This
extension was coded by Dr. Chris Hennigar in 2020 and derived in principle from
SBW DSS methods (Erdle and MacLean 1999; MacLean et
al. 2001) with notable upgrades over the past 20 years described below. This
extension is currently accessible only from the Acadian variant.
This extension provides means to:
1) Submit and remember spruce
budworm historical observations and/or future projections of either:
a.
annual
current-shoot defoliation
OR
b.
annual
L2 populations.
2) Estimate host annual current-shoot
defoliation from annual L2 levels, with consideration of:
a.
stand
hardwood content,
b.
natural
enemy buildup during the outbreak,
c.
foliage
protection efficacy and frequency,
d.
white
spruce foliage endophyte efficacy, and
e.
host
susceptibility differences.
3) Translate annual
current-shoot defoliation to whole-tree defoliation according to the SBW DSS
(MacLean et al. 2001).
4) Predict SBW DSS impact rates
as a function of whole-tree defoliation, host species, and tree size,
including:
a.
Tree
diameter growth reduction multipliers, and
b.
Tree
additive mortality rates (probability of tree death).
5) Convert periodic defoliation
and impact relationships to annual rates for use in OSM annual or periodic
simulation cycles.
6) And finally, modify OSM
host-tree base growth and mortality rates with these predictions during
simulation.
These
properties and methods are accessible using the command SIMULATION.MODEL.SBW
|
Property | Method |
Data Type |
Constraint |
Default |
Description |
|
Outbreak Definition |
||||
|
SetPopulationPattern |
Text |
≥
0 |
None |
Pre-coded
outbreak patterns (MacLean et al. 2001): ·
Moderate (based on 1980s NB
‘Normal’ Outbreak) ·
Severe (based on 1970s Cape
Breton, NS Outbreak) ·
None (no outbreak) Defoliation
is calculated as a function of L2 values, host susceptibility,
foliage protection properties. |
|
SetPopulation |
Float
Array |
≥
0 |
0 |
Sets
a user-defined annual population (L2/m2) sequence for
an outbreak. Defoliation
is later calculated as a function of L2, host susceptibility,
foliage protection, etc. |
|
SetDefoliation |
Float
Array |
0
– 100 |
0 |
Sets
a user-defined annual current-shoot defoliation (%) sequence for an outbreak.
Annual defoliation levels are influenced by foliage protection. Defoliation
is assumed to be based on actual observed levels. Levels should already be
adjusted by the user to account for past or future protection, if any. If
set, then any previously set population pattern is reset to 0. |
|
OutbreakOffset |
Int32 |
-5
– 100 |
0 |
Number
of years to offset the start of the outbreak; e.g., a
value of -5 causes the outbreak pattern to shift backward five years from the
start year of the simulation, and likewise end five years sooner. |
|
Foliage protection |
||||
|
ProtectFoliage |
Boolean |
TRUE|FALSE |
FALSE |
Turn
foliage protection on (TRUE) or off (FALSE). |
|
ProtectEfficacy |
Float |
0
– 1 |
0.5
(50%) |
Ratio
of L2 that will die as a result of
foliage protection. 1 = 100% effective and 100% budworm mortality. Single
applications of Btk and Mimic efficacies
vary in the range of 40-80%. Double or triple applications may assume
slightly higher protection levels. |
|
ProtectionThreshold |
Float |
≥
0 |
40
(40%) |
Threshold
percent annual current-year shoot defoliation on balsam fir that would
trigger a foliage protection treatment. |
|
ProtectionDelay |
Byte |
≥
0 |
0 (none) |
Number
of years to delay protection after first eligible year (when protection
threshold is first achieved); default = 1 (skip first eligible year). If = 0,
protection will occur in the first eligible year. |
|
ProtectionLock |
Byte |
≥
0 |
0
(none) |
Number
of years to skip between annual protection treatments; default = 0 (all
years). If = 1, protection will occur in every second year. |
|
EndophyteEfficacy_WS |
Float |
0
– 1 |
0
(none) |
Assume
planted white spruce is inoculated with endophytes that reduce budworm survival; e.g., 0.3 = 30% mortality in every outbreak
year. This effect is multiplicative to effects of foliage protection. |
SIMULATION.MODEL.SBW.SetPopulationPattern Moderate
SIMULATION.MODEL.SBW.SetPopulationPattern Severe
SIMULATION.MODEL.SBW.SetPopulationPattern Severe
SIMULATION.MODEL.SBW.OutbreakOffset -5
SIMULATION.MODEL.SBW.SetPopulationPattern None
SIMULATION.MODEL.SBW.SetPopulation 20.00 20.51
21.05 21.62 22.22 80.00 105.88 225.00 112.50 103.23 106.67 96.55 57.14 14.81
15.38 16.00
SBW
SetPopulationPattern Severe
ProtectFoliage TRUE
SBW
SetPopulationPattern Severe
ProtectFoliage TRUE
ProtectionEfficacy 0.6
ProtectionThreshold 50
ProtectionDelay 2
ProtectionLock 1
The
SIMULATION.MODEL.SBW object provides stand-impact routines derived from
the Spruce Budworm Decision Support System (Erdle and
MacLean 1999, MacLean et al. 2001) with some significant method upgrades
over the past 20 years, including:
1)
Reduced
defoliation on spruce, especially red and black, relative to fir according to
Hennigar et al. (2008).
2)
Reduced
defoliation with increasing stand hardwood content (Su
et al. 1996, Zhang et al. 2018).
3)
Ability
to input annual SBW population levels [larval 2nd instar (L2)/m2],
with internal functions to predict defoliation on respective host species as a
function of defoliation: population relationships, stand hardwood basal area
ratio, and foliage protection assumptions.
4)
Ability
to use annual stand-level defoliation levels (as in the original SBW DSS), but
with auto-prorating of stand defoliation onto hosts (balsam fir and white, red,
and black spruce) as a function of their basal area ratio and relative
susceptibility to SBW defoliation; i.e., if the stand is composed half of fir and
half of black spruce, then the fir will sustain more relative defoliation
compared to the black spruce, but basal area weighted stand defoliation will
still match aerially observed or projected input levels.
5)
Preprogrammed
‘Moderate’ and ‘Severe’ SBW outbreak scenarios expressed in annual population
levels. These population patterns were estimated from the original SBW DSS
defoliation-based outbreak scenarios using the same defoliation: population
relationships as discussed above, so outcomes should be similar.
6)
Foliage
protection efficacy assumptions and treatment options that can be modified
easily.
Technical
details of these assumptions and calculations are provided in SbwOutbreakAssumptions.xlsx. This Excel file demonstrates the effect of modifying some of the
foliage protection assumptions on resulting host defoliation and calculation of
periodic average whole-tree defoliation. These calculations are nearly
identical to the calculation performed by the SBW extension. Growth multipliers
& additive mortality rates are also provided as a function of host, tree
size, and periodic average whole-tree defoliation (aka cumulative defoliation
in MacLean et al. 2001). These tree impact: defoliation relationships
have been modified slightly from the original SBW DSS to 1) use tree size (DBH
class) as a surrogate for tree age (age is not a readily available attribute on
all trees in most inventories), and 2) predict impact rates with equations
rather than a lookup table for application coding/performance reasons.
Before translating population levels to defoliation levels,
annual populations (L2/m2) are adjusted for
multiplicative effects of stand
hardwood content, natural enemy buildup during the outbreak, foliage
protection, white spruce plantation endophyte efficacy, and host susceptibility
to defoliation, in that order. These adjustments are provided in SbwOutbreakAssumptions.xlsx, including assumptions and
cell comments regarding the conversion of the SBW DSS Moderate and Severe
outbreak patterns from annual fir defoliation to annual L2 density. Dr. David
MacLean provided technical review of this work as part of Eric Ye Liu’s masters
project at UNB in 2019 studying the effects of SBW on the NB forest sector and
economy.
Increased
population density is assumed here to cause linear increases in annual
current-shoot defoliation for fir, up to 100% defoliation. Likewise, we assume
studies that predict relative defoliation reductions from increased hardwood
content or reduced host susceptibility, can be interpreted as equivalent
reductions to larva survival in most cases.
Here, annual population is converted to annual host current-shoot % defoliation (AD) using: AD = L2 * HW * NE * FP * wsENDO * HOST, where AD is capped at 100 (100% defoliation); HW is the stand hardwood effect = 1 – hardwood basal area ratio; NE is outbreak natural enemy buildup factor = Max (1 – 0.035 * outbreak_year, 0.6); FP = 1 – the foliage protection efficacy ratio; wsENDO = 1 – efficacy ratio of planted white spruce inoculated with endophytes; and Host = 1.00 for fir, 0.72 for wS, 0.42 for rS, and 0.28 for bS. This assumes that 50 L2/m2 will result in 50% defoliation on fir, assuming: i) the stand has 0% hardwood, ii) the outbreak is in year 1 (no buildup of natural enemies), and iii) no foliage protection occurs in that year. For white, red, and black spruce, if L2 density post HW, NE, FP, and wsENDO adjustments is > 70/m2, defoliation is predicted from adjusted L2 levels using a logistic relationship (see worksheet ‘Host L2 à DEFOL relationship’) to yield higher spruce defoliation levels under severe populations and to cap red and black spruce defoliation at 90 and 80%, respectively.
This
population-based input approach is useful for simulating severe population
years where budworm densities are far above levels needed to cause 100% fir
defoliation and where no amount of foliage protection would be sufficient to
avoid significant stand defoliation. In past SBW DSS work, protection was
assumed to always reduce defoliation to 40% regardless of L2 density.
Here, if populations are protected at 200 L2/m2 and we
assume 50% protection efficacy (50% more L5-L6 die), the
surviving budworm, despite protection, would remain high enough to cause
moderate-severe defoliation; and conversely, if there is only enough budworm to
yield 40% defoliation, then we would expect foliage protection to push
defoliation much lower than 40%. Alternatively, from a host susceptibility
perspective, if populations substantially exceed levels to cause 100% fir
defoliation, we would expect higher than 28% defoliation on black spruce
(following from Hennigar et al. 2008) from droves of starving budworm
dispersing to less-desirable hosts like black spruce in search of food. Using
these equations, we assume 200 L2/m2 would result in 78%
defoliation on black spruce.
Whole-tree defoliation each year is calculated
using MacLean et al. (2001) as the weighted average of current-shoot
foliage loss over the past six years: TDt = 0.28Ct +
0.26Ct–1 + 0.22Ct–2 + 0.13Ct–3 + 0.08Ct–4 + 0.03Ct–5 where t is year, TD is
whole-tree defoliation (%), and C is current-shoot defoliation (%).
Coefficients for each term in the equation equal the proportion of total
foliage mass by age-class on a healthy balsam fir crown, derived from
Baskerville and Kleinschmidt (1981).
In
the SBW DSS, whole-tree defoliation was averaged over 5-years and called
periodic cumulative-defoliation by MacLean et al. (2001). Periodic cumulative
defoliation was used to build defoliation impact relationships used in the SBW
DSS and here. However, as discussed in following sections, to permit annual
simulation cycles, this extension departs from this periodic convention by
instead using annual whole-tree defoliation to predict annualized versions of
the SBW DSS growth and mortality rates with minor bias correction scaling.
See SbwOutbreakAssumptions.xlsx worksheet ‘Host DEFOL
--> Impact’ for growth and mortality relationships used in the SBW DSS.
Note that this impact table has been adjusted to use DBH ranges as a
surrogate for age ranges in the original SBW DSS calibration (Erdle and MacLean 1999) by using DBH: Age relationships
developed from New Brunswick’s extensive network of timber cruises collected in
the 1990s and 2000s.
Base
growth rates are reduced proportional to the % of whole-tree foliage remaining.
Technically, the extension uses a simple linear equation (growth rate
multiplier = -0.0099 * defoliation + 1.0182; bounded between 0 and 1; r2 = 0.9984)
to closely match the SBW DSS impact table in SbwOutbreakAssumptions.xlsx. If the simulation cycle is
between 2 and 5 years, these growth rate multipliers are averaged over the
cycle. SBW growth reduction multipliers are applied to both DBH and height
increments.
Periodic
additive mortality rates are calculated with a 3rd order polynomial
regression equation for each host in the form of: Intercept + DBH + DBH*D + D2
+ D3, where D is whole-tree defoliation. The SBW extension equation
closely matches the periodic mortality rates in the SBW DSS impact table in SbwOutbreakAssumptions.xlsx, when predicted rates are
bounded between 0 and 1 and DBH is bounded to the maximum upper range of DBH in
the impact table. Predicted periodic rates are then annualized using the
inverse of the compound interest formula. Annual rates are then scaled down by
6% (rate * 0.94) for reasons discussed in the following section. If the
simulation cycle is between 2 and 5 years, these scaled annual mortality rates
are averaged over the cycle. The cycle average annual tree mortality rate is added
to the base tree annual mortality rate.
Secondary
additive mortality post-outbreak from wind, beetles, and rot, while not
accounted for in the SBW extension, can be introduced manually with Amend
commands.
The extension supports 1 – 5-year cycle lengths.
In the original SBW DSS, the start date of the annual
defoliation pattern within the 5-year periods windows had an unintended
influence on the magnitude of stand mortality by the end of the outbreak. This
was because mortality risk increases exponentially with defoliation, so if all
severe years by chance occurred in one period, stand impacts by the end of the
outbreak would be predicted higher, compared to distributing the same number of
severe years over two or more periods. In other words, when severe defoliation
is spread across multiple periodic windows, it results in less cumulative
mortality, compared to one period with very high average defoliation. This is
not an issue for growth multipliers, as growth linearly declines with
defoliation and linear calculations are not sensitive to cycle duration or
defoliation timing.
Here, we attempt to avoid major differences in stand impact
results due to 1) the number of years in a simulation cycle (1-5), and 2) the
timing of defoliation (e.g., outbreak start offset between -2 and 2 years)
within fixed 5-year cycle windows. Following many different approaches
explored, the simplest and most reliable solution was to avoid periodic
defoliation calculations entirely and instead sum annualized mortality
rates predicted from annual whole-tree defoliation. A 6% reduction
factor was required on annual whole-tree defoliation to avoid mortality
overestimation, compared to the original SBW DSS, which relied on average
whole-tree defoliation rates over 5-years. These annualized
defoliation-mortality calculations and scaling reduced variation in stand
volume impacts to +/-3% from effects of cycle length changes (1-5 years) and
resulted in acceptable alignment with average impacts predicted from the
original SBW DSS when outbreak start year was varied between -2 and 2 years for
a wide range of outbreak patterns.
Whenever possible, it is recommended to run OSM with 5-year
cycles to align with data collection periods used to measure and model tree
growth and mortality.
See also: SIMULATION.YPC
Adjustments
can occur once in each simulation cycle when OSM calls the method ‘Predict Stand Dynamics’. During this call, variant tree growth (DBH and Height)
increments, tree mortality rate, and ingrowth recruits are predicted by the
variant model. Before returning from the call, the Acadian-Variant version of
this method checks to see if the SBW extension is active (true if a population
or defoliation sequence was assigned). If the SNW extension is active, then the
trees, along with their baseline predicted increments and survival rates, are
passed to the SBW disturbance extension for adjustment.
Additional
user-defined growth and survival amendments, if any, occur after the SBW disturbance extension
adjustments.
· Recent Estimates and Future Projections of Maine’s Timber Inventory, Forest Carbon Stocks, and Sequestration Rates: 2003 -2048, Executive Summary. Maine Forest Service, Department of Agriculture, Conservation and Forestry, Augusta. Pre-release draft version Dec 20th, 2021.
·
Maine
Wood Volume and Projection Study 2021 Update. Sewall Forestry & Natural
Resource Consulting. Sept. 20, 2021.
· Projections of Spruce Budworm Impact on Potential Wood Supply in Nova Scotia from 2020-2060. Report prepared by FORUS Research, Fredericton for the Nova Scotia Department of Natural Resources and Renewables. Oct 4th, 2021. 20 p.
·
Liu
EY, Lantz VA, MacLean DA, Hennigar C. 2019. Economics of early intervention to
suppress a potential spruce budworm outbreak on Crown land in New Brunswick,
Canada. Forests: 10: 481. https://doi.org/10.3390/f10060481
These
limitations below all result in underestimated stand impacts, and therefore should
be considered when using this extension, and where possible, factored into
results before decision making.
·
Post-outbreak
secondary disturbances (wind, beetle, rot) are not considered here. In pure fir-spruce stands
that have experienced prolonged moderate-severe defoliation, these secondary
disturbances can result in complete stand replacement in 10-15 years post
outbreak.
Taylor and MacLean (2009) showed rate of wind-related mortality peaked at 11 m3/ha/yr 11–15 years after defoliation ceased in mature-old
stands dominated by fir and spruce. In forest-level modeling, where outbreak is
severe enough and the stand vulnerable enough to expect complete stand
replacement, it is recommended that these areas be transitioned to an entirely
new regenerating state (e.g., one or more regenerating yield curves). For
example, in forest SBW-impact modelling, if a stand’s volume declines more than
about 40% during outbreak, then one may assume eventual stand replacement with
additional secondary disturbance and immediately transition the stand to a new
regenerating condition post-outbreak. While this additional secondary mortality
and stand regeneration can be accomplished in OSM through user-defined amendments of mortality and recruitment, the outcome is generally
similar and can be simpler to accomplish in the forest model.
·
Top
kill is a likely outcome during severe outbreaks, especially for fir, but is
excluded in this extension. While some trees may live on with top kill, their growth
becomes stunted with multiple unmerchantable tops growing above the point of
top kill. Saw material downgrades can be very significant if the point of top
kill is low on the bole. In the future, we hope to introduce a top kill model
presented by Virgin et al. (2017) that documents top kill on 40-88% of trees,
with top kill height at 32-88% of bole length, after 4-6 years of severe defoliation
in Cape Breton, NS. This study suggest that large saw material and volume
downgrades should be considered on surviving trees that have been severely
defoliated.
·
Back-feeding
on previous
years foliage is not simulated here but should be considered under very severe
populations such as in Cape Breton, NS during the 1980s outbreak. Ostaff and MacLean (1989) reported 20 plots in Cape Breton
had an average L2/m2 of 690 in 1977 and 540 in 1980, with
as high as 1340-1570 in individual plots! Piene (1989)
in the same study area observed trees stripped of all age classes by very high
populations in 1977 and 1978. In the future, we hope to introduce a threshold
assumption in the population-defoliation model that removes all tree foliage if
L2 densities are projected to be this extreme. The current ‘Severe’
outbreak pattern peaks slightly over 300 L2/m2, but this
by no means represents the most extreme population levels observed in recent
history.
·
Stand
mortality during outbreak can be clumpy, resulting in regeneration gaps where
clumps of mature fir have died. As OSM is non-spatial, and because SBW
mortality rates are based on average impacts over many stands, even during
severe outbreak, the simulated post-disturbance condition tends to retain some
trees and grow like a well-spaced naturally thinned stand. Actual SBW-caused
tree mortality patterns are often spatially complex and stochastic (MacLean
1980). Some fir stands, or parts of stands, will undergo complete stand
replacement while others will be relatively unscathed, and yet others will have
well-spaced mortality. It is unclear whether the extra effort to model this
stochastic spatial mortality in a spatially explicit individual-tree
environment would make a meaningful difference in average stand impact results,
but some additional impact is likely due to reduced full site occupancy.
Increased difficulty scheduling salvage and future harvest operations in these
clumped post-outbreak forest conditions (regenerating/mature) is another
consideration.
Baskerville, G., and Kleinschmidt, S. 1981. A dynamic model
of growth in defoliated fir stands. Can. J. For. Res. 11: 206–214
Erdle, T.A., and MacLean, D.A. 1999. Stand
growth model calibration for use in forest pest impact assessment. For. Chron. 75:
141-152.
Hennigar, C.R., MacLean, D.A., Porter, K.B., and Quiring D.T.
2007. Optimized insecticide application and harvest planning to reduce volume
losses to spruce budworm. Can. J. For. Res. 37: 1755-1769.
Hennigar, C.R., MacLean, D.A., Quiring D.T., and Kershaw J.A.
Jr. 2008. Differences in spruce budworm defoliation among balsam fir and white,
red, and black spruce. For. Sci. 54:
158-166.
MacLean, D.A. 1980. Vulnerability of fir-spruce stands during
uncontrolled spruce budworm outbreaks: a review and discussion. For. Chron. 56:
213-221.
MacLean, D.A., Erdle, T.A.,
MacKinnon, W.E., Porter, K.B., Beaton, K.P., Cormier, G., Morehouse,
S., and Budd, M. 2001. The Spruce Budworm Decision Support System: forest
protection planning to sustain long-term wood supplies. Can. J. For. Res. 31:
1742-1757.
MacLean, D.A., and Ostaff, D.P. 1989. Patterns of balsam fir mortality caused
by an uncontrolled spruce budworm outbreak. Can. J. For. Res. 19: 1087-1095.
Ostaff, D.P., and MacLean, D.A. 1989.
Spruce budworm populations, defoliation, and changes in stand condition during an
uncontrolled spruce budworm outbreak on Cape Breton Island, Nova Scotia. Can.
J. For. Res. 19: 1077-1086. https://doi.org/10.1139/x89-164
Su, Q.,
Needham, T.D., MacLean, D.A., 1996. The influence of hardwood content on balsam
fir defoliation by spruce budworm. Can. J. For. Res. 26: 1620-1628.
Taylor, S.L., and MacLean, D.A.
2009. Legacy of insect defoliators: increased wind-related mortality two
decades after a spruce budworm outbreak. Forest Science 55: 256-267.
Virgin, G.V.J., MacLean, D.A., Kershaw,
J.A,. Jr. 2018. Topkill and
stem defects initiated during an uncontrolled spruce budworm outbreak on Cape
Breton Island, Nova Scotia. Forestry: An International J. For. Res.: 91:
63-72. https://doi.org/10.1093/forestry/cpx035
Zhang, B., MacLean, D., Johns R.,
Eveleigh E. 2018. Effects of hardwood content on balsam fir defoliation during
the building phase of a spruce budworm outbreak. Forests 9: 530. https://doi.org/10.3390/f9090530