RESIDUAL EFFECTS OF THINNING AND HIGH WHITE-TAILED DEER
DENSITIES ON NORTHERN REDBACK SALAMANDERS IN
SOUTHERN NEW ENGLAND OAK FORESTS
ROBERT T. BROOKS,' U.S. Forest Service, Northeastern Research Station, 201 Holdsworth NRC, University of Massachusetts,
Amherst, MA 01003, USA
Abstract: Research has demonstrated that even-aged regeneration han-ests, especially clearcutting, can have
a major and long-lasting detrimental effect on forest amphibians, but the effects of less intensive sil\icultural
treatments have not been well documented. Additionally, the chronic overabundance of white-tailed deer
(Oclocoileus uirginianrls) has become a problem in many parts of North America, with associated effects on
vegetation con~positionand structure and on other wildlife. I assessed the effects of crown thinning and deer
overabundance on the relative abundance of forest-floor salamanders in a southern New England mixed oak-
hardwood forest. I surveyed salamanders by using cover boards in 16 forest stands with thinned or untilinned
treatments and with histories of low (3-6 deer/km2) or high (10-17 deer/km2) deer densities. Suneys were
conducted 5 times a year for 3 years. Northern redback salamanders (Plethodon cinereus) were the dominate
species in all sun7eysand in all treatment classes. Redbacks were most abundant in spring and fall surveys and
in the second and third year of the study. Neither thinning nor white-tailed deer density had a significant effect
on the number of redback obsen~ations: stantis with high numbers of redbacks occnrred in all treatment classes.
At the stand level, numbers of redback observations were positively correlated wit21 the number of pieces and
area of coarse woody debris and with the density of tall (21 m ) woody stems. The study suggests that a stand
disturbance, where a large percentage of the canopy is retained and that results in an increase in cover of
understory vegetation, would result in no long-term effect on forest-floor salamanders.
JOURNAL OF WILDLIFE MANAGEMENT 63(4):1172-1180
Key words: commercial thinning, cover-board survey, mixed oak-hardwood forest, Plethodun cineret~s,
back salamander, southern New England.
Although the policies and procedures of eco- numbers (Pough et al. 1987, Petranka et al.
system management of forests are evolving, 1993), but this effect is not certain (Chazal and
there is consensus that management should Niewiarowski 1998). In New England, this de-
have a long-term neutral effect on the structural cline is seen principally in numbers of the ter-
and functional integrity of the ecosystem (Healy restrial-breedmg northern redback salamander
1994).?iithin forest ecosystems, the role of am- (DeGraaf and Yamasaki 1992) and the pond-
phibians and the relation between forest man- breedng wood frog (Rana sylvatica) and mole
agement and amphibian ecology has been thor- salamanders (Ambystoma spp.; deMaynadier
oughly reviewed by deMaynader and Hunter and Hunter 1998). The effects of intermedate
(1995). Salamanders can account for a signifi- levels of forest harvesting on amphibians (i.e.,
cant component of the biomass of forest fauna thinning or selection-based silviculture) are less
in the eastern United States (Burton and Likens well known (Harpole and Haas 1999), but
1975a, Hairston 1987), and thus can also play small-scale modfications (e.g., firewood cutting)
an important role in detritus processing as pred- appear to have little detrimental effects on pop-
ators of forest-floor invertebrates (Burton and ulations of some northern salamanders (Pough
Likens 1975b, Wyman 1998). et al. 1987).
Although forestry practices can have a nega- White-tailed deer overabundance also affects
tive effect on some amphibians, maintaining forested ecosystems in many parts of North
sustainable populations can be compatible with America (\T7arren 1997). With sustained high -
timber harvesting if precautions are taken deer densities, for example, browsing can neg-
(deMaynadier and Hunter 1995, Waldick 1997, atively affect habitat composition and structure
Harpole and Haas 1999). Forest clearcutting for other wildlife species (Waller and Alverson
can result in a temporary decline in amphibian 1997), but the effects of chronic high densities
of white-tailed deer on amphibiais are un-
' E-mail: rbrooks@fon~ld.umass.ed~~ known.
J. Wildl. Manage. 63(4):1999 COMMERCIAL
THINNINGND SALAMANDERS rooks
A B 1173
Over the last 10 or more years, studies in the of the same type. Slope and aspect varied
oak forests of the Quabbin Reservation in cen- among the stands and within individual stands.
tral Massachusettshave documented changes in Soils across the study area were loamy and
vegetation, birds, gray squirrels (Sciurus caro- sandy glacial till, shallow to deep, and well
linensis), and small mammals following inter- drained (Soil Conservation Service 1967, 1989).
memate timber harvests in areas of hgh- and I selected 4 stands in each of 4 treatment
low-density whte-tailed deer populations (Hea- categories of a 2 x 2 factorial design: thinned
ly et al. 1987, Brooks and Healy 1988, DeGraaf and unthinned control in low deer density and
et al. 1991, Healy 1997). I surveyed the com- high deer density areas (Healy 1997). Densities
position and relative abundance of terrestrial of white-tailed deer in the low deer density,
amphibians of these forests to assess the lasting, hunted stands were estimated from harvest re-
long-term effects of intermediate timber har- cords to be about 3-6 deer/km2 (Healy 1997).
vests and chronic high levels of deer browsing. Deer densities in the high deer density, sanc-
Harvested and control stands were located both tuary stands were estimated by &stance-based,
in areas that supported low-density deer popu- line transect sampling at 10-17 deer/km be-
lations and in areas that had supported high- tween 1983 and 1992 (Healy 1997). A con-
density populations for many years. trolled hunt for white-tailed deer was begun on
the h g h deer density portion of the study site
STUDY AREA in 1991, 4 years prior to the first year of h s
The study was conducted within the water- study, and has continued every year thereafter
shed of the Quabbin Reservoir in Franbn, and is drastically reducing deer numbers.
Hampshire, and Worcester counties in central Thinnings in the silviculturally treated stands
Massachusetts (approximate centroid of study occurred between 12 and 21 years prior to the
sites at 42'25'N, 72'22'W). The watershed con- study (Hedy 1997). Thinning was done by re-
tains a 9,713-ha reservoir and 22,663 ha of sur- moving trees of undesirable species, form, or
roundmg uplands (O'Connor et d. 1995). Ap- condition from the main canopy to promote
proximately 19,000 ha of the uplands are man- growth on residual trees in accordance with the
aged forest. The area immediately adjacent to iilvicultural guidelines for upland oaks, and left
the reservoir was set aside as a wildLfe sanctu- a residual stocking of 5040%. In 1983, follow-
ary, and public access has been controlled since ing this si1vicultu;al treatment, thinned stands
1938. The property is managed by the Metro- were found to have significantly fewer trees
politan District Commission (MDC) to provide 22.5 cm in diameter, with significantlyless bas-
municipal water, forest products, wildlife, and al area, than unthinned stands (Healy et al.
recreation. 1987, Healy 1997). Stands with high densities
Study stands were identified from MDC for- of deer had fewer trees 22.5 cm in hameter
est inventory data. All stands met the following than stands with low deer densities because of
criteria: (1)white pine (Pinusstrobus)-northern a lack of recruitment into the smaller &ameter
red oak (Quercus rubra)-red maple (Acer rub- classes. Average stand mameter was unaffected
rum) forest type (Type 20; Eyre 1980:27-28); by thinning but was larger in high deer density
(2) even-aged, 60-100 years old; (3) sawtimber- stands because of reduced numbers of small-
sized with dominant trees having a diameter at hameter trees.
breast height (dbh) >30 cm; and (4) not on ex- Thinning resulted in an increase in seedlings
tremely steep or rocky sites (Healy et al. 1987, and saplings <2.5 cm in dameter, especially for
Hedy 1997). This forest type and other com- indviduals 230 cm in height (Hedy et al.
positionally similar types occur from central 1987). In stands with chronic high deer densi-
Maine, across southern New England, south ties, numbers of these stems were greatly re-
through the Middle Atlantic States, west down duced. Thinnings had no consistent effect on
the Ohio River and central Mississippi River other understory herbaceous and shrub vege-
watersheds, and in southern portions of the tation cover, but high deer numbers resulted in
Great Lakes States (Eyre 1980). The stands less forb and greater graminoid cover. In
generally originated in the first 2 decades of this thinned stands with high deer numbers, fern
century, following removal of the pine compo- cover increased dramatically (Healy et al. 1987).
nent. All stands were surrounded by extensive The most recent survey of understory vege-
forest land, predominantly sawtimber size and tation composition and structure in the study
SALAMANDERS J. W'ildl. Manage. 63(4):1999
stands occurred in 1994, >10 years since the 25 cm x 4 cm, in contact with the humus, 2.5
most recent thinning and 3 years after deer m apart at each of 30 vegetation survey plots in
numbers were reduced (FV. H. Healy, U.S. For- each stand. The plots were systematically locat-
est Service, unpublished data). Thinned stands ed 220 m apart along transects that paralleled
had a greater density of stems <2.5 cm in di- the long axis of the stands. The boards had been
ameter and a greater cover of herbaceous and air-dried prior to purchase and installation.
shrub cover than &d unthinned stands. Signif- Individual salamanders and other vertebrates
icant differences existed for woody stems 30-99 under each board were counted 5 times each
cm in height (unthinned stand average = 3,500 year between the months of May and October,
stemska; thinned stand average = 8,600 stems/ 1995 through 1997. Animals were handled only
ha; F1,l, = 8.85, P = 0.025), stems 2100 cm in if there was a question as to species identifica-
height (unthinned stand average = 1,400 stems/ tion; otherwise animals remained where ob-
ha; thinned stand average = 6,600 stems/ha; served when the board was carefully lowered.
F1,12 = 7.02, P = 0.038), and for forb cover Northern redback salamanders (hereafter, red-
(unthinned stand average = 12.2%; thinned back salamanders) were visually identified as
stand average = 30.5%; F1,12 = 9.95, P = unsexed or young-of-the-year (<30 mm snout-
0.017). Stands with a history of high densities vent length), immature or 1 year old ( 3 0 4 0
of deer had nonsignificantly lower densities of mm), or mature or 2 2 years old (>40 mm; Say-
woody stems 230 cm in height, greater fern ler 1966). The interval between successive sur-
cover, and lower shrub cover than stands that veys was never <2 weeks and generally was 4
had always supported low densities of deer. weeks. The surveys were not selected to follow
precipitation events. The exact dates of a survey
METHODS were scheduled to avoid having a precipitation
I surveyed the relative abundance of terres- event occur during a survey. It took between 1-
trial salamanders by using cover boards (De- 4 days to turn boards in all stands, but usually
Graaf and Yamasaki 1992, Fellers and Drost between 2-3 days.
1994, Davis 1997). The use of cover boards, a Coarse woody debris (CWD),including both
type of artificial cover object (ACO), for sur- logs and stumps, was measured on 33.3-m2cir-
veying terrestrial salamanders has not been cular plots centered on a point midway between
thoroughly validated (North American Amphib- the 2 boards. To be surveyed, CWD had to be
ian Monitoring Program [NAAMP]. 1997. The in contact with the forest floor. The minimum
terrestrial salamander monitoring program. diameter for CWD was 10 cm; no minimum
Bias: the disguiser of population trends. URL: length was set (Harmon and Sexton 1996). The
http://www.im.nbs.gov/sally3.html). A small CWD diameters were measured with calipers,
number of studies have found that the numbers length with a tape, and each piece was assigned
of salamanders under boards were correlated to 1 of 4 decay classes (U.S. Forest Service
with independent inlces of salamander abun- 1995). Decay classes ranged between sound
dance (DeGraaf and Yamasaki 1992; NAAMP. (Class 1) to fully rotted (Class 4) and were as-
1997. The terrestrial salamander monitoring signed based on structural integrity, texture of
program: recommended protocol for running rotten portions, and bark conditions. Depths of
cover object arrays. URL: http://www.im.nbs.gov/ organic soil horizons (litter, fragmentation, hu-
sally4.html). The use of boards or other types mus) were measured in a shallow profile exca-
of ACOs is being recommended for monitoring vated about 30 cm from the outer edge of each
terrestrial salamanders (Fellers and Drost 1994, cover board on each CWD survey plot. Because
Davis 1997, NAAMP. 1997. The terrestrial sal- of time constraints, CTVD was surveyed only on
amander monitoring program: recommended even- or odd-numbered plots (n = 15),with the
protocol for running cover object arrays. URL: decision determined by the toss of a coin in
http://www.im.nbs.gov/sally4.html). I chose to each stand.
use cover boards based on these recommenda- The relative abundance of redback salaman-
tions, the nondestructive nature of the meth- ders and forest floor habitat attributes were an-
odology, and the ability to implement a highly alyzed for the effects of deer density and thin-
standardized sampling design. ning by analysis of variance. Redback salaman-
In early December 1994, I placed 2 rough- ders observations were summed across plots to
cut hemlock (Tsuga canadensis) boards, 1 m x the stand because no salamanders were seen at
J. Wildl. Manage. 63(4):1999 COMMERCIAL
A Brooks 1175
Table 1. Number of observed fauna under cover boardsaby year and species, Quabbin Reservation, central Massachusetts,
1995 1996 1997 Total
Red-spotted newt (Notophthalmus uiridescensh) 15 36 33 84
Spotted salamander ( A m b y s t o m m c u l a t u m ) 1 3 6 10
Northern dusky salamander (Desmgnathusfuscus) 2 0 3 6
Northern two-lined salamander ( E uy c e a bislineata) 1 3 2 6
Four-toed salamander (Hemidactylium scutatum) 1 0 0 1
Northern redback salamander (Plethodon cinereus) 388 985 907 2,280
Northern spring peeper (Hyla crucfer cnrcfer) 0 1 0 1
Pickerel frog (Rana palustris) 0 1 0 1
Northern ringneck snake (Diadophis punctatus) 0 3 4
Eastern garter snake (Thamnophis sirtalis sirtalis) 3 0 3 6
Northern short-tailed shrew (Blarina hreuicauda) 0 0 1 1
IVhite-footed mouse (Peromyscus 1eucopu.s) 17 1 1 19
Southern redback vole (Clethrionomys gapperi) 13 0 0 13
= 4.ROO/>ear (2 boards/plot X 30 plots/stand X 16 stands X 5 surveyslyear)
Nomenclature follows Colhns (1997) for amphib~ans and reptlle5, and Ranks et al. $198;) for mammals
many plots. These values were transformed by Seven additional amphibian species were ob-
the square root of the number of observations served (Table 1). Red-spotted newts (Noto-
per stand per survey occasion to improve nor- vhthalmus uiridescens) were most often ob-
mality of dstribution (Mateu 1997), and then served as terrestrial efts and occurred in 4
analyzed via a repeated-measures design, where stands located near permanent ponds. Spotted
both year and survey occasion were indvidually salamander (Ambystoma maculatum) observa-
evaluated (Damon and Harvey 1987). The d s - tions were of immatures, presumedly dspersing
tribution of redback salamander total numbers from natal ponds. Northern dusky (Desmogna-
by size class and treatment class was analyzed thus fuscus) and northern two-lined salaman-
via the r x c contingency table (Conover 1971: ders (Euycea bislineata) breed in or near
149-154). All analyses were performed with streams (DeGraaf and Ruds 1983) and were
SYSTAT (kliilkinson et al. 1992). Significance observed on plots near streams.
level was a = 0.05. Neither thinning nor deer density had a sig-
nificant effect on the number of redback sala-
RESULTS mander observations (F1,lZ,deerl 0.998, P =
Redback salamanders were the most abun- 0.338; F1,lZtihi,,, 2.606, P = 0.132).Thinning
dant vertebrates observed under cover boards and the interaction of thinning and deer density
(Table 1).Of the 2,280 redback salamanders ob- were greater sources of treatment variance
served during the study, 1,950 occurred as sin- (MSEtkli,= 5.62; MSEdee,. th,n = 5.79) than
gle specimens under a board. There were 148 deer density alone (MSEd,,,. = 2.15). Redback
observations where 2 redback salamanders oc- salamander numbers were frequently lowest in
curred under a single board, 10 triples, and 1 the unthinned, low deer density treatment and
observation with 4 salamanders. Redback sala- generally greatest in the thinned, low deer den-
manders were observed under 2,111 of 14,400 sity treatment, but no treatment class consis-
boards turned during the 15 surveys. The max- tently had the greatest number of redback sal-
imum number of boards in a stand (n = 60) amander numbers every year (Table 2).
with redbacks was 43; no redbacks were ob- Numbers of redback salamander observations
served in >1 stand in several surveys. During varied significantly by year (F2.24 44.8, P <
the study, there was a redbacks salamander ob- 0.001). Numbers of redback salamanders in-
served under an average of 3.8 boards in a creased from 388 the first year of study to 985
stand. More than 75% of the redback salaman- the second year, and then declined slightly the
ders were sexually mature, 13%were immature, third year. Numbers of redback salamander ob-
and 12% were young-of-the year. No dfference servations by month (survey occasion) varied
was found in the distribution of redback sala- considerably, between a low of 16 salamanders
manders by size class across the 4 treatment in July 1997 to a high of 443 in October 1996.
classes (xZ6= 8.1, P = 0.231). The significant effect of survey occasion (FA,,
1176 COMMERCIAL Brooks
THINNINGN D SALAMANDERS
A J. Wildl. Manage. 63(4):1999
Table 2. Northern redback salamanders obsewed under cover boards by treatment, year, and month, Quabbin Reservation,
central Massachusetts, 1995-97.
Low deer densi? High deer densit)
No thin Thin No thln Thln
Month r 95% CI r 95% CI T 95% CI T 95% CI
= 75.9, P < 0.001) was the single largest source among treatments. Litter depths averaged 2 4 -
of variation in redback salamander observations. 33 mm by treatment class (Table 3). Fragmen-
The Pearson correlation between total redback tation depths averaged 32-37 mm, and humus
salamander numbers and precipitation for the 3 depths averaged 1 4 4 6 mm among treatment
days prior to the survey was 0.651 (Bonferroni- classes. The greatest fragmentation and humus
adjusted probability = 0.051). Generally, num- depths were recorded in 2 stands in the
bers of redback salamanders were highest in the thinned, high deer density treatment that con-
spring and fall surveys and lowest in the sum- tained small wetlands.
mer surveys. Density and area of CWD were greatest in
Organic soil horizon depths I d not I f f e r the thinned, low deer density treatment but not
Table 3. Organic soil horizon depths and coarse woody debris amounts by treatment, Quabbin Rese~ation, entral Massa-
Low deer density High deer densit).
No thin Thin No thin Thin
Attribute r SD r SD T SD r SD
Organic soil horizon depths (mm)
Litter 23.8 3.4 27.0 5.7
Fragmentation 35.9 6.6 36.1 5.2
Humus 24.3 12.3 15.4 9.6
Coarse woody debris counts (no./100 m2) by decay classa
1 0 0
2 0.05 0.1 0.35 0.342
3 3.05 0.943 3.35 1.29
4 3.25 0.998 5.15 1.05
All classes 6.4 1.681 8.85 1.32
Coarse woody debris area (m2/100 m2) by decay classa
1 0 0
2 0.017 0.035 0.061 0.061
3 0.4 0.323 0.52 0.316
4 0.434 0.29 0.826 0.291
All classes 0.854 0.588 1.406 0.52
See METHODS for defimtlon of coarse wood) debns decay cla~ses
J Wildl. Manage. 63(4):1999 TAI G
C O M ~ ~ E R C I H LS ~ I NA N D SALAMANDERS
statistically different from other treatment clas- of 66% of the basal area in the control, unthin-
ses (Table 3). There was very little sound CWD ned stands (Healy et al. 1987).
(Classes 1 and 2 ) , and generally equivalent Renroduction from these survivors could re-
amounts of CWD in the more decayed Classes place any salamanders lost during or subse-
3 and 4. quent to the thinning (Pough et al. 1987). Al-
The number of redback salamander obser- ternatively, resident salamanders may have been
vations differed among the 16 forest stands. able to survive within the patches where trees
Stands with high numbers of redback salaman- were removed by moving h t o deeper soil ho-
der observations occurred in all treatment clas- rizons or by tolerating the changed temperature
ses, whereas 3 of the 4 stands with the lowest and moisture conditions of the forest floor in
number of redback salamander observations oc- gaps created by thinning canopy trees (Messere
curred in the low deer density, unthinned treat- and Ducey 1998). It is unlikely that thinned
ment. Stands 11, 13, and 16 regularly had few stands weie functioning as sink habitats (Griffis
redback salamander observations, while stands and Jaeger 1998) populated by emigrating ma-
1, 12, and 18 had high numbers of obsen~ations. ture salamanders from adjacent unthinned
At the stand level (n = 16), total redback sala- stands, as the distribution of redback salarnan-
mander observations were positively correlated ders by size class in the thinned stands was the
(Spearman's rho) with the average area of CWD same as in the control stands. Size class is a
(rs = 0.568, P = 0.01-0.025), average count of good surrogate measure of age class and sexual
CWD (rs = 0.479, P = 0.025-0.05), and tall maturity (Sayler 1966). The equivalent abun-
woody stem ( 2 1 m tall) density (rs = 0.381, P dance of immatures in the treated stands indi-
= 0.05-0.1). cates that redback salamander reproduction was
similar to that in the control stands.
DISCUSSION The thinnings that were evaluated in this
The predominance of redback salamanders study are typical of many timber harvests
observed in this study was typical of terrestrial (Smith 1986) and are a recommended inter-
salamander suneys in the northeastern United memate treatment for oak in New England
States (Burton and Likens 1975a, Wyman 1988, (Hibbs and Bentley 1983) and elsewhere
Gibbs 1998),especially when artificial cover ob- throughout the range of northern red oak
jects were used for the surveys (DeGraaf and (Johnson 1994). The residual stands following
Yamasaki 1992. Bonin and Bachand 1997). The the thinnings of this study were similar in struc-
terrestrial herpetofauna of New England is not ture to the reverse-1 mameter distribution ex-
very diverse, especially compared to southeast- pected from uneven-aged or selection regen-
ern regions of the United States (DeGraaf and eration systems (Smith 1986). This structure
Rums 1983, \Vyman 1998),so the dominance of suggests that selection harvests, while not rec-
a single species was expected. ommended for stands dominated by northern
The silvicultural treatment evaluated in this red oak (Johnson 1994), would have a nonsig-
study was an intermediate treatment, not a re- nificant effect on redback salamanders (Pough
generation harvest where a larger portion of the et al. 1987, Messere and Ducey 1998), as ob-
canopy would be removed. The effects of thin- served following thinning in this study.
ning on forest-floor fauna associated with The lack of significant dfferences in the
closed-canopy forests would therefore be ex- numbers of redback salamanders between
pected to be less severe than even-aged regen- thinned and control treatments found in this
eration harvests and especially clearcutting, a study, and the results of 2 other wildlife studies
practice that has been shown to have negative conducted in these stands, suggests this com-
effects on some forest amphibians (Pough et al. mon silvicultural practice is compatible with the
1987, Petranka et al. 1993, deMaynadier and maintenance of faunal diversity. The effects of
Hunter 1995, Harpole and Haas 1999). Stands thinnings on vegetation structure and compo-
thinned to silvicultural guides (Hibbs and Bent- sition in southern New England oak stands
ley 1983, Johnson 1994) should have sufficient were sufficiently minor as to have no significant
residual canopy trees to create shaded refugia effect on small lnammal comnosition or-the rel-
for resident salamanders throughout the stand. ative abundance of dominant small mammal
In 1987, 4-12 years after thinning, basal area in species (Brooks and Healy 1988). Thinning in-
the thinned stands of this study was an average creased the richness of the breeding-bird com-
1178 COMMERCIAL I \ ~ \ IA NG
TH ~ D shL~\14\DFRb Brooks J Wildl Manage 63(4) 1999
munity by increasing the numbers of species ac- Bachand 1997). Additionally, precipitation dur-
tive in understory vegetation (DeGraaf et al. ing the spring months of 1995 was 3.9 (Mar),
1991). 3.8 (Apr), and 4.6 (May) cm below the 30-year
My failure to identify a lasting effect of normal for the area (National Oceanic and At-
chronic high numbers of white-tailed deer on mospheric Administration 1995). Drought con-
redback salamander numbers differed from ditions could have resulted in redback salaman-
other studies, which found significantly lower ders retreating to lower soil depths and avoiding
numbers of red-backed voles (Clethriononzys the drier forest floor (Heatwole 1962) and, con-
gapperi) and short-tailed shrews (Blarina brev- sequently, the cover boards, resulting in the re-
icauda) and canopy-gleaning birds in the high duced numbers observed in the May and June
deer densit>. stands (Brooks and Healy 1988, surveys in 1995.
DeGraaf et al. 1991). Decreased bird popula- Differences in numbers of redback salaman-
tions have been obsewed in other locations with der observations also differed significantly
chronic high densities of white-tailed deer, among survey occasions. Numbers of redback
which resulted in structural and compositional salamander observations were generally higher
changes in mid- and understory vegetation due in spring and fall surveys. Soil moisture in New
to overbrowsing (decalesta 1997, McShea and England is typically high in the spring and fall
Rappole 1997). and low in the summer (Hornbeck and Leak
Residual effects of thinnings on understory 1992),reflecting seasonal precipitation patterns,
vegetation structure and compktion continued and could account for seasonal patterns in red-
to be observed in 1994, >12 years after the back salamander numbers. Optimum tempera-
stands were thinned, with greater densities of tures for redback salamander surface activity
seedlings and saplings and greater forb cover in also occur in the spring and fall (Burton and
thinned stands. Likewise, residual effects of Likens 1975a). Some variation in redback sala-
chronic high deer numbers were observed in mander numbers among indvidual survey oc-
lower numbers of seedlings and saplings, great- casions was also due to precipitation during the
er fern cover, and lesser shrub cover 3 years week prior to the survey. Precipitation can tem-
after deer densities were reduced by hunting. porarily increase moisture levels of the forest
These residual effects in understory composi- floor, allowing salamanders to be active on the
tion and structure had no apparent effect on ground surface and take refuge under cover
redback salamander numbers. boards (Jaeger 1980).
No significant residual effects of thinning
were observed on CWD density or area. If dif- MANAGEMENT IMPLICATIONS
ferences had occurred immediately following Differences in the composition and structure
thinnings, natural recruitment of CWD has of residual vegetation, in the forest floor cover,
since removed any differences that may have and in the observations of redback salamanders
existed between thinned and unthinned stands leave the impression that a stand dsturbance,
Likewise, no differences were found in the where a large percentage of the canopy is re-
thicknesses of soil organic horizons between tained and that results in an increase in cover
treatment classes. Pough et al. (1987) demon- of understory vegetation, would result in no
strated that depth of leaf litter is an important long-term effect on redback salamanders. This
predictor of aboveground salamander activity. I effect was observed following thinning and fol-
do not know how litter depths changed imme- lowing chronic overbrowsing by white-tailed
diately following the thinnings (Ash 1995), but deer. While not measured, it is likely that in-
if reduced, they had recovered by the time of creased low herbaceous cover increased humid-
this studv. ity and reduced temperatures at the ground lev-
The largest sources of variation in numbers el, which improved microhabitat quality for red-
of redback salamander observations were tem- back salamanders (Heatwole 1962, Pough et al.
poral effects. The dramatic increase in numbers 1987). These findings demonstrate that the
between the first and second year of study may maintenance of sustainable redback salamander
have been a result of aging of the boards, the populations are compatible with timber har-
length of time of board placement, or an in- vesting if precautions are taken to minimize dis-
crease in carrying capacity of the habitat caused turbance to the overstory canopy and to forest
by the addition of cover objects (Bonin and floor microhabitats.
J. Wildl. Manage. 63(4):1999 C ~ M M E R C IT H I N N I NAsD SALAMANDERS
AL G Brooks 1179
ACKNOWLEDGMENTS on southern \'ancouver Island, British Columbia.
Ileipetological Conservation 1:161-173.
R. M. DeGraaf suggested the study. I very DECALESTA, S. 1997. Deer and ecosystem man-
much appreciate the help of K. L. Kripp, J. M. agement. Pages 267-279 in W J. McShea, H. B.
Green, and T. J. Maier in the physically de- Undenvood, and J. EI. Happole, editors. The sci-
ence of overabundance: deer ecology and popu-
mandmg task of distributing the cover boards. lation management. Smithsonian Institution
T. J. Maier, C. E. Baker, and L. P. Levesque Press, I'i'ashington, D.C., USA.
helped with the many surveys of fauna under DEGH.AAF, . hf., 1j7. I. IIEALY,AND R. T. BROOKS.
the boards. C. E. Baker and L. P. Levesque con- 1991. Effects of thinning and deer browsing on
breeding birds in New England oak wood1:inds.
ducted the CWD survey. Vegetation survey data Forest Ecology and Management 41:179-191.
were provided by LV. M. Healy. R. M. DeGraaf, , AKD D. D . RUDIS. 198:3. Amphibians and
LV. M . Ilealy, P. G. deMapadier, J. W. Petranka, reptiles of New England: habitats and natural his-
R. L. Wyman, and 2 anonymous referees pro- toy. Lniversit) of Massachusetts Press. -4mherst.
vided constructive reviews of drafts of the man- Massachusetts, USA.
. .4ND Xf. Y.~&IASAKI. A nondestnlctive
uscript. technique to monitor the relative abundance of
terrestrial salamanders. Wildlife Societ). Brllletin
LITERATURE CITED 20:260-264.
DEMAYNADIER, .\XI>M. L. HUNTER, 1995. Jn.
ASH, A. N. 1995. Effects of clear-cutting on litter pa- The relationsltip between forest manag~ment nd a
rameters in the southern Blue Ridge Mountai~ls. of
amphibian ecolog): a r e ~ i e w the North Ainer-
C:astanea 60:89-97. icail literature. En\ironmental Re\ie\v 3:230-261.
BANKS, . C., R. rj'. V.ICDIARMID,N D A. L. GARD-
R A ~ ASD -1998. The effects of sil\~cl~ltural
~ER. 1987. Checklist of vertebrates of the United edges on the distribution and abundance of am-
States, the U.S. Territories, and Canada. U.S. phil~iansin Maine. Consen-ation Biolog) 12340-
Fish and \Vildlife Service Resource Publication 352.
166. EVRE.F. H. 1980. Forest cover types of the United
B o s r s , J., 4ND Y. BACHAND. 1997. The use of artifi- States and Canada. Society of American Forest-
cial covers to sun-ey terrestrial sala~riandersin er$. LVashirigton, D.C., USA.
Q ~ ~ e b eHerpetological Conservation 1:179-179. FELLEHS, X i . , AND i A. DROST. 1994. Samplitlg
G. ; .
BROOKS, . T., AND \L7. V.1. HEALY.1988. Response
R ~ 4 t hrtificial cover. Pages 14C150 in \V. K. Hey-
of small mammal coni~nunitiesto sil\:icultural er. XI. A. Donnellq: R. \V hlcDiarmid, L. C. Hay-
treatments in eastern hardwood forests of West ek, and hl. S. Foster, editors. Measuring and
Virginia and Massacl~usetts.Pages 313-318 in R. monitoring biological diversity standard methods
C . Szaro, K. E . Severson, and D . R. Patton, tech- for amphibians. Srnitlisoniall Institt~tionPress,
nical coordinators. Proceedings of the syrr~posinm \t7:tshington. D.C., USA.
on management of amphibians, reptiles, and GIBBS,J. P. 1996, Distribution of woodland amphib-
small mammals in North America. U.S. Forest ia11s along a forest fragmentation gradient. Land-
Service General Technical Report RM-166. scape Ecolog). 13263-268.
BURTON, hl., .AKD G. E. LIKENS. 975a. Salaman- GRIFFIS,b1. R., %ZD R. G. J.\EGER.1998. Competi-
der populations and biomass in the Hubbard tion leads to an extinctio~i-pronespecies of sala-
Brook Experimental Forest, New Hampshire. mander: interspecific territorialit) in a rnetapopu-
Copeia 1975:541-546. lation. Ecology 79:2494-2502.
, .%SD -1975b. Energy flow and nutri- H ~ ~ R s r oN ., G. 1987. <:on?inunity ecolog- and sal-
ent cycling in salamander populations in the amander guilds. Cztmbridge UniversiQ- Press,
Hnhbard Brook Experimental Forest, New Camlxidge, United Kingdorn.
Hampshire. Ecolog?; .56:1068-1080. HARMON,V. E., A N D J. SEXTON. 996. Guidelines
CH.~ZA[., C:., AND P. H. NIE\VIARO~'SKI. Re-
A. 1998. for lneasurements of woody detritus in forest eco-
sponses of mole salamanders to clearcutting using systems. U.S. Long-Term Ecological Research
field experiments in forest management. Ecolog- h e w o r k Office P~~blic;ition20.
ical Applications 8:1l3:3-1143. ; .
HARPOL~,:, 5 . .ASD i A. H.4.i~.1999. Effrcts of
COLLINS, . T. 1997. Startdard com~nonand current se\.en sil~icnltnraltreatments on terrestrial sala-
scientific names for North American aalnpliibians manders. Forest E c o l o k ~ and V.fanagenient 114:
and reptiles. Fourth cdition. Societ) for the Study 349-,356.
of Amphibians and Reptiles, Herpetological Cir- HEAL).\Ir. E 1 1994. Ecosystem management: policies
cular 25. and procedures for effective implementation in
C O N O ~ X R . J. 1971. Practical nonparametric statis-
\V. the Northeast. Nortlieast \t7i1dlife 51:81-89.
tics. John tWley & Sons. New York, New York, 1997. Influence of deer on the structure and
LJSA. composition of oak forests in central Massachu-
D.AMOS,R. A,, .4ND \V. R. HARVEY. 1987. Experi- setts. Pages 2-19-5266 i n \V. J. SlcShea, 11. B. Un-
mental design, ANO\'A, and regression. IIarper denvood, and J. H. Kappole, editors. The science
& How, New York, h'ew York, USA. of overab~~ndance: deer e c o l o p and population
D.%\-IS, h1. 1997. Non-dismptive monitoring of ter-
T. management. Smithsonian Institution Press,
restrial salamanders with artificial cover objects \V:~shington, D.C., USA.
1180 THINNINGND SALAMANDERS
COLISIERCIAL A Brooks J. \ViIdl. Manage. 63(4):1999
, R. T. BROOKS, ASD P. J. LYONS. 987. Deer
1 District Commission, \tratershed blanagement,
and forests on Boston's municipal watershed after Boston, hlassachusetts, USA.
50 years as a wildlife sanctuary. Pages 3-21 in PET RANK.^, J. bv., hf. E. ELDRIDCE. AND K. E. HA-
Proceedings of papers presented at the sympo- LEY.1993. Effects of timber harvesting on south-
sium deer, forestry, and agriculture: interactions ern Appalachian salamanders. Consen.ation Bi-
and strategies for management. Allegheny Society ology 7:363-370.
of Arnerican Foresters, Warren, Pennsylvania, F
POUGH, . H., E. M. ShllTH, D. H. RIIODES, AND A.
USA. COLUZO.1987. The abundance of salamanders
HEATWOLE, 19@2.Environmental factors influenc-
H. in forest stands with different histories of distur-
ing local distribution and activity of the salaman- bance. Forest Ecology and Management 20:l-9.
der, PletJ1(1d(111inereus. Ecology 43:460-472.
S.~YLER; 1966.The reproductive ecology of the red-
HIBBS,D . E., AND \V. R. BENTLEY. 1983. A manage- backed salamander, Plethoclor~citlcreus, in Mar-
ment guide for oak in New England. Cooperative viand. Copeia 1966:183-193.
Extension Senice, University of Connecticut, SMITH, D. M. 1986. The practice of sil~icdture. ohn J
Storrs, Connecticut, USA. \17iley & Sons, New York, New York, USA.
HORNBECK, ?\', AND \V. B. LEAK.1992. Ecology SOIL COXSERVATIOX 1
SERVICE. 967. Soil survey of
and management of northern hardwood forests in Franklin County Massachi~srtts,U.S. Depart-
New England. U.S. Forest Senice General Tech- ment of Agriculture, Washington, D.C., USA.
nical Report NE-159. . 1989. Soil sun-ey of IIampden and Hamp-
J.%EGER, G. 1980. Microhabitats of a terrestrial sal- shire counties. Massachusetts. eastern art. U.S.
amander. Copeia 1980:265,268. Department of Agriculture, ~ a s h i n g t h n ,D .C.,
JOHYSON, S. 1994. The silviculture of northern red
oak. Pages 33-68 in J. G. Isebrands and R. E. U.S. FOREST SERVICE. 1995. Field instructions for the
Dickson, compilers. Biology and silviculture of fourth inventory of Maine. Northeastern Forest
northern red oak in the north central region: a Experiment Station, Forest Ir~ventoy and Anal-
synopsis. U.S. Forest Senice General Technical ysis, Radnor, Pennsylvania, USA.
Report NC-173. W7ALDICK, R. 1997. Effects of foresty practices on
MATEU, . 1997. Methods of assessing and achieving
J amphibian populations in eastern Nortlt America.
normality applied to endronmental data. Envi- 1
Herpetological Conservatio~l :191-20.5.
ronmental Management X:767-777. D
WALLER, . M., A N D ' l S. AL\,ERSON.
l< 1997. The
SICSHEA, J., AND J. H. RAPPOLE. 1997, Herbivores white-tailed deer: a keystone herbivore. ivildlife
and the ecology of forest understory birds. Pages Society Billletin 25:217-226.
298-309 in \V. J. McShea, H. B. Underwood, and MT.4KREN, R. J. 1995. The challenge of deer over-
J. H. Rappole, editors. The science of overabun- abundance in the 21st century. FViIdlife Societ).
dance: deer ecology and population management.
WILKINSON, M. HILL,J. P. ~VELNA, D G. K. AN
Smithsonian Institution Press, Washington, D.C.,
BIRKENBEUEL. SYSTAT for \Vi~ldows:sta-
tistics. Version 5 edition. SYSTAT, Evanston, 11-
MESSERE, ASD P. DUCEY. 998. Forest floor dis-
tribution of northern redback salamanders, PletR- R.
WYMAN, L. 1988. Soil acidity and moisture and the
odotl cinercrru, in relation to canopy gaps: first distribution of amphibians in five forests of south-
year following selective logging. Forest Ecology central New York. Copeia 1988:394-399.
and Management 107:319-324, . 1998. Experimental assessment of salaman-
NATIONAL OCEANIC .4ND ATMOSPHERIC DMISIS- A ders as predators of detrital food wells: effects on
TRATIOS. 1995. Clilnatological data: annual sum- invertebrates, decompositioll and the carbon cy-
mary, New England, 1995. National Climatic cle. Biodiversity and Conservation 7:641450.
Data Center, Ashe\ille. North Carolina. USA.
O'CONNOR, T. KYKER-SNO\VMAN, ~ Y S ,AND P. L Y Receiued 8 September 1998.
B. SPENCER. 995. Quabbin Watershed: MDC Accepted 18 April 1999.
land management plan, 1995-2004. Metropolitan Associate Editor: Diefenbach