Forestry
An International Journal of Forest Research
Forestry 2014; 87, 85 – 97, doi:10.1093/forestry/cpt028
Advance Access publication 29 August 2013
Predicting height growth of sugar pine regeneration using stand
and individual tree characteristics
N. Angell1,2, K.M. Waring1* and T.A. Graves1,3
1
School of Forestry, Northern Arizona University, PO Box 15018, School of Forestry, Flagstaff, AZ 86011, USA
Present address: Arapaho Roosevelt National Forest, United States Forest Service, Idaho Springs, CO 80452, USA
3
Present address: Department of Fish, Wildlife, and Conservation, Colorado State University, Fort Collins, CO 80523, USA
2
Received 19 January 2013
Concern exists among managers and researchers that sugar pine (Pinus lambertiana), a valuable, moderately
shade-tolerant timber species, regeneration appears to be declining. Management and restoration require understanding factors leading to sustained sugar pine regeneration growth and overstorey recruitment. The primary research objective was to identify factors influencing sugar pine regeneration height growth. Data were collected on
sugar pine regeneration, including height growth and stand characteristics across six managed and eight unmanaged stands in the Lake Tahoe Basin, CA and NV, USA. Individual tree- and stand-level analyses were conducted
using non-parametric statistical comparisons and regression. Results indicated low mean height growth rates
and no relationship between canopy closure and either height growth or management history. Individual sugar
pine seedlings grew significantly taller under unmanaged stand conditions with higher canopy closures while
sapling growth did not differ statistically by management history. Individual tree-level height growth models
never explained more than 35 per cent of the variation. Stand-level models explained over 50 per cent of the variation with fewer variables than the individual tree-level models. More research should be conducted to determine
whether the regeneration that is persisting in the understorey would respond positively to more aggressive unevenaged silvicultural treatments designed for enhancing understorey pine growth.
Introduction
Successful establishment and growth of regeneration in forest
stands largely depends on stand structure (i.e. species composition, vertical and horizontal patterns) and site factors (i.e. soil
type, site index, available water and nutrients (e.g. Coomes and
Grubb, 2000 and references within). Stand structure influences
available resources necessary for successful establishment and
growth of regeneration (Oliver and Larson, 1996). In particular,
light can be a critical limiting factor for seedling growth and may
be more important than available soil resources in many locations
(Pearson, 1930; Oliver and Larson, 1996; van Pelt and Franklin,
2000). Several studies have found that soil resources, such as available water, limit seedling growth, particularly in the western United
States (e.g. Pearson, 1930; Royce and Barbour, 2001). Additionally,
complex competitive interactions in mixed species stands may
result in difficulty discerning the relative importance of different
factors related to understorey growth and development (Larson,
1963; Zald et al., 2008).
Abiotic site factors, such as soil resources, can greatly affect tree
growth and overall forest productivity (Smith et al., 1997).Poor sites
may have greater light availability but fewer soil resources, leading
to slow understorey growth rates (Chapin et al., 1987; Kranabetter
and Simard, 2008). Some pine species allocate more resources to
belowground than aboveground growth during the first several
years (Lenart, 1934; Fowells and Schubert, 1956; Larson, 1963;
Pharis, 1966); this occurs more frequently on lower quality sites
where long tap roots may be needed to access available soil moisture and or nutrients. Thus, site factors may have a greater impact
on the sustained growth of pine seedlings than that of other
species even though the more open conditions frequently found
on lower quality sites may be conducive to germination and
initial establishment (Pearson, 1930). Environmental factors,
such as soil water availability, temperature and precipitation,
also influence pine seedling survival and growth (North et al.,
2005a,b; Larson and Kipfmueller, 2010), likely due to the early
emphasis on belowground growth in these species.
Historically in Sierra Nevada mixed-conifer forests, frequent,
mixed-severity fire maintained open stands with groups of evenly
spaced large trees interspersed with groups of smaller trees
(Parsons and DeBenedetti, 1979; North et al., 2007). Historic
species composition was not dominated by the shade-tolerant
species frequently found in modern stands because fire created
openings and exposed mineral soil, generating conditions conducive to establishment of the relatively shade-intolerant pine
species (Zald et al., 2008). For the best survival and continued
growth of pine, these species may require some initial protective
shade followed by creation of more open conditions conducive to
rapid height growth (York et al., 2004; North et al., 2005b; Legras
et al., 2010).
# Institute of Chartered Foresters, 2013. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.
85
Downloaded from http://forestry.oxfordjournals.org/ at DigiTop USDA's Digital Desktop Library on March 14, 2014
*Corresponding author. Tel.: +011 19285234920; Fax: +011 19285231080; E-mail: kristen.waring@nau.edu
Forestry
Methods
Study area
This study was located in the TB, an 88 000 ha watershed found in the northcentral Sierra Nevada Range along a transition zone between the Mediterranean climate of California and the drier continental climate of the Great Basin,
Nevada (Figure 1). This transition results partly from the pronounced rain
shadow effect, with the east side of the TB receiving as little as one-third of
the precipitation as the west (50–150 cm, respectively). The majority of the
annual precipitation arrives as snow between December and March with
mean summer and winter temperatures averaging 30 and 268C,
respectively, at lake elevation (1900 m a.s.l.) (Manley et al., 2000). Most TB
soils are young Inceptisols or Entisols derived from granite, characterized
by a mixture of scree and rock outcrops on shallower soils and glacial till supporting deeper soils (USDA NRCS, 2007). Volcanic rocks overlie the granite on
the northern shores and in a few small pockets in other localities (Elliott-Fisk
et al., 1996; USDA NRCS, 2007). Basalts and andesites comprise parent materials in these areas (Elliott-Fisk et al., 1996; USDA NRCS, 2007). This large variation in soils sustains a broad range of vegetation (Manley et al., 2000).
86
Fourteen second-growth stands were selected from the mixed-conifer
forest of the lower montane vegetation zone (1396–2200 m) across
private, federal and state ownerships for sampling (Figure 1). Stands were
selected based on species composition and management history. Designated stands contained sugar pine regeneration (seedlings and/or saplings), overstorey sugar pine and at least three additional overstorey tree
species. The most common associates of sugar pine in the TB include
Jeffrey pine (P. jeffreyi Balf.), incense cedar (Calocedrus decurrens (Torr.)
Florin) and white fir (Abies concolor (Gord. & Glend.) Lindl.). At higher elevations (2134–2734 m), white fir is replaced by red fir and may be associated
with western white pine (P. monticola Dougl.) and lodgepole pine (P. contorta (Grev. & Balf.) Critch.) (Tappeiner, 1980). Green alder (Alnus tenuifolia)
and Scouler’s willow (Salix scouleriana Barr.) are dominant high elevation
riparian species found along stream banks and in meadows but are minor
components above 1680 m (USDA, 1981).
Six sampled stands were ‘managed’ since the Comstock era in the late
1800s (three prescribed burn, one thinned and prescribed burned, two
thinned and salvaged); the other eight were ‘unmanaged’ stands
(Table 1). The stands ranged in size from 2 to 30.4 ha and had a median
stand size of 4.9 ha. The managed stands varied widely in management
goals, although recent management had the following two main objectives: 1. fuel reduction to mitigate fire hazard and 2. salvage projects to
harvest red and white fir trees that were dead or dying from a fir engraver
(Scolytus ventralis Lec.) outbreak. In the two prescribed burned stands, all
the stems ,36 cm were thinned or girdled in the mid-1990s; the dead
stems and slash were piled and a fall understorey burn conducted two
years post-thinning. In the thinned and burned stand, residual basal area
decreased between 23 and 34 m2 ha21 with 6 m spacing between tree
boles in 2004; stems were piled and burned in the autumn of 2005 (Rich
Adams, personal communication, March 2010). The two thinned stands
underwent a salvage harvest in the mid-1990s for a fir engraver and Jeffrey
pine beetle (Dendroctonus jeffreyi Hopk.) outbreak on federal lands. Treatment consisted of tractor or hand-felling down to 34 m2 ha21 of residual
basal area; logs were removed by helicopter. Desired spacing between the
edges of the residual tree crowns was 1.5 m and species retention occurred
in this order of preference: Jeffrey pine/sugar pine, white fir and incense cedar.
Slash was lopped and scattered within the stand (USDA, 1994).
Field methodology
During the summers of 2008 and 2009, 10 plots were installed in each stand
(but 17 plots in D.L. Bliss State Park due to its large size) on a systematic grid
after randomly locating the first grid point. To avoid edge bias, a minimum
distance of 20 –30 m from stand edge was randomly selected to start grid
transects. Plot centres were placed at a minimum spacing of 30 m from
each other; actual spacing varied by stand size. Plots were circular
(12.62 m radius (0.20 ha) ‘overstorey plots’) with smaller nested plots
(5.64 m radius (0.01 ha) ‘understorey plots’) to assess regeneration (trees
less than 1.37 m tall and ,12.7 cm diameter at breast height (DBH)) of
all species. Plots falling on roads or in streams were randomly offset. Belt
transects (50 –150 m length by 11.28 m width) were installed in each
stand to study sugar pine regeneration; transects were placed on the
same grid and at the same azimuth as the circular overstorey and understorey plots. The transects were placed every 200 m beginning from a randomly selected starting point between 1 –50 m from the stand boundary.
At each overstorey plot centre, aspect, slope, elevation, disturbance
history and UTM coordinates were recorded. Species, tree status (live/
dead), strata and canopy class (Smith et al., 1997), and damage/defect
for all overstorey trees (≥12.7 cm DBH) falling within the overstorey plot
were also recorded. Overstorey trees were then sub-sampled by species
and size class, according to Forest Inventory and Analysis protocol (USDA,
2007) for the following additional characteristics: age, total height,
height to live crown base, crown width, sapwood and bark thickness.
Canopy cover was recorded at each plot centre and the four cardinal directions at 6.31 and 12.62 m from plot centre.
Downloaded from http://forestry.oxfordjournals.org/ at DigiTop USDA's Digital Desktop Library on March 14, 2014
Forest management practices (i.e. fire exclusion and suppression) have led to dense conditions favouring shade-tolerant individuals in many Sierran mixed-conifer stands (Parsons and
DeBenedetti, 1979; Ansley and Battles, 1998). Current stand structures are often comprised of mostly shade-tolerant species that
are able to establish and persist in dark understorey conditions.
The more shade-intolerant species cannot successfully compete
in these conditions (Ansley and Battles, 1998). The shade-tolerant
species regenerate more successfully even when treatments are
implemented, providing additional light resources and bare
mineral soil; this is likely due to the overstorey abundance and prolific seeding capabilities of shade-tolerant species (Grayet al., 2005;
Zald et al., 2008). Sugar pine (Pinus lambertiana Dougl), a moderately shade-tolerant and significant component of the California
mixed-conifer vegetation type, has declined in abundance due to
both forest management practices and the introduced pathogen
white pine blister rust (Kinloch and Scheuner, 1990; Ansley and
Battles, 1998; van Mantgem et al., 2004). The observed slow
growth rates and patchy nature of sugar pine decline indicate
that the application of restoration strategies to promote sugar
pine seedling establishment, growth and overstorey recruitment
in Sierran mixed-conifer forests could slow the decline (van
Mantgem et al., 2004; Maloney et al., 2011). Restoration to increase
proportions of sugar pine could be integrated with ongoing ecosystem management approaches in the Sierra Nevada (e.g. North
et al., 2009). Such approaches utilizing modified uneven-aged
management strategies are conducive to both restoration objectives and regeneration requirements of sugar pine. Increased survival and maintenance of sugar pine regeneration is requisite for
sustaining sugar pine indefinitely on the landscape.
This research focused on more clearly defining stand structures
that promote successful sugar pine regeneration growth and overstorey recruitment in the Lake Tahoe Basin (TB), CA and NV, USA.
Specifically, the following objectives were addressed: 1. to define
relationships between canopy closure, stand structure and sugar
pine height growth rates in stands with differing management histories; 2. to identify the most important stand, microsite and individual tree attributes explaining past growth rates and 3. to
develop models to predict stand structures that promote growth
at the individual seedling/sapling- and stand-levels.
Predicting height growth of sugar pine regeneration
Downloaded from http://forestry.oxfordjournals.org/ at DigiTop USDA's Digital Desktop Library on March 14, 2014
Figure 1 Location of the fourteen sampled stands of different management types in the Lake Tahoe Basin.
Per cent species composition, basal area, trees per hectare and stand
density index for each stand and by regeneration size class were calculated.
Site productivity was assessed for each stand using the following: 1. absolute site index values and 2. site index categories. Absolute site index was
calculated based on the average height and age of dominant and
co-dominant sugar pine (base age 50 years (Krumland and Eng, 2005)).
Site index categories were based on average height and age of dominant
and co-dominant trees of all species. Site index averages by species were
then used to classify each stand into a category: poor (,12.2 m), fair
(12.2–18.3 m) or good (.18.3 m) (base age 50 years (Krumland and Eng,
2005)). Predicted available soil moisture between 0 –15 and 15– 30 cm
for each stand was obtained from published soil survey data (Manley
et al., 2000).
To determine site tree ages and an approximate age distribution for each
stand, rings were counted on tree increment cores after mounting and
sanding. Sapling ring counts were validated using a microscope; overstorey
tree cores were scanned into a computer for ring width analysis using WinDendro (Regent Instruments Inc. 2007). The analysed cores were visually
cross-dated and validated using COFECHA (Holmes, 1983). On samples
without pith, age was estimated with the methodology of Applequist (1958).
Regeneration measurements
On understorey plots (all trees) and belt transects (sugar pine only), trees
were separated into the following regeneration height and DBH classes:
seedlings (,1.37 m height) and saplings (≥1.37 m and ,12.7 cm DBH).
87
Forestry
88
Table 1 Stand-level characteristics for fourteen sampled stands in the Lake Tahoe Basin
Stand
TRT
Age structure
SP1
PB
DLB
PB
SP2
EB
TP
PB
TH and
PB
TH and
SA
TH and
SA
UM
1 cohort w/
remnants
2 cohorts w/
remnants
1 cohort
1 cohort w/
remnants
1 cohort w/
remnants
1 cohort
VS
UM
MY
UM
GC
CL2
GR
CL1
UM
UM
UM
UM
SP3
UM
LC
SS
2 cohorts w/
remnants
2 cohorts w/
remnants
1 cohort w/
remnants
2 cohorts
3 cohorts
1 cohort
2 cohorts w/
remnants
1 cohort w/
remnants
Stand
size (ha)
Site index
class
Canopy
cover (%)
Live basal area
(m2 ha21)
7.3
M
44.4
35.4 (0.23)
41.1 (0.33)
284.2 (0.00)
520.5 (15.99)
677 (17.43)
30.4
M
39.0
44.7 (0.11)
60.9 (0.17)
389.7 (0.00)
791.4 (4.05)
760 (4.71)
7.7
2.4
M
M
61.6
36.0
59.3 (0.07)
42.4 (0.42)
68.4 (0.09)
45.2 (0.78)
632.5 (0.00)
164.1 (0.00)
5585.4 (80.41)
800.8 (10.03)
1474 (92.25)
619 (9.63)
4.5
L
40.9
27.5 (0.08)
29.6 (0.12)
210.2 (0.00)
480.5 (5.12)
474 (5.11)
5.3
L
45.6
41.4 (0.08)
43.3 (0.12)
376.3 (0.00)
930.9 (20.13)
709 (19.1)
3.6
L
27.9
14.3 (0.13)
19.1 (0.17)
108.1 (0.00)
420.4 (1.67)
247 (2.75)
21.0
L
33.7
22.5 (0.10)
23.9 (0.21)
260.3 (0.00)
700.7 (2.39)
439 (3.4)
4.9
M
40.0
41.8 (0.14)
51.4 (0.24)
412.3 (0.00)
1791.8 (6.67)
802 (8.05)
13.0
2.0
3.6
2.5
M
M
H
M
43.0
42.2
54.4
53.3
45.7 (0.06)
56.9 (0.42)
62.9 (0.22)
85.7 (0.23)
52.6 (0.17)
60.2 (0.46)
94.5 (0.31)
95.2 (0.34)
450.3 (0.00)
368.3 (0.00)
452.3 (0.00)
616.5 (0.00)
4444.5 (17.59)
1361.4 (10.12)
1281.3 (6.10)
2674.7 (2.7)
1115 (19.22)
950 (21.66)
1040 (6.65)
1437 (2.92)
5.7
M
61.6
91.3 (0.37)
108.7 (0.52)
670.6 (0.00)
3365.4 (8.32)
1563 (11.44)
Total basal area
(m2 ha21)
Overstorey
trees ha21
Understorey
trees ha21
SDI
Except stand size, numbers are means with standard error in parentheses. Values are based on individual tree measurements averaged at the plot level, except for site index, which is
averaged at stand level.
TRT ¼ treatment type; PB ¼ prescribed burn: mid-1990s; TH and SA ¼ thinned and salvaged: light understorey thin and salvage harvest in the mid-1990s; TH and PB ¼ basal area reduction:
2004 and prescribed burn 2005; UM ¼ unmanaged since harvesting during 1860 –1930s; age structure ¼ remnants are stems left prior to Comstock era harvesting between 1860s and
1930s; site index class ¼ H (high); M (moderate); L (low); see text for definitions (base age 50 (Krumland and Eng, 2005)); SDI ¼ stand density index was calculated using the summation
method (Long and Daniel, 1990); remnants were much older overstorey trees (widely scattered, few per acre) that were left during original harvesting in the 1800s.
Downloaded from http://forestry.oxfordjournals.org/ at DigiTop USDA's Digital Desktop Library on March 14, 2014
Predicting height growth of sugar pine regeneration
Statistical analysis
To account for potentially different influences on height growth patterns,
regeneration data were separated and analysed by height class: seedlings
and sapling. Individual tree-level comparisons were conducted after calculating overall summary statistics and pooling trees by management
(managed or unmanaged) and treatment type; this analysis allowed for
initial selection of important variables for inclusion in height growth prediction models (described later). Stand-level comparisons were made after the
data were appropriately scaled and summarized at the stand level; the
stands were also compared by management and treatment type (unmanaged, thinned and salvaged, thinned and prescribed burned, and prescribed
burned only). Data from the thinned and prescribed burned stand were
omitted from all treatment type comparisons due to the small sample
size (n ¼ 2 trees and 1 stand). Statistical comparisons were conducted
using the non-parametric Wilcoxon rank-sums to test for an effect of management and non-parametric Kruskal–Wallis (test statistic ¼ x2) tests followed by a post-hoc Dunn’s test (z statistic) to test for an effect of treatment
and, if a treatment effect was found, to discern which treatments were significantly different from each other. The significance level was set at a ¼
0.05 for all tests. Statistical comparisons were performed with the use of
JMP (Version 8) (SAS Institute Inc. 2007).
Height growth prediction models
Models of sugar pine regeneration height growth rates were fitted to test
the effect of environmental, stand-level, and individual tree measurements
on individual tree-level growth. Because of the expectation that samples
within a transect or stand might be spatially autocorrelated (i.e. not independent) due to the sampling design, transects and stands were modelled
as random effects in mixed-effect, linear regression models. All other variables were treated as fixed effects (Zuur et al., 2009).
The response variable was height growth averaged over the last 10 years
or the life of the seedling. The height growth was skewed (non-normal) so
the data were transformed using a natural logarithm. Only slight violations
in normality and homogeneity of variance assumptions remained in the
height growth after natural log transformations. The natural log of past
seedling height growth was modelled as a function of stand-level measurements (canopy cover (%), site index (m), stand density index and stand live
basal area (m2 ha21)), individual tree-level measurements (available soil
moisture between 0 –15 and 15– 30 cm of the soil profile, canopy closure
(per cent), live, dead and total basal area (m2 ha21), distance, transformed
azimuth (radians), species and status of closest shade tree and crown
surface area (m2)). Sapling models were similar but status of closest
shade tree was not used as a potential predictor because few shade trees
were observed. Potential predictors also included DBH, height-to-diameter
ratio and white pine blister rust (presence/absence). White pine blister rust
was not used as a variable in seedling models because few seedlings were
observed with confirmed white pine blister rust. Interactions between variables were tested in all tree-level models. Variance inflation factors were
calculated between variables to test for multicollinearity, and one variable
was omitted from the model if the factor was 10 or greater (Ott and Longnecker, 2010). Multiple linear mixed-effect, backwards stepwise regression
models were fitted with program R statistical software and the NLME
package (R Development Core Team, 2010). Variable retention was set at
P , 0.05.
The variables influencing height growth were modelled separately for
seedlings and saplings, while also separating the models using a management variable (‘management’ and ‘treatment’). To evaluate whether management or specific treatment type was an important determinant of past
height growth, height growth was assessed with three different models for
both seedlings and saplings, for a total of six models: 1. the hypothesized
growth model (that past height growth would be a function of canopy
closure, live basal area and crown surface area), 2. the management
model: using management as a variable, with all treatment types pooled
and each tree assigned to either ‘managed’ or ‘unmanaged’ and (3) the
treatment model: using treatment type as a variable, with treatments
separated into treatment type (i.e. unmanaged, prescribed burned or
thinned and prescribed burned) with each tree assigned to the appropriate
category. R program code for the six most complex models (prior to backwards stepwise reduction) can be found in the Supplementary data.
The natural log of sugar pine seedling height growth was modelled at
the stand level, but sapling growth was not modelled at the stand level
due to the lack of observations. Growth was first modelled using only our
hypothesized best predictors (canopy closure, live basal area and crown
surface area) and was then modelled as a function of the same predictors
as used in the individual tree-level models, using backwards stepwise regression to determine the best model, assuming independence between
stands. Due to the small stand sample size, variable interactions were not
included to avoid model overfitting. All variables were scaled appropriately
and averaged at the stand level.
Results
Sugar pine growth and stand attributes
Stand characteristics were summarized for all the fourteen
sampled stands (Table 1). The stands typically contained one or
two age cohorts with a few stands also having remnants from
pre-Euro-American settlement (prior to 1850). Canopy cover
ranged between 28 and 62 per cent, with no clear trends
between the managed and the unmanaged stands (Table 1).
Live and total basal areas varied considerably across stands and
management types, with the highest live basal area over six
89
Downloaded from http://forestry.oxfordjournals.org/ at DigiTop USDA's Digital Desktop Library on March 14, 2014
For each sugar pine seedling and sapling, the following characteristics were
also measured: age, height, height growth, live crown length, crown width
and DBH, presence of damage/defects and presence/absence of white pine
blister rust. White pine blister rust was classified as present if we found aecia
spores or two of the following characteristics: white pine blister rust canker
with resin, bark stripping, branch flagging or rodent chewing (Maloy, 2003).
Local competition around each seedling or sapling was described with per
cent canopy closure, live tree basal area, and distance, azimuth to and
species of closest shade tree. Shade trees were defined as the nearest
sapling-size or greater neighbour without respect to direction. All azimuths
were transformed prior to analysis from degrees to radians using the
method of Beers et al. (1966). Per cent canopy closure was measured
with a spherical densiometer above each seedling or sapling at the outermost tip of the branch in the four cardinal directions and was averaged.
Basal area per unit area was also recorded at each tree using a relaskop
(20 ft2 ac21 basal area factor, transformed to m2 ha21 for analysis).
Past ten-year height growth (cm year21) of the dominant leader (or
tallest branch) was measured using digital callipers or a height pole by
measuring insertion point of each branch whorl (saplings) or bud scar (seedlings), down the length of the stem starting with the year 2007. Measurements were then averaged over the last ten years for saplings, or for the
life of the seedling if ,10 years old, to estimate average annual height
growth rate. If a tree was over 2.7 m in height, increment growth was not
measured and the densiometer readings were taken from the outermost
branch in each cardinal direction. If the canopy of the sapling biased the
densiometer measurement, a ladder was used to take the measurement
above the canopy of the sample tree.
Crown length was measured to the height of the lowest live branch at
point of insertion. Crown diameter was calculated by averaging the
longest live branch lengths in the N– S and E– W directions; crown surface
area was then calculated using a modified equation from Zarnoch et al.
(2004), as used for Sierran conifers in Waring and O’Hara (2009). On saplings, an increment core on the north side of the tree base was extracted
to determine age using the techniques as described earlier.
Forestry
Table 2 Individual tree-level differences in microsite and sugar pine seedling and sapling characteristics (height growth and height) between 14
managed and unmanaged stands
Size class
MGD
n
Live basal area
(m2 ha21)
n
Total basal area
(m2 ha21)
n
Canopy closure
(%)
n
Height growth
(cm year21)
n
Total height
(m)
Seedlings
UM
M
UM
M
396
223
134
45
33.8 (0.65)*
29.5 (0.78)*
23.3 (1.15)*
36.2 (2.11)*
396
223
134
45
40.3 (0.80)*
37.3 (1.04)*
26.1 (1.16)*
46.9 (2.68)*
397
223
134
45
76.9 (0.98)*
70.3 (1.24)*
52.9 (2.10)*
79.1 (2.21)*
283
184
75
16
2.53 (0.06)*
1.80 (0.05)*
4.10 (0.19)
4.99 (0.56)
351
206
123
39
0.28 (0.017)*
0.13 (0.007)*
2.78 (0.10)*
3.64 (0.26)*
Saplings
times greater (92.1 m2 ha21) than the lowest (14.3 m2 ha21). For
total basal area, the highest value (109.46 m2 ha21) was over
5.5 times greater than the lowest (19.1 m2 ha21). The median
live and total basal areas were 42.0 and 51.6 m2 ha21, respectively. Understorey and overstorey trees per hectare also varied widely
between stands (overstorey ¼ 108.1–670.5 and understorey ¼
420.4– 5585.4), as did stand density index, ranging between 247
and 1563 (Table 1).
At the individual tree-level, sugar pine seedlings growing in the
unmanaged stands had significantly higher basal areas (live
(Z ¼ 24.52, P , 0.0001) and total (Z ¼ 22.59, P ¼ 0.0097),
canopy closure (Z ¼ 24.86, P , 0.0001), height growth
(Z ¼ 28.42, P , 0.0001) and total height (Z ¼ 25.55, P , 0.0001)
compared with seedlings growing in the managed stands
(Table 2). The converse is true for saplings, with those growing in
managed stand conditions having significantly higher mean
basal areas (live (Z ¼ 5.34, P , 0.0001) and total (Z ¼ 6.69, P ,
0.0001)), canopy closures (Z ¼ 6.39, P , 0.0001), and sapling
total height (Z ¼ 3.00, P ¼ 0.0027) than those growing in unmanaged stand conditions (Table 2). The mean sapling height growth
rate was also higher under managed than that under unmanaged
stand conditions, although the difference was not statistically significant (Z ¼ 1.53, P ¼ 0.1267; Table 2). There was no clear relationship between height growth of seedlings or saplings and canopy
closure (Figure 2). When separated by treatment type, seedlings
growing in unmanaged and prescribed burned stand conditions
had similar live and total basal areas and canopy closures but
had significantly greater live and total basal areas and canopy
closure than thinned and salvaged stand conditions (Table 3).
However, height growth rate means were significantly different
among all treatment types, with growth rate highest for seedlings
growing in unmanaged followed by prescribed burned, then
thinned and salvaged stand conditions. Seedling total height was
significantly greater in unmanaged than prescribed burned stand
conditions but not for thinned and salvaged stand conditions
(Table 3).
Fewer differences were found for saplings at the individual treelevel (Table 3). The saplings growing in the unmanaged stands had
significantly lower live and total basal areas than those in prescribed burned stands but were only statistically different from
those in thinned and salvaged stands for canopy closure
(Table 3). Mean total height was significantly higher for the saplings
growing in unmanaged than those in prescribed burned stand
90
Figure 2 Relationship between seedling and sapling height growth rates
and individual tree-level canopy closure.
conditions, but again not significantly different from those in
thinned and salvaged stand conditions.
Stand level comparisons revealed no significant differences in
any microsite variables or sugar pine characteristics between the
unmanaged and the managed stands (Table 4). We also found
no significant differences in any microsite variables or sugar pine
characteristics between treatment types; however, some categories had very small sample sizes and high variability (Table 4). Only
seedling total height and height growth were compared at the
stand level due to small sample sizes of saplings in several
stands. We also found no clear relationships between mean seedling height growth and canopy closure or live basal area when separated by treatment type (Figure 3).
Height growth models and regression analyses
Individual tree-level models
The best models predicting height growth of both seedlings and
saplings at the individual tree-level were complex, containing
stand and transect as random effects and multiple explanatory
variables and interactions as fixed effects (Table 5; management
Downloaded from http://forestry.oxfordjournals.org/ at DigiTop USDA's Digital Desktop Library on March 14, 2014
Numbers are means with standard error in parentheses. Wilcoxon rank-sums were used to test for differences between unmanaged and managed stands.
Values with an asterisk are significantly different at P , 0.05 level.
Seedlings (,1.37 m height and ,2.54 cm DBH); saplings (.1.37 m height and 2.54– 12.7 cm DBH); MGD ¼ management; M ¼ managed since initial harvesting between 1860s and 1930s; UM ¼ unmanaged since harvesting during 1860s– 1930s.
Predicting height growth of sugar pine regeneration
Table 3 Individual tree-level differences in microsite and sugar pine seedling and sapling characteristics (height growth and height) between the
fourteen stands with different management histories
Treatment
type
n
Live basal area
(m2 ha21)
n
Total basal area
(m2 ha21)
n
Canopy
closure (%)
n
Height growth
(cm year21)
n
Total height
(m)
Seedlings
UM
PB
SH
TH and PB
UM
PB
SH
TH and PB
396
160
61
2
134
36
7
2
33.8 (0.65) A
31.7 (1.03) A
24.7 (0.73) B
20.7 (2.3)
23.3 (1.15) A
37.9 (2.48) B
30.2 (3.59) AB
27.6 (4.59)
396
160
61
2
134
36
7
2
40.3 (0.80) A
42.4 (1.26) A
25.3 (0.70) B
36.7 (13.78)
26.1 (1.16) A
50.8 (2.97) B
30.8 (3.12) AB
34.4 (6.89)
397
160
61
2
134
36
7
2
76.9 (0.98) A
75.3 (1.31) A
60.4 (2.28) B
58.4 (28.86)
52.9 (2.10) A
81.3 (2.40) B
76.0 (3.07) B
49.3 (1.30)
283
127
55
2
75
13
3
0
2.53 (0.06) A
1.94 (0.07) B
1.49 (0.06) C
2.69 (0.24)
4.10 (0.19) A
4.95 (0.69) A
5.13 (0.41) A
–
351
147
57
2
123
30
7
2
0.28 (0.017) A
0.12 (0.01) B
0.13 (0.01) AB
0.39 (0.18)
2.78 (0.10) A
3.74 (0.31) B
2.97 (0.47) AB
4.57 (0.80)
Saplings
See Table 1 for treatment type and Table 2 for size class descriptions. Wilcoxon and Kruskal– Wallis rank-sums were used to test for significant differences
between different treatment types. Within each treatment type, values that are not connected by the same letter are significantly different at the P , 0.05
level. Values that share the same letter are not statistically significant. TH and PB not included in statistical analysis due to low sample size. Means (standard
error) shown; n ¼ number of observations.
Table 4 Stand-level differences in microsite and sugar pine seedling characteristics (height growth and height) between 14 stands
Management
UM
M
Treatment type
PB
SH
TH and PB
n
Live basal area (m2 ha21)
Total basal area (m2 ha21)
Canopy cover (%)
Height growth (cm year21)
Total height (m)
8
6
53.0 (8.5)
41.9 (9.0)
63.5 (10.3)
48.1 (11.1)
44.5 (3.8)
44.6 (4.2)
2.40 (0.16)
1.92 (0.22)
0.36 (0.07)
0.17 (0.08)
3
2
1
46.5 (13.8)
34.5 (16.9)
42.6 (23.9)
56.9 (16.8)
36.5 (20.5)
45.3 (29.0)
48.3 (6.3)
43.3 (7.7)
36.0 (10.9)
2.03 (0.27)
1.38 (0.33)
2.69 (0.46)
0.14 (0.12)
0.12 (0.14)
0.39 (0.20)
No significant differences were found between management and treatment type (P ≥ 0.05). Treatment type comparisons included unmanaged stands;
the thinned and prescribed burned stand (TH and PB) was not included in statistical comparisons (n ¼ 1). Means (standard error) shown; n ¼ number of
stands. See Table 1 for treatment type and Table 2 for size class descriptions.
models not shown). These complex models failed to explain more
than 30 per cent of the variation but included the majority of possible parameters.
Stand-level models
At the stand level, the hypothesized model, which included crown
surface area, live basal area and per cent canopy closure as predictors, was a similar predictor of average 10-year sugar pine seedling
height growth (R 2adj. ¼ 0.56) as the best model, which included
only live basal area and crown surface area (R 2adj. ¼ 0.53;
Table 6). Stand management or treatment history were not significant predictors, thus supporting the same final, best model regardless of stand history.
Discussion
Individual tree-level
At the individual tree-level, densities and canopy closures were
greater in the unmanaged stands as expected. No apparent relationship was found between microsite characteristics (basal area
or canopy closure) and height growth rates by treatment type; it
was expected that height growth would be greatest under the
most open conditions. The expectation that growth rates and
microsite characteristics would be influenced by treatment type
was not supported by the data. The greatest height growth rates
for seedlings were found in unmanaged stand conditions, which
also had the highest densities and canopy closures. Growth rates
and microsite values varied considerably within treatment types,
complicating interpretation. While it appears that sugar pine seedlings require a sheltered microclimate for growth and are less influenced initially by density and canopy closure, this result could be an
artefact of better site productivity leading to greater seedling
growth and stand density.
The greater height growth rates and total heights of seedlings
growing in unmanaged stand conditions were not anticipated although they support some findings in recent studies (Dulohery
et al., 2000; Gray et al., 2005; Legras et al., 2010). When separated
by management type, mean densities and canopy closures at the
seedling level were statistically similar for prescribed burned and
unmanaged stand conditions, which were statistically greater
than in thinned and salvaged stand conditions. These findings
were anticipated since the management objectives for the
91
Downloaded from http://forestry.oxfordjournals.org/ at DigiTop USDA's Digital Desktop Library on March 14, 2014
Size class
Forestry
prescribed burned stands were not aimed at reducing stand densities but rather at reduction of hazardous fuels. The significantly
greater seedling height growth rates under prescribed burned conditions as compared with thinned and salvaged stand conditions
were expected because prescribed fire decreases litter and duff,
which can in turn increase the amount of soil moisture available
for seedling growth (Zald et al., 2008; Legras et al., 2010). Greater
height growth rates were expected for seedlings under prescribed
burned conditions than in the unmanaged stand conditions, but
the results did not support this expectation.
For saplings, higher densities and canopy closures were found at
the individual tree level in managed stand conditions. This was unexpected and again may relate to site productivity or management
history. Sapling total heights and height growth rates were greater
in managed than unmanaged stand conditions, although the difference was not statistically significant for height growth rates.
The greater total heights and height growth rates observed
under prescribed burned stand conditions (as compared with unmanaged stand conditions) may be explained by the benefits of
fire, which reduces competition and litter depth and increases
availability of understorey light (Sugihara and McBride, 1996;
North et al., 2005b) but must be interpreted with consideration
of the small sample size (n ¼ 2). Height growth was greatest,
although not statistically significant, under thinned and salvaged
stand conditions, which had the lowest densities and canopy
closures. However, it is important to recognize that both thinned
and salvaged stands were on low quality sites, which may
explain why the difference in growth was not statistically significant. For sustained sapling growth and overstorey recruitment,
92
saplings should benefit from treatments reducing both overall
stand density and more focused canopy closure reduction
around individual saplings.
The best individual tree-level models explained less than half of
the variation, indicating that important variables, such as nutrient
availability, are likely missing from the predictive models. More
complex models were supported for seedling than sapling height
growth, suggesting that more variables influence seedling height
growth. The influence of genetics may also affect the ability to
predict height growth at the individual tree-level; environmental
influences, which tend to be more important than genetics in the
phenotypic expression of tree growth, may overshadow the influence of genetics at the stand level (White et al., 2008).
Significant variables in the seedling models included factors
linked to establishment and early growth, such as available soil
moisture, closest shade tree characteristics and crown surface
area. Sapling models included variables more closely tied to competition effects and expected patterns of height growth such as live
basal area, DBH and stand density index. However, several variables were significant in both seedling and sapling models, and
management or treatment type was important in most models.
Stand level
Mean stand level basal area and canopy cover in the managed
stands were higher than expected. Current management objectives and treatment intensities for the stands were not aimed at
substantially reducing density except for in the thinned and
burned stand. The large range in values for densities and canopy
Downloaded from http://forestry.oxfordjournals.org/ at DigiTop USDA's Digital Desktop Library on March 14, 2014
Figure 3 Relationship between average stand-level sugar pine seedling height growth rate (cm year21) and the microsite characteristics of (a) canopy
closure (%) and (b) live basal area (m2 ha21) by management type for 14 stands in the Lake Tahoe Basin.
Predicting height growth of sugar pine regeneration
Table 5 Results of multiple linear regression modelling to predict the natural log of past sugar pine height growth (cm year21) at the individual
tree-level in the Lake Tahoe Basin
Seedlings
Stand 14.88
Parameter
Transect 15.80
Estimate
Residuals 69.30
P value
Intercept
TPHA
CC
TRT
Site index
SPP1
SPP2
SPP3
SPP status
H2030
CSA
DI
BAT
AZ
0.262
,0.001
20.006
0.661
0.045
0.440
20.216
20.111
20.047
20.093
1.129
20.033
0.025
20.183
0.650
0.862
0.169
0.862
0.199
0.085
0.128
0.528
0.402
0.226
,0.001
0.014
0.001
0.007
DF 375
Parameter
TPHA*SPP1
TPHA*SPP2
TPHA*SPP3
TPHA*CSA
TRT*CC
TRT*SPP1
TRT*SPP2
TRT*SPP3
TRT*CSA
Site index*CC
Site index*BAT
SPP Status*CSA
H2030*CSA
Distance*AZ
BAT*AZ
Estimate
20.001
,0.001
,0.001
,0.001
20.006
0.372
20.035
,0.001
20.519
,0.001
20.002
20.289
20.078
0.034
0.004
P value
0.036
0.252
0.938
,0.001
0.003
0.008
0.731
0.999
,0.001
0.020
,0.001
0.004
0.040
0.003
0.007
% Variation explained
Stand ,1.00
Parameter
Transect 32.90
Estimate
Residuals 67.10
P value
Intercept
TPHA
CC
BAL
Site index
SPP1
SPP2
DBH
SPP status
H2030
CSA
DI
AZ
6.277
0.004
0.033
20.434
0.219
3.342
1.050
0.804
21.119
21.951
20.109
0.099
0.191
0.110
0.044
,0.001
0.001
0.037
0.001
0.134
,0.001
,0.001
0.101
0.081
,0.001
0.084
DF 31
Parameter
Estimate
P value
TPHA*CC
BAL* H2030
Site index*SPP1
Site index*SPP2
Site index*DBH
Site index*CSA
SPP1*CSA
SPP2*CSA
DBH*CSA
DBH*AZ
,20.001
0.092
20.258
20.031
20.046
0.014
20.006
20.053
20.009
20.055
,0.001
0.002
0.003
0.577
0.006
0.009
0.618
0.014
0.002
0.015
Best treatment models (treated or untreated) are presented for both seedlings and saplings. Stand and transect were included as random effects within the
mixed-effects models.
DF ¼ degrees of freedom; TPHA ¼ trees per hectare; BAL ¼ live basal area (m2 ha21); CC ¼ individual tree-level canopy closure (%); CSA ¼ crown surface
area of sugar pine seedling or sapling (m2); Site index (m, base age 50 years (Krumland and Eng 2005); TRT ¼ treatment history (treated or not); SPP ¼
species of nearest shade tree; DBH ¼ diameter at breast height (cm); SPP status ¼ nearest shade tree status (live or dead); DI ¼ distance (m) to nearest
shade tree; AZ ¼ direction to nearest shade tree; H2030 ¼ available soil moisture between 15– 30 cm.
cover for the unmanaged stands was also contrary to expectations. The unmanaged stands were expected to have higher
average densities and canopy cover than the managed stands.
A wide range in site productivity across stands was found, particularly in the unmanaged stands, which partially explains the wide
variation in density and canopy cover between the managed and
the unmanaged stands and within the unmanaged stands. The inability to accurately measure site productivity in these multi-aged
stands may contribute to the variability in our results. Additionally,
age structure differences in these stands may affect light regimes
and understorey growth rates. For example, many saplings were
quite old and suppressed (Angell, 2011). Accessible stands
meeting the selection criteria likely affected the results; many
stands with an abundance of sugar pine are on sites of poor productivity and low canopy cover. Time since management, which
varied from only a few to 15 years, was not accounted for and
can greatly impact canopy cover. In comparison to the other
studies conducted in California, seedlings and saplings grow
more slowly in the TB (Larsen and Woodbury, 1916; Fowells and
Schubert, 1956; Fowells and Stark, 1965; York et al., 2004). The TB
lies near the eastern edge of sugar pine’s range and may be marginal habitat. Additionally, natural sugar pine regeneration under
93
Downloaded from http://forestry.oxfordjournals.org/ at DigiTop USDA's Digital Desktop Library on March 14, 2014
Saplings
% Variation explained
Forestry
Table 6 Results of multiple linear regression modelling to predict the
natural log of past sugar pine seedling height growth (cm year21) at the
stand level across fourteen stands in the Lake Tahoe Basin
Parameter
1.091
0.013
20.010
0.209
0.126
0.010
0.238
P value
R2
R2 adj.
RMSE
DF
AIC
0.010
0.151
0.004
0.203
0.011
0.007
0.472
0.004
0.005
0.663
0.561
0.182
10
22.65
0.598
0.527
0.189
11
22.26
Backward stepwise regression resulted in the same best model for both
management and treatment history.
HYP ¼ hypothesis model; BAL ¼ live basal area (m2 ha21); CC ¼ average
canopy closure (%); CSA ¼ crown surface area of sugar pine seedling or
sapling (m2); R2 ¼ coefficient of multiple determination; R2 adj. ¼ coefficient of multiple determination adjusted for the number of explanatory
terms in the model; RMSE ¼ square root of the variance (standard error of
the regression); DF ¼ degrees of Freedom; AIC ¼ Akaike’s information
criterion (Burnham and Anderson, 2002).
completely open conditions was not found in the TB, but such conditions should produce higher growth rates.
Valid predictions of stand level height growth of seedlings can be
easily attained by managers using the models presented here and
with metrics relatively accessible to managers such as canopy
closure, live basal area and crown surface area. Of these, canopy
closure and live basal area can be easily calculated from standard
stand exam data. Crown surface area would require the most
effort, but data collection could be incorporated into regeneration
plots during stand exams, resulting in a straight-forward calculation (Zarnoch et al., 2004; Waring and O’Hara, 2009).
Factors affecting sugar pine regeneration
Sugar pine has been shown to establish well on sandy soils (Fowells
and Schubert, 1956), allocating its resources to a taproot with few
lateral root branches. The taproot may grow in length up to 61 cm
in 2 –3 months, averaging 43 cm in length (Fowells and Schubert,
1956).This is in contrast to the establishment of sugar pine germinating on duff-covered soils that had roots of an average of 18 –
23 cm and a maximum of 30.5 cm in length (Fowells and Schubert,
1956). Sugar pine allocates a large portion of the available
resources to root development in more open, resource poor sites
such as duff-covered soils (Fowells and Schubert, 1956). The granitic derived soils on many of the TB sites have less available soil
moisture (Hubbert et al., 2001; Witty et al., 2003), likely impeding
continued seedling growth on poor quality sites if resources are limiting and the bedrock is not friable for continued root development
(Witty et al., 2003).
Water is the primary limitation to productivity in many Mediterranean forests of California (Witty et al., 2003). Soil water is often
depleted by mid-July in the TB (Rogers, 1974). With the majority
of the precipitation delivered in the form of snow in the TB, soil
94
Downloaded from http://forestry.oxfordjournals.org/ at DigiTop USDA's Digital Desktop Library on March 14, 2014
HYP
Intercept
BAL
CC
CSA
Best model
Intercept
BAL
CSA
Estimate
moisture availability to seedlings depends on the water-holding
capacity of the soil and competition from other vegetation (Gray
et al., 2005).Although available soil moisture was not directly measured in this study, sugar pine seedlings have been found to be very
sensitive to changes in soil moisture, being both the least drought
resistant and tolerant in comparison to its mixed-conifer associates (Pharis, 1966). Other mid-tolerant white pine species have
also been found to be limited by soil moisture during the seedling
stage (Raymond et al., 2006). Sugar pine appears to require
ample soil moisture until it reaches sapling or greater sizes
(Tappeiner and McDonald, 1996), thus explaining the negative relationship found at the stand level with available soil moisture
between 15 and 30 cm. Recent research looking at the resource
requirements for diameter growth in mixed-conifer saplings on
the Plumas National Forest north of the TB found a slight negative
correlation between diameter growth and available water for
sugar pine saplings (Bigelow et al., 2009).
Shrub cover and density may have influenced growth rates of
seedlings; however, seedling locations relative to shrubs was not
recorded. Seedling establishment improves in shaded areas due
to the facilitative effects of shrubs that increase available soil moisture, decrease light intensity and heat, and reduce soil temperatures and evaporation (Dulohery et al., 2000; Gray et al., 2005;
Legras et al., 2010). Desiccation from direct solar radiation
causes high seedling mortality rates (Gray et al., 2005). Legras
et al. (2010) also found that seedling survival was lowest following
treatments that involved canopy removal during initial seedling establishment. Poor survival rates were attributed to water limitations caused by rapid drying of the soil from the additional heat
absorbed by exposed, darkened litter.
In other species, the benefits of an overstorey canopy on seedling growth are short-lived, switching to a competitive interaction
after several years (Dulohery et al., 2000; Milakovsky et al., 2011);
the effects are more pronounced on sites of poor productivity
(Zhang et al., 2006). Sugar pine has slow initial height growth
rates even on the most productive sites (Fowells and Schubert,
1956; York et al., 2004). It will persist at lower light levels and
grow at reduced rates in suboptimal growing conditions (Oliver
and Dolph, 1992; Waring and O’Hara, 2009). In the low and moderate quality sites such as the majority of the TB stands, growth may
be retarded for an even longer period of time, explaining why the
average height growth rates in this study were very low. It
appears that sugar pine seedlings can persist as advanced regeneration in the understorey, but it is unclear whether or not the
established regeneration will respond to treatments that increase
belowground resource availability in stands of low site productivity.
For continued growth of saplings, a recent study suggests that
even after vigorous fuels treatments have been applied, the conditions created are not sufficient to support rapid growth of shade
intolerants in the Central Sierras (Bigelow et al., 2009). However,
this same study was unable to identify specific light requirements
for sugar pine saplings, citing a small sample size (Bigelow et al.,
2009). Sugar pine tends to be less shade tolerant as it matures
(Minore, 1979) and may require lower densities and canopy closures for sustained growth and recruitment into the overstorey
(Gray et al., 2005).Sugar pine growing under a canopy of ponderosa
pine responded well to overstorey removal and grew as fast as
some other conifers at 58 per cent full sun (Oliver and Dolph,
1992). McDonald (1976) substantiated that sapling growth was
highest in treatments with the most aggressive reductions in
Predicting height growth of sugar pine regeneration
density and canopy closure such as clear-cut and seed tree
methods. However, uneven-aged treatments are more conducive
to ecosystem management and restoration than even-aged treatments yet can provide similar conditions if applied appropriately.
native species and climate change that impact the ability of
species to adapt.
Supplementary data
Supplementary data are available at Forestry Online.
Conclusion
Acknowledgements
Jonathan Long (USFS Pacific Southwest Research Station), Dave Fournier
(USFS Lake Tahoe Basin Management Unit), Rich Adams (California State
Parks), Judy Clot (California Tahoe Conservancy), John and Maria Pickett
(Sugar Pine Foundation) and Brig and Mary Ebright provided field site
suggestions, permission and access. Jason Brown, Nathan Burgess, Chris
Erickson, Alan Griffin, Anna Higgins, Adam Polinko, Russ Potter, Brent
Prusse, Alex Spannuth and Jay Williams provided field and lab assistance.
Conflict of interest statement
None declared.
Funding
This work was supported by grant #08-DG-11272170-011 from the United
States Department of Agriculture Forest Service Pacific Southwest Research
Station. Funding has been provided by the Bureau of Land Management
through the sale of public lands as authorized by the Southern Nevada
Public Land Management Act (to support restoration of Lake Tahoe).
Northern Arizona University School of Forestry provided additional financial
support to N. A.
REFERENCES
Adams, R. 2010. Forester, California State Parks, Personal Communication.
Angell, N. 2011 Determinants of pygmy sugar pine in the Lake Tahoe Basin,
CA and NV. Master of Science Thesis, Northern Arizona University, 101 p.
Ansley, J.A. and Battles, J.J. 1998 Forest composition, structure, and change
in an old-growth mixed conifer forest in the northern Sierra Nevada. J. Torrey
Bot. Soc. 125(4), 297– 308.
Applequist, M.B. 1958 A simple pith locator for use with off-center
increment cores. J. For. 56(2), 141.
Beers, T.W., Dress, P.E. and Wensel, L.C. 1966 Aspect transformation in site
productivity research. J. For. 64(10), 691– 692.
Bigelow, S.W., North, M.P. and Horwath, W.R. 2009 Resource-dependent
growth models for Sierran mixed-conifer saplings. Open For. Sci. J. 2, 31 –40.
Burnham, K.P. and Anderson, D.R. 2002 Model Selection and Multimodel
Inference: A Practical Information-Theoretic Approach, 2nd edn. Springer,
New York.
Chapin, F.S., Bloom, A.J., Field, C.B. and Waring, R.H. 1987 Plant responses to
multiple environmental factors. BioScience 37(1), 49– 57.
Coomes, D.A. and Grubb, P.J. 2000 Impacts of root competition in forests
and woodlands: a theoretical framework and review of experiments. Ecol.
Mono. 70(2), 171–207.
Dulohery, C.J., Kolka, R.K. and McKelvin, M.R. 2000 Effects of a willow
overstory on planted seedlings in a bottomland restoration. Ecol. Eng. 15,
57– 66.
Elliott-Fisk, D.L., Rowntree, R.A., Cahill, T.A., Goldman, C.R., Gruell, G., Harris,
R., Liesz, D., Lindstrom, S., Kattelmann, R., Machida, D., Lacey, R., Bucks, P.,
Sharkey, D.A. and Ziegler, D.S. 1996 Lake Tahoe case study. In Sierra
95
Downloaded from http://forestry.oxfordjournals.org/ at DigiTop USDA's Digital Desktop Library on March 14, 2014
Height growth at the individual tree-level is complex and difficult to
predict. Variables such as nutrient availability, soil type and shrub
competition may explain more of the variation in height growth
at this scale but were not included in this study. At the individual
tree-level, sugar pine seedlings are less affected than saplings by
dense stand conditions and canopy closure. These trees probably
respond to both competitive interactions with the overstorey
canopy and potentially the shrub layer in addition to abiotic site
factors. It is unclear whether seedlings or saplings will respond favourably to treatments that reduce density and canopy closure,
particularly if they are already suppressed (Angell, 2011). For saplings, the hypothesis that individual tree growth is maximized in
microsites with reduced densities and canopy closures was supported. Continued growth of saplings may require creating more
open microsites with increased resource availability through
more aggressive uneven-aged silvicultural treatments.
At the stand level, two models best explained seedling height
growth and can be easily utilized by managers. To enhance sugar
pine seedling growth at this level, creating shaded microsites for
initial seedling establishment and following-up with intermediate
treatments to reduce belowground competition and increase resource availability are recommended. For continued growth of saplings, treatments that reduce densities and canopy closures should
aid in sapling recruitment into the overstorey. More research should
be conducted to determine whether the suppressed regeneration
that currently exists in the understorey will respond positively to
more aggressive uneven-aged silvicultural treatments. Treatments designed for restoring stand structure and composition to
conditions more similar to those occurring prior to the late 1800s
are also likely to favour sugar pine regeneration and growth
(Taylor, 2004; North et al., 2007, 2009).
Guidelines for improving growth of sugar pine regeneration that
can be incorporated into restoration treatments include the following: 1. reduce above- and below-ground competition and provide
shade for greater seedling growth and 2. reduce aboveground
competition and create gaps around saplings for greater sapling
growth. Sugar pine seedling and sapling height growth is a
complex issue determined by a range of abiotic and biotic site variables. Understanding the importance of these variables will allow
for greater success in sugar pine restoration efforts.
While the TB represents only a minor component of sugar pine’s
range, the results of this research are applicable outside of the TB.
The results presented here largely support other research indicating that treatments not designed to create openings are unlikely
to enhance regeneration and recruitment of sugar pine to the overstorey. Sugar pine is similar in shade tolerance to other white pines,
such as eastern and western white pines (P. strobus and P. monticola), also impacted by white pine blister rust. Similar factors are
likely important in regeneration and recruitment of such species.
Comprehensive knowledge of species silvics and regeneration
ecology, coupled with management recommendations, is critical
for sustainable forestry, particularly given factors such as non-
Forestry
Nevada Ecosystem Project: Final Report to Congress. pp. 217–276.
University of California Wildland Resource Center Report Number 37.
control on survival and growth of white spruce (Picea glauca (Moench)
Voss) seedlings. For. Ecol. Manag. 261(3), 440– 446.
Fowells, H.A. and Schubert, G.H. 1956 Silvical Characteristics of Sugar Pine.
Pacific Northwest Forest and Range Experiment Station Technical Paper
No. 14.
Minore, D. 1979 Comparative Autecological Characteristics of Northwestern
Tree Species: A Literature Review. USDA Forest Service, Pacific Northwest
Forest and Range Experiment Station Research Paper. Gene Technology.
Report. PNW-87.
Fowells, H.A. and Stark, N.B. 1965 Natural Regeneration in Relation to
Environment in the Mixed Conifer Forest Type of California. USDA Forest
Service Pacific Northwest Forest and Range Experiment Station Research
Paper PSW-24.
North, M., Hurteau, M., Fiegener, R. and Barbour, M. 2005a Influence of fire
and El Nino on tree recruitment varies by species in Sierran mixed-conifer.
For. Sci. 51(3), 187– 197.
North, M., Oakley, B., Fiegener, R., Gray, A. and Barbour, M. 2005b Influence
of light and soil moisture on Sierran mixed conifer understory communities.
Plant Ecol. 177, 3– 24.
Holmes, R.L. 1983 Computer-assisted quality control in tree-ring dating and
measurement. Tree-Ring Bull. 44, 69 –75.
North, M., Innes, J. and Zald, H. 2007 Comparison of thinning and prescribed
fire restoration treatments to Sierran mixed-conifer historic conditions.
Can. J. For. Res. 37, 331–343.
Hubbert, K.R., Graham, R.C. and Anderson, M.A. 2001 Soil and weathered
bedrock: components of a Jeffrey pine plantation substrate. Soil Sci. Soc.
Am. J. 65, 1255– 1262.
Kinloch, B.B. and Scheuner, W.H. 1990 Pinus lambertiana Dougl., sugar pine.
In Silvics of North America. pp. 370– 379. 1. Conifers; 2. Hardwoods. Edited
by R. M. Burns and B.H. Honkala (technical coordinators). Agriculture
Handbook 654. USDA Forest Service, Washington, D.C. 877 pp. U.S.
Department of Agriculture, Forest Service, Washington, DC. 877 pp.
Kranabetter, J.M. and Simard, S.W. 2008 Inverse relationship between
understory light and foliar nitrogen along productivity gradients of boreal
forests. Can. J. For. Res. 38, 2487– 2496.
Krumland, B. and Eng, H. 2005 Site Index Systems for Major Young-Growth
Forest and Woodland Species in Northern California. State of California
Research, Agency of California, Department of Forest and Fire Protection.
Rep. No. 4.
Larsen, L.T. and Woodbury, T.D. 1916 Sugar Pine. USDS Agriculture Bulletin.
No. 426, 40 p.
Larson, E.R. and Kipfmueller, K.F. 2010 Patterns in whitebark pine
regeneration and their relationships to biophysical site characteristics in
southwest Montana, central Idaho, and Oregon, USA. Can. J. For. Res. 40,
476–487.
Larson, M.M. 1963 Initial root development of ponderosa pine seedlings as
related to germination date and size of seed. For. Sci. 9(4), 456– 460.
Legras, E.C., Vander Wall, S.B. and Board, B.I. 2010 The role of germination
microsite in the establishment of sugar pine and Jeffrey pine seedlings.
For. Ecol. Manag. 260(5), 806– 813.
Lenart, D.Y. 1934 Initial root development of longleaf pine. J. For. 32,
459–461.
Long, J.N. and Daniel, T.W. 1990 Assessment of growing stock in
uneven-aged stands. West. J. Appl. For. 5(3), 93 –96.
Maloney, P.E., Vogler, D.R., Eckert, A.J., Jensen, C.E. and Neale, D.B. 2011
Population biology of sugar pine (Pinus lambertiana Dougl.) with reference
to historical disturbances in the Lake Tahoe Basin: implications for
restoration. For. Ecol. Manag. 262, 770–779.
North, M., Stine, P., O’Hara, K., Zielinski, W. and Stephens, S. 2009 An
Ecosystem Management Strategy for Sierran Mixed-Conifer Forests. USDA
Forest Service Pacific Southwest Research Station, Albany, CA, 49 p.
Oliver, C.D. and Larson, B.C. 1996 Forest Stand Dynamics, Update Edition.
John Wiley & Sons, Inc., New York, NY, p. 520.
Oliver, W.W. and Dolph, K.L. 1992 Mixed-conifer seedling growth varies in
response to overstory release. For. Ecol. Manag. 48, 179–183.
Ott, R.L. and Longnecker, M. 2010 Statistical Methods and Data Analysis. 6th
edn. Duxbury, Pacific Grove, CA, p. 652.
Parsons, D.J. and DeBenedetti, S.H. 1979 Impact of fire suppression on a
mixed-conifer forest. For. Ecol. Manag. 2, 21 –33.
Pearson, G.A. 1930 Light and moisture in forestry. Ecology 11(1), 145–160.
Pharis, R.P. 1966 Comparative drought resistance of five conifers and foliage
moisture content as a viability index. Ecology 47, 211– 221.
Raymond, P., Munson, A.D., Ruel, J.C. and Coates, K.D. 2006 Spatial patterns
of soil microclimate, light, regeneration, and growth within silvicultural gaps
of mixed tolerant hardwood white pine stands. Canadian J. For. Res. 36,
639–651.
R Development Core Team. 2010 R: A language and environment for
statistical computing. R Foundation for Statistical Computing, Vienna,
Austria. http://www.R-project.org.
Regent Instruments Inc. WinDendro 2007 Tree Ring, Stem, and Wood
Density Analysis and Measurement. Regent Instruments Inc., Quebec City,
Canada.
Rogers, J.H. 1974 Soil Survey of the Tahoe Basin Area, California and Nevada.
U.S. Department of Agriculture, Soil Conservation Service and Forest
Service, in Cooperation with the University of California Agricultural
Experiment Station and the Nevada Agricultural Experiment Station,
Davis, CA.
Royce, E.B. and Barbour, M.G. 2001 Mediterranean climate effects. I. Conifer
water use across a Sierra Nevada ecotone. Am. J. Bot. 88(5), 911– 918.
Maloy, O.C. 2003 Updated 2008. White pine blister rust. The Plant Health
Instructor. DOI:10.1094/PHI-I-2003-0908-01.
SAS Institute Inc. 2007 1989–2007 JMP. Version 8. SAS Institute Inc.,
Cary, NC.
Manley, P.N., Fites Kaufman, J.A., Barbour, M.G., Schlesinger, M.D. and Rizzo,
D.M. 2000 Chapter 5: biological integrity. In Lake Tahoe Basin Watershed
Assessment, Vol. 1. Murphy, Dennis D. and Knopp, C.M. (eds). pp.
403–598. USDA Forest Service, Lake Tahoe Basin Management Unit Gene
Technology Report. PSW-175.
Smith, D.M., Larson, B., Kelty, C. and Ashton, P.M.S. 1997 The Practice of
Silviculture: Applied Forest Ecology. John Wiley & Sons, New York.
McDonald, P.M. 1976 Forest regeneration and seedling growth from five
major cutting methods in north-central California. USDA For. Serv., Pacific
Southwest Forest and Range Expt. Stn. Res. Pap. PSW-115.
Milakovsky, B., Frey, B.R., Ashton, M.S., Larson, B.C. and Schmitz, O.J. 2011
Influences of gap position, vegetation management and herbivore
96
Sugihara, N.G. and McBride, J.R. 1996 Dynamics of sugar pine and
associated species following non-stand replacing fires in white-fir
dominated mixed-conifer forests. In Proceedings of a Symposium
Presented by the California Sugar Pine Management Committee: Sugar
Pine: Status, Values, and Roles in Ecosystems. Kinloch, B.B. Jr, Marosy, M.
and Huddleston, M.E. (eds). pp. 39– 44. Pub. 3362, Davis, CA., 30 March–1
April 1992. University of California, Division of Agriculture and Natural
Research, Davis, CA.
Downloaded from http://forestry.oxfordjournals.org/ at DigiTop USDA's Digital Desktop Library on March 14, 2014
Gray, A.N., Zald, H.S., Kern, R.A. and North, M. 2005 Stand conditions
associated with tree regeneration in Sierran mixed-conifer forests. For. Sci.
51(3), 198–210.
Predicting height growth of sugar pine regeneration
Tappeiner, J.C. 1980 Sierra Nevada mixed conifer 243. In Forest Cover Types
of the United States and Canada. Edited by Society for American Foresters,
Washington, DC, pp. 118–119.
van Pelt, R. and Franklin, J.F. 2000 Influence of canopy structure on the
understory environment in tall, old-growth conifer forests. Can. J. For. Res.
30, 1231 –1245.
Tappeiner, J.C. and McDonald, P.M. 1996 regeneration of Sierra Nevada
forests. In Sierra Nevada Ecosystem Project: Final Report to Congress. Vol. II.
pp. 501–513. University of California Wildland Resource Center Report
Number 37.
Waring, K.M. and O’Hara, K.L. 2009 Stand development and tree growth
response to sugar pine mortality in Sierran mixed-conifer forests.
Northwest Sci. 83(2), 89– 100.
Taylor, A.H. 2004 Identifying forest reference conditions on early cut-over
lands, Lake Tahoe Basin, USA. Ecol. Appl. 14(6), 1903– 1920.
USDA Forest Service 1981 Calveg: A Classification of Californian Vegetation.
USDA Forest Service Regional Ecology Group, San Francisco, CA.
USDA Forest Service, Forest Inventory and Analysis 2007 National Core Field
Guide Volume I: Field Data Collection Procedures for Phase 2 Plots Version 4.0.
U.S. D.A. Forest Service, Washington, DC, 362 p.
USDA Natural Resources Conservation Service 2007 Soil Survey of the
Tahoe Basin Area, California and Nevada [online]. soils.usda.gov/survey/
printed_surveys/ (accessed on 15 February, 2009).
van Mantgem, P.J., Stephenson, N.L., Keifer, M.B. and Keeley, J. 2004 Effects
of an introduced pathogen and fire exclusion on the demography of sugar
pine. Ecol. Appl. 14(5), 1590– 1602.
York, R.A., Heald, R.C., Battles, J.J. and York, J.D. 2004 Group selection
management in conifer forests: relationships between opening size and
tree growth. Can. J. For. Res. 34, 630– 641.
Zald, H.S.J., Gray, A.N., North, M. and Kern, R.A. 2008 Initial tree regeneration
responses to fire and thinning treatments in a Sierra Nevada mixed-conifer
forest, USA. For. Ecol. Manag. 256, 168–179.
Zarnoch, S.J., Bechtold, W.A. and Stolte, K.W. 2004 Using crown condition
variables as indicators of forest health. Can. J. For. Res. 34, 1057– 1070.
Zhang, J.W., Oliver, W.W. and Busse, M.D. 2006 Growth and development of
ponderosa pine on sites of contrasting productivities: relative importance of
stand density and shrub competition effects. Can. J. For. Res. 36, 2426–2438.
Zuur, A.F., Ieno, E.N., Walker, N.J., Saveliev, A.A. and Smith, G.M. 2009 Mixed
effects models and extensions in ecology with R. Springer, New York, NY.
97
Downloaded from http://forestry.oxfordjournals.org/ at DigiTop USDA's Digital Desktop Library on March 14, 2014
USDA Forest Service 1994 East Shore Project: Final Environmental Impact
Statement. USDA Forest Service Lake Tahoe Basin Management Unit, Lake
Tahoe, CA.
Witty, J.H., Graham, R.C., Hubbert, K.R., Doolittle, J.A. and Wald, J.A. 2003
Contributions of water supply from the weathered bedrock zone to forest
soil quality. Geoderma 114, 389–400.