Oak Regeneration Potential Increased
by Shelterwood Treatments
Richard C. Schlesinger, Ivan L. Sander, and Kenneth R.
Davidson, USDA Forest Service,North Central Forest Experi-
ABSTRACT. In muchof the CentralHardwoodForestRegion,oak species not regenerating well, even
thoughlargeoak treesare common forests.Theshelterwood
method beensuggested as
a potentialtoolfor establishing developing advanced regeneration where it is lacking.The 10-yr results
from a studyof several variants of the shelterwoodmethodshowthat on good sitesin the Missouri Ozarks
Region,a heavyunderstory treatmentis mostimportant,while on averagesites,reducingtheoverstory density
levelis mostimportant.The right combinations overstory and understory will
treatments likelybe different
for differentstandsand locations,but the shelterwoodmethoddoesappear to be a usefulsilviculturaltoolfor
centralhardwood forests.North. J. Appl. For. 10(4): 149-153.
Within of Central Forest there was about60, while the meansiteindexfor the 30 goodsite
•slongstanding that are
concern oakspecies notregenerating plotswas about75.
well, andthatmanyof theforests by
nowdominated oakswill on
The primary oak species the studyareaswere white
be dominated other species
by and species groupsin the oak, black oak, northernred oak, and scarletoak. Post oak,
future (Sander1988, Lorimer 1989). Althoughsuchchange in
present small numbers, was includedwith the lessdesir-
bad,thevariousoak species considered able tree species.Other overstoryspeciesincluded black
components the forests, providingwoodprod- gum, hickory (mainly shagback and mockernut),and red
ucts,food for wildlife, andvarietyto the landscape. maple.The primaryunderstory specieswere dogwoodand
Advanced oak regeneration longbeenrecognized
has as sassafras.
the key to maintaining oak component centralhard- Three overstorydensitytreatments [leaving40, 50, or
wood forests(Carvell and Tryon 1961), but how to obtain 60% stocking basedon the oak-hickorystocking equation
amounts advanced regeneration whereit is not (Gingrich 1967)] were combinedfactorially with threeun-
present beenlessobvious. Several people havesuggested derstorytreatments(heavy,medium,or none)andreplicated
thattheshelterwood regenerationsystem maybe aneffective within three blocks on each site class.In addition, controlled
methodfor doing so (Clark 1970, Sander 1979). We con- burningwas testedas an understorytreatmentat 50% over-
ducteda study to test severalvariantsof the shelterwood The controlled
storystocking. burningtreatment consistedof
methodfor increasing oakregeneration potentialon sites two burns,onein the springimmediatelyafter the overstory
whereoakreproduction notsufficient meetregeneration and 3
treatment a second yr later.Thus,therewere 10 distinct
goals. shelterwood treatments.
Thesetreatments were slightlydifferentfor the two site
Methods In the
classes. bothcases, overstory leavetreeswereprima-
rily the larger,morevigorous oaks,andthe overstory treat-
The study area is locatedin the heart of the Missouri mentremoved treesdownto 1.6in. dbh.On thegoodsite,the
long.91.3øW) the SinkinExperimen- of
cut stumps nonoakspecies were treatedwith herbicideto
The soils theareaarepredominantly
tal Forest. in Clarksville On
reducesprouting. the averagesite, all stumps including
stonyor gravelly loams (loamy-skeletal,siliceous,mesic, oak stumps were treatedwith herbicideto reducesprouting.
Typic Paleudults).There were 30 treatment plotson eachof of
The objective theheavyunderstory was
treatment to kill all
two siteclasses. treatment
The plotswere 1.04acin sizewith woody vegetationlessthan 1.6 in. dbh exceptoakson the
a 0 3 ac measurement in the center.The meanblack oak goodsite,andit included oakstallerthan3 ft on theaverage
s•teindex(heightin feetat 50 yr) for the30 average plots
site site.The mediumunderstory treatment to
all woodyvegetation taller than6 ft andlessthan 1.6 in. dbh understory
•ntenslve treatment stock-
on the goodsite,andit includedoakstaller than6 ft on the affectthenumbers saplings
averagesite.The third understory left
treatment all woody The numberof saplings by
increased 221 stemswithin the
vegetation both sites. heavy understorytreatment to
plotscompared increases of
was annually duringthefirst 3 114 and 110 stemswithin the medium andcontroltreatments,
yr aftertreatment. trapsin eachof 3 plots respectively.
withineachoverstory density level,for a totalof 30 samples treatment about
The heavyunderstory was twiceaseffec-
per overstory densitylevel per site.The trapshad a surface in the
tive astheothertwo treatments increasing number of
area were to
collecting of 3.25ft2. Theacorns examined largeoak saplings. per
However,if 435 stems acre4.5 ft tall
determine they werematureandsound. to
or taller are necessary provide adequateregeneration
Reproduction measured twenty 1/735-acsubplots
was on (Sanderet al. 1976), not even the besttreatmentcombinaUon,
within eachmain plot. The subplots were measured before the heavyunderstory at
treatment 60% stocking, met this
treatment, or 2 yr aftertreatment (averageandgoodsites, goal.
respectively), and 10 yr after treatment. Measurements in- For the understory and
(dogwoodand sassafras) shrub
cludeda complete by and
tally of all stems species 0.305 ft (primarilywitchhazel)species the
height classes. addition, all oak stemsand a sampleof of
ing level favoredthe development more large saplings
stemsof otherspecies were identifiedandmappedso they (Table2). On theotherhand,therewerefewerlargesaplings
could be remeasured. For these stems,height and stem the
within the plotsthat received heavyinitial understory
diameterl in. above ground to
weremeasured thenearest treatment. for
0.4 in. and0.04 in., respectively. petition fromthese was
species themost intensiveunderstory
Analysisof varianceandcovariance wasusedto examine and
treatment the highest overstory density.
the response the reproduction the treatments. Because Initially, the oak saplings were only 2.8% of all large
the treatments were not identical on the two sites, each site reproduction stems(Table3), rangingfrom 0 to 26.9% onthe
separately. response of
variables primary of
27 individualplots.Sixteen theplotshadno oaksaplings
interestwere the numbersof large (taller than 4.5 ft) oak The understoryspecies(mainly dogwoodand sassafras)
after10yr andthechange numbers during the At
madeup69.8% of thesaplings. theendof 10 yr, oakshad
10-yr period.Regeneration stockingvalueswere computed to
increased 5.5% of the large saplings, and understory
using the method of Sander et al. (1984) and were also to
specieshad decreased 61.5%. And within the heavy
with analysis variance andcovariance. understory the
treatment, oaksmade up 10% of the large
The analyses for
were run separately the 3 x 3 factorial and
saplings, theunderstory species madeup 55.2%. During
involvingtheoverstory understory and
treatments for the thefirst3 yr followingtreatment, there were142,000,201,000,
four understorytreatments(includingcontrolled burning)at and 169,000sound per
acorns acreproduced underthe 40%,
an overstory densityof 50% stocking. The latter analyses 50%, and60% overstory densities,respectively.
some thesame plotsasthe3 x 3 factorialanalysesand The covarianceanalysisfor the 4 understory treatments
arethusnot completely independent. includingcontrolled a
burningshowed significant treatment
Analysesof the data collected beforethe shelterwood at
effect.The numberof largesaplings the endof 10 yr was
treatmentswere applied indicated significant differences 295 and250 on the burnedandheavyunderstory treatment
amongplotsin the numbers large oak reproduction and but
plots,respectively, wasonly 163 and 127 on the control
valuesonthegoodsitebeforetreatment, noton andmediumunderstory treatment plots,respectively.
theaverage Therefore, of
analyses covariance wereused
for the goodsitedata,with the originalnumberof largeoak
The heavy understorytreatment the
reproduction stocking value)asthecovariate. Analysesof
largeoaksaplingsfrom 151to 0, themedium from
variancewereusedfor the average data. 135 to 12, and the control treatment from 82 to 74. Since the
understorytreatmentsreducedthe large oaks by different
Results and Discussion the in
amounts, changes numbers from immediatelyafter
treatment of primaryinterestbecausethey reflectthe
of oak to
response theremaining reproduction theshelterwood
usingnumbers large oak reproduction yieldedthe same treatments.
conclusionsabout treatmenteffects. Becausethe stocking
The analysisof variancefor the 10 yr data showedno
valuecalculations notbeappropriate smalltreatment
significantunderstory treatmenteffects (Table 4), but a
the was for
plotsbecause procedure developed largerstands,
significantoverstoryeffect. The lowest stockinglevel re-
we present based thelargeoakreproduc-
only theresults on
sulted themostlargeoakregeneration. analysis the of
change numbers fromimmediately aftertreatmentuntilthe
Good Site loth year showedthe sameresult.For the averagesite, the
Based theanalysis covariance, absolute
the number per
numbers acre for the loweststocking level exceedthe
of large oak saplings acreafter 10 yr was significantly suggested number of 435/ac thoughtto be sufficient to
affectedby the understorytreatment,especiallythe most in
providethe oakcomponent the next stand.
Table 1. Number per acre of large oak reproduct,onstems 10 yr after treatment and ( ) before treatment, good s,te only.
Understory removal treatment 40% 50% 60% All levels a
Heavy 306 (37) 208 (0) 343 (159) 286a (65)
Medium 208 (37) 98 (12) 86 (0) 130b (16)
Control 86 (0) 172 (49) 184 (61) 147b (37)
200a (25) 159a (20) 204a (73) 188 (39)
aMeans by are
Table 2. Number per acre of large understoryspecies (dogwood,sassafras,and shrub) reproductionstems 10 yr after treatment,good
Understoryremovaltreatment 40% 50% 60% All levels a
Heavy 2377 1764 600 1580b
Medium 2671 2744 1936 2450a
Control 2658 1838 2242 2246a
All removalsa 2569a 211 5ab 1593b 2092
aMeans by letter not
Table 3. Percentageof large reproductionstems that were oaks 10 yr after treatment and ( ) beforetreatment, good site only.
Understory removal treatment 40% 50% 60% All levels a
Heavy 7.6 (3.8) 6.7 (0.0) 23.7 (16.9) 10.0 (5.8)
Medium 4.8 (2.3) 2.6 (1.0) 2.7 (0.0) 3.5 (1.2)
Control 2.6 (0.0) 4.8 (2.7) 4.9 (3.5) 4.1 (2.2)
5.1 (1 .9) 4.5 (1 .4) 7.3 (5.3) 5.5 (2.8)
stems10 yr aftertreatmentand ( ) immediately
Table4. Numberper acre of largeoak reproduction aftertreatment,averagesite only.
Understoryremovaltreatment 40% 50% 60% All levels a
Heavy 588 (0) 306 (0) 74 (0) 323a (0)
Medium 576 (25) 294 (12) 306 (0) 392a (12)
Control 527 (135) 441 (49) 319 (37) 429a (74)
All a a
564 (5.3) 347
b (20) b
233 (12) 381 (29)
aMeans by different
followed the sameletterare notsignificantly to Multiple
according Duncan's RangeTest.
Therewerefewerlargesaphngs theunderstory species excellent
islikely to create growing for
on the averagesitethan on the goodsite:329/ac versus976 species. methodhas succeeded,
Where the shelterwood an
(before treatment)and 828 versus2092 (10 yr after treat- appropriate opening
mix of overstory treat-
ment). The numberof saplings10 yr after treatmentwas not mentshas been found (Loftis 1990, Wolf 1988).
significantlyaffectedby eitherthe understory treatment or One key to obtainingan adequate is
oak component to
the overstorytreatment(Table 5). haveenough advancedreproduction of
stems sufficient size
The combinationof more large oak saplingsand fewer with the nonoakrepro-
that they can competesuccessfully
large saplings the understory speciesresultedin a higher ductionfollowing overstoryremoval. "Enough"has been
proportion large oak saplings the averagesitethan on
of on defined as 435 stems/ac if an oak-dominated stand is the
thegood (Table6). Ontheaverage theoaks made up management et
objective(Sander al. 1976),butcouldbe less
15% of the largesaplingsbeforetreatment and 19% 10 yr in
if a smalleroak component the next standis acceptable
aftertreatment.Beforetreatment,only 1of the27 averagesite "Sufficient size" hasbeendefined as about4.5 ft or larger
plotshadnolargeoakreproduction. thissite,sound acorn (Sander 1972).
productionduring the first 3 yr following treatmentwas A secondkey to successful regenerationis for the oak
130,000, 145,000, and 160,000/ac for the 40%, 50%, and reproduction to
stems be ableto compete with othertreeand
60% overstory densitylevels,respectively. shrubregeneration et
(Sander al. 1984,Dey et al. 1991). The
The analysis the understory treatments includingcon- of stem
competitiveness anoakregeneration depends both on
trolled burningshowedno significanttreatmenteffects.On the absolute size of the oak and its size relative to the
the average per
site,therewere441 largeoak saplings acre competing treesand shrubs (Sander1972).
after 10 yr for the controltreatment, 380 for the controlled In the Appalachians, Loftis (1990) foundthat reducing
burning treatment,307 for the heavy understorytreatment, stand density was
from belowusingherbicides effectivefor
and294 for the mediumunderstory On
treatment. this site, regenerating red he
northern oak.Furthermore, foundthatthe
controlled burningdid not benefitthe regeneration didbut betterthe site quality, the higher the shelterwoodstocking
damage numberof the residual overstory trees. level to leave. For the two siteclassesstudiedin Missouri, our
The role of advanced in
reproduction theoakregeneration for
data supportthis samegeneralconclusion mixed oak
processis well known (Sander and Clark 1971). The On
regeneration. the bettersite, the higheststocking level
methodhasbeen suggested a procedure for combined with a heavyunderstory treatment to
reproduction regenerate the mostsuccessful,whileonthepoorer site,theloweststoclung
oak componentin standswhere it is insufficient,but the level was best.
methodhas frequentlyfailed (Sander 1979). Its primary Fire is thoughtto be an importantfactorin providinga
drawback thatcreating appropriate conditions oaks
for competitive for in
advantage oak species Midwest ecosys-
Table 5. Number per acre of large regeneration stems of understory (dogwood, sassafras, and shrub) species 10 yr after treatment,
average site only.
Residual stocking level
Understow removal treatment 40% 50% 60% All levels a
Heavy 931 809 478 739
Medium 906 1176 380 821
Control 1078 931 760 923
Allremovals 972 972 539 828
Table 6. Percentage of large reproduction stems that were oaks 10 yr after treatment and ( ) before treatment, average site only.
Understory removal treatment 40% 50% 60% All levels a
Heavy 25.3 (23.1) 16.3 (14.1) 5.6 (15.0) 17.5 (17.7)
Medium 25.4 (22.2) 11.8 (10.3) 22.7 (14.9) 19.3 (15.5)
Control 19.5 (7.9) 22.2 (19.5) 18.1 (13.4) 19.9 (11.8)
All a 23.2 (17.2) 16.4 (13.4) 15.8 (14.5) 19.0 (15.2)
152 NJAF 10(4)1993
tems (Crow 1988, Rouse 1986, Reich et al. 1990). The oak neededona variety
andamountof advanced regeneration
in wasaseffective of sites.
as the heavy understorytreatmenton the better site. We
probably not,andperhaps cannot, the
duplicate effects of Literature Cited
wildfires by usingcontrolledburning,but we can suggest
K.L., andE.H. TRYON.
CARVELL, 1961. The effect of environmentalfactorson
controlledburningcould be a usefulregeneration tool on
the abundance oak regeneration For. Sci.
somesiteswhereunderstory controlis needed. 7:98-105.
CLARK,F.B. 1970. Measures for of
necessary naturalregeneration oaks,
sweetgum, blackwalnut.USDA For. Serv.Res.Pap.
NE-144. P. 1-16.
Management Implications T.R. 1988.Reproductive
of northernred oak (Quercusrubra)--a review. For. Sci. 34:19-40.
On averagesitesin the MissouriOzarks(black oak SI of H.E. and 1991.
DEY,D., P.S. JOHNSON, GARRE•FF, P.L. SPECKMAN. Interfacing
55 to 65), numbers advanced reproduction often
oak will modelwith growthandyield models.
a regeneration USDA For. Serv.
Gem Tech. Rep. SE-70. P. 221 228.
be adequate.Simply reducingthe overstorydensitywill S.F. and
GINGmCH, 1967.Measuring evaluating and
stocking stand densityin
allow thisadvanced to
reproduction developsothatthe oak upland hardwood in
forests the CentralStates. For. Sci. 13:38 53.
saplings can competesuccessfully oncefully released. On LOFTIS,D.L. 1990.Predicting post-harvest performance advance oak
reproduction the southern Appalachians. For. Sci. 36:908-916.
thesesites,therearefewer nonoakspecies cancompete C.G.
LOmMER, 1989.Theoakregeneration problem: on
New evidence causes
successfully theoaks. theprimarymanagement objec- and possible solutions.For. Res. Anal. R3484 No. 8. Univ. Wisc. Sch.
tive is retention an oak component within the next stand, 31
REICH,P.B., M.D. ABRAMS, E.L. and
D.S. ELLSWORTH, KRUGER, T.J. TABONE.
the final overstory removalcut couldbe madeat ten years, and
1990. Fire affectseLDphysiology community of
andperhaps earlier. oak regeneration.
Wisconsin forest 71:2179-2190.
However,on the bettersites(black oak S170 andabove), RousE,C. 1986. Fire effects in northeasternforests:Oak. USDA For. Serv.
Gen. Tech. Rep. NC-105.7 p.
reducing overstory densityalonemaynotbe sufficient to I.L.
SANDER, 1971.Heightgrowthof new oak sprouts on
produce competitive oak advanced It
regeneration. may be advance J.
reproduction. For. 69:809-811.
necessary reducethe numbers competing
to of nonoaksap- SANDER, 1972.Sizeof oakadvance
I.L. reproduction: togrowth
ing harvest cutting.USDA For. SerF.Res.Pap.NC-79.6 p.
lingsdirectly and the
withfire orherbicide to retard develop- Sander, 1979. Regenerating with theshelterwood
I.L. oaks P.
mentof the competition maintaining densera overstory. in Regenerating oaksin upland Holt, H. (ed.). The 1979JohnS.
Furthermore, appears that developing adequate oak ad- WrightFor. Conf. Proc.Purdue W.
SANDER, 1988. Guidelinesfor regenerating oak stands. P.
vancedregeneration will requiremore than 10 yr and/or 189 198in Smith,H.C., A.W. Perkey,andW.E. Kidd, Jr. (eds.).Work-
additionalunderstory treatment. for
shopproc.:Guidelines regenerating Appalachianhardwood stands.
Adapting the shelterwood systemto the regeneration SAF Publ. 88-03. Soc. Am. For., Bethesda,MD.
SANDER, and F.B. CLARK. of
1971. Reproduction uplandhardwood
needsof the oaksseems be quite possible, although a in
forests theCentralStates. USDA For. Serv.Agric. Handb.405.25 p.
universalsystemfor all sitesis not likely. Obtainingad- Sander, and
LL., P.S.Johnson, R.F. Watt. 1976.A guidefor evaluating the
vancedoak regeneration sufficientsize and in sufficient of
adequacy oakadvance USDA For.Serv.Gen.Tech.Rep.
numbersto perpetuate this importantforest speciesmay Sander,I.L., P.S. Johnson,and R. Rogers. 1984. Evaluatingoak advance
requirea series bothunderstory overstory treatments in
reproduction theMissouri Ozarks.USDA For.Serv.Res.Pap.NC-251.
onthebetter sites; increasing radiation
the withintheseedling 16p.
WOLF, W.W., JR.1988.Shelterwood to
space beall thatisneeded lowerquality sites.Butas P. and
experience. 210-218in Smith,H.C.,A.W. Perkey, W.E. Kidd,Jr.
this, and other, research has shown, some variant of the (eds.).Workshop for
proc.:Guidelines regenerating Appalachianhard-
system beeffective developing type
shelterwood can for the SAF Publ. 88-03. Soc. Am. For., Bethesda,MD.