter, height, and stem volume from the time of planting through the fifth year following postplanting herbicide treatment in the low, moderate, and high productivity classes. The bars represent the mean values and the error bars represent one standard error. Asterisks at each time point indicate significant differences between treatments at p ≤ .05.
All metrics of western larch seedling size did not differ between treatments across the three productivity classes (figure 6). In addition, the variation between stands was much greater than for Douglas-fir, as indicated by the error bars (figure 6).
Change in western larch stem base diameter, height, and stem volume from the time of planting through the fifth year following postplanting herbicide treatment in the low and high productivity classes. The bars represent the mean values whereas the error bars represent one standard error. No differences between treatments were detected at p ≤ .05.
Discussion
The postplanting clopyralid herbicide treatment resulted in lower total forb cover the first year after treatment, although the response varied by site productivity class and tree species. Forb cover was only significantly lower the first year following treatment in the low and high productivity classes in the Douglas-fir plots. By 5 years after treatment, Douglas-fir seedlings in the treated plots were larger for all size metrics in these productivity classes. Comparatively, forb cover was not different between treated and untreated western larch plots following treatment, and the size of seedlings did not differ by treatment. Treatments did effectively reduce the cover of thistles, a genus of forb species that are controlled by clopyralid.
Forb cover was 11.1% and 6.8% lower the first year after treatment than before treatment in the Douglas-fir treated plots in the low and high productivity classes, whereas forb cover increased during the same period in the western larch treated plots. This resulted in significantly lower forb cover in the Douglas-fir plots the first year after treatment. The different responses of forbs by tree species suggests western larch plots may have (1) contained forb either not targeted by clopyralid, (2) had a greater seed bank of forbs that germinated after treatment, (3) been invaded by seed dispersed from outside the plot area, (4) experienced inconsistencies in herbicide application, and/or (5) had greater variability in forb cover among plots and sites. Common mullein (Verbascum thapsus), houndstongue (Cynoglossum officinale), and St. John’s wort (Hypericum perforatum) are common forbs in recently established plantations in the Inland Northwest (Oester and Fitzgerald 2016) but are not controlled by clopyralid. Unfortunately, the cover of individual forb species other than thistles were not measured for all species over the 3 years of observation to confirm whether these species dominated forb cover, but the western larch plots could have been dominated by such species. In addition, many of the common forb species can remain viable in the seed bank for many years (Burnside et al. 1996), with common mullein remaining viable for up to 100 years (Kivilaan and Bandurski 1981). Seed of other species, such as Canada thistle, can disperse up to 11.35 m via wind (Sheldon and Burrows 1973), whereas prickly lettuce can disperse up to 43 km via wind (Lu, Baker, and Preston 2007). The significantly lower cover of thistles following treatment suggests treatments effectively controlled species targeted by clopyralid, but recovery of other species besides thistles cannot be confirmed.
Nontree vegetation can quickly develop following stand-replacing disturbance and subsequent plantation establishment. Removing this vegetation, even if only selectively removing some species, leaves room for other vegetation to occupy the growing space (Balandier et al. 2006). Vegetation can quickly rebound, often with a positive effect of favoring native vegetation and enhancing diversity (Miller and Miller 2004). Yet results from this study show that grasses and shrubs did not occupy the growing space created by the reduction of forbs in the Douglas-fir plots. The one exception was greater grass cover in the moderate productivity class. Grass cover was greater before treatment and continued to be greater following treatment at these sites. Comparatively, grasses and shrubs in the low and high productivity classes, where there was a treatment effect on forbs, remained similar between treatments, suggesting they did not occupy the growing space created within 2 years after selective forb control. In this region with a distinct growing season drought, lower competition cover could have increased resource availability for Douglas-fir seedlings (sensu Nambiar and Sands 1993).
Clopyralid treatment did not affect survival of western larch or Douglas-fir seedlings. Survival can be less sensitive to competition than growth for some species (Wagner 2000) because plant growth can slow when resources are limited while conserving reserves for survival. The effects of competition on survival are often site-specific and can more broadly be examined in relation to site productivity. Seedling survival tends to be lower on lower productivity sites when competition is present due to a smaller overall pool of available soil resources (Harper, Comeau, and Biring 2005; Schneider, Knowe, and Harrington 1998). Higher productivity sites have a larger pool of available soil resources, and thus seedlings can often overcome the pressures from competition. Controlling competition is still important on higher productivity sites, as vegetation can develop to high densities and thus push seedlings over a competition-induced tipping point (Rose, Ketchum, and Hanson 1999). The inability to detect a treatment effect on seedling survival in this study could be related to the overall site conditions. Even the low productivity class is considered suitable for timber production in northern Idaho. All sites were on mesic north or east aspects and had volcanic ash or loess surficial deposits that helps maintain soil moisture during the summer drought (McDaniel and Wilson 2007). To fully understand the effects of postplanting forb control across a broad site productivity gradient would require stands on south- and west-facing aspects with shallow to no ash/loess surficial deposits. Longer-term monitoring would also be worthwhile as Yildiz et al. (2011) found competition had no effect on early survival of coastal Douglas-fir followed by a delayed effect of greater mortality after 15 years.
Douglas-fir seedlings were larger after 5 years in plots that were treated in the low and high productivity classes. Diameter was only greater in the high productivity class starting the second year after treatment and persisted through the fifth year. Diameter growth is highly sensitive to competition (Wagner 2000) and forb cover remained 4% lower in treated versus untreated plots in the high productivity class compared with the low productivity class two years post-treatment. The magnitude of diameter response to competition can also vary depending on moisture availability. For example, Shovon, Gagnon, and Vanderwel (2021) examined the growth response of white spruce (Picea glauca) at the species’ northern range limit and found that the presence of competition strongly reduced diameter growth during a drought year. Increased diameter growth of Douglas-fir in the absence of competition was also observed in northern Idaho where grass cover was experimentally manipulated through seeding (Eissenstat and Mitchell 1983). In that study, larger diameter with lower competition was observed on both a southwest and northeast aspect. The high productivity class in our study likely had greater growing season soil moisture given it was selected based on a site index model that incorporated soil metrics including ash mantle depth.
Compared with diameter, height of Douglas-fir seedlings following weeding treatment was only greater in the fifth year after treatment, but the effect was observed in both the low and high productivity classes. Generally, height growth is insensitive to nonovertopping competition such as forbs (Zedaker, Burkhart, and Stage 1987), but height responses to competition control can still occur. The delay in height response was likely due to the determinant growth habit of Douglas-fir (Hermann and Lavender 1990). Species with a determinant growth habit develop leaf and shoot primordia in buds that determine potential growth the following year (Oliver and Larson 1996). Thus, the predetermined vertical and lateral growth potential of the Douglas-fir seedlings the first 1 to 2 years after treatment were developed the year before treatment, whereas the growth potential in later years would be in response to modified conditions of lower competition.
The larger diameter in the second year posttreatment in the high productivity class and greater height at year five in the low and high productivity classes resulted in Douglas-fir stem volume gains in the second and fifth year in the high productivity class and only year 5 in the low productivity class. It is possible that increased growth was due to lower competition for moisture from forbs. Cowden, Wightman, and Gonzalez-Benecke (2022) found that Senecio sylvaticus abundance was positively correlated with soil moisture depletion and increased water stress of Douglas-fir seedlings across a soil moisture gradient, which could lead to slower seedling growth. The gains in size of Douglas-fir in the present study are likely to persist. Wagner et al. (2006) conducted a review of multiple studies examining the effects of competition control on wood volume gains and found that gains in the Pacific Northwest ranged from 4% to 11,800%. In one of the only long-term studies to examine growth responses to early competition control in Idaho, Cherico et al. (2020) found interior Douglas-fir trees were 101.3 dm3 larger in stem volume at age 34 compared with no vegetation management. The result was a 2.7-year age shift in stem volume with vegetation management (i.e., 34-year-old trees were the same size as what would be expected for trees at age 36.7, thus potentially shortening the rotation to a target mean stem volume).
The considerable variability in western larch seedling size in both treatments resulted in no effect of the postplanting competition control. This was unexpected because it was assumed that due to the indeterminant growth of western larch, the species would exhibit an immediate response that would sustain for 5 years after treatment. Similar insensitivity of western larch seedlings to competition within the first 2 years after planting has been observed on other sites in northern Idaho (Pinto et al. 2018). Greater variability in stem volume of western larch than Douglas-fir was also observed by Reely and Nelson (2021) in relation to microsite and aspect, suggesting the species can be highly sensitive to microsite soil and vegetation conditions, and requires additional investigation to identify treatments and planting-spot selection that maximize growth.
Conclusion
Determining whether to selectively control forbs postplanting in recently established northern Idaho plantations is often financially driven by whether the treatment will produce significant increases in survival and growth and thus shorten rotations and increase yield at the end of the rotation. Results from this study show that across a site productivity gradient of timber-producing sites, seedling survival was unaffected for both species, and only Douglas-fir size was positively affected by clopyralid application. One of the main benefits of this herbicide is the ability to apply a broadcast application over western larch without considerable damage, but the lack of response by this species suggests efforts may be better diverted towards Douglas-fir plantations. Other chemicals are available that can be applied over recently planted Douglas-fir, especially when trees are still dormant (Kelpsas and Landgren 2023). These chemicals can have a broader spectrum of species targeted, which may provide additional gains in growth due to lower competition for soil water.
Acknowledgments
We thank PotlatchDeltic Corp. for providing funding for this project and providing access to study sites. Specifically, Chance Brumley was instrumental by helping develop ideas for the project and identifying sites. Noah Olaff and Casey Flannigan assisted with plot establishment and initial measurements. Lauren King, Thomas Foltz, Jean Smith, and Daria Paxton assisted with posttreatment measurements.
Funding
Funding for this project was provided by PotlatchDeltic Corporation.
Conflict of Interest
PotlatchDeltic Corporation funded this research project and supported Joshua Mullane for his Masters of Science degree.
Data Availability
The data underlying this article cannot be shared publicly due to its propriety nature of containing information specific to the private landowners landbase.
Literature Cited
Author notes
Current affiliation of Joshua A. Mullane: Nez Perce-Clearwater National Forest, USDA Forest Service, 1700 ID 6, Potlatch, ID, 83855, USA
