Short-Term Breeding Bird Response to Two Harvest Practices in a Bottomland Hardwood Forest

2004·6.8 MB·English

by  Charles A Harrison

Wilson Bulletin 1 16(4):314-323, 2004
SHORT-TERM BREEDING BIRD RESPONSE TO TWO HARVEST
PRACTICES IN A BOTTOMLAND HARDWOOD FOREST
CHARLES A. HARRISON* ^ AND JOHN C. KILGO^
—
ABSTRACT. Clearcutting is the preferred timber harvest method in bottomland hardwood forests because
it is most likely to result in regeneration of preferred species. However, clearcutting generally has negative
impacts on forest birds. Patch-retention harvesting may provide similar silvicultural benefits, but its effects on
birds are unknown. We surveyed breeding birds in uncut control, clearcut, and patch-retention treatment areas
(1 1-13 ha) for one season prior to harvest and two seasons postharvest in a bottomland hardwood forest in the
Lower Coastal Plain of southeastern South Carolina. Bird observations recorded along line tran.sects were ana-
lyzed using the software Estimates to estimate species richness and program Distance to estimate densities. We
found greater species richness and bird densities in the patch-retention treatment than in the clearcut in both
postharvest seasons. We detected no forest-interior birds in the clearcut after the harvest, but by the second
postharvest season in the patch-retention treatment, the density offorest-interior birds had returned to approxi-
mately half of its preharvest level. Thus, based on density response, patch-retention harvesting appears to be
less detrimental to forest birds than clearcutting. However, additional work is needed to determine whether
retained patches influence avian survival and productivity. Received 30April 2004, accepted30 October2004.
Bottomland hardwood forests in the south- the decline of birds (Pashley and Barrow
eastern United States serve as critical breeding 1993). Indeed, much ofthe extant bottomland
habitat for numerous avian species, including hardwood forest is now in streamside man-
many considered by Partners in Flight to be agement zones and drainages <50 m wide
of high conservation concern (Hunter et al. (Kilgo et al. 1998).
1993, Rich et al. 2004). Historical loss and Ofthe approximately 194 million ha offor-
fragmentation of these forests by conversion ested land in the United States during 1989,
to agriculture, development, and other activi- 85 million ha were in the southern U.S. (Wig-
ties, concurrent with possible functional ley and Sweeney 1993). The forest products
changes in the remaining forests, have likely industry (18.8%) and individual landowners
contributed to the reduction in bird popula- (71.1%) combined held title to nearly 90% of
tions (Pashley and Barrow 1993). According these forested lands (Wigley and Sweeney
to the National Resource Inventory of 1992 1993). Ownership of bottomland hardwoods
(Shepard et al. 1998), the area covered by was apportioned in roughly the same manner,
wooded palustrine wetlands in the South de- and nearly 20 million ha of non-federally
clined by only about 1.5% from 1982 to 1992, owned, palustrine-forested wetlands were
a marked reduction in the rate of loss com- present in the South (Shepard et al. 1998). The
pared with that in preceding decades. Despite primary use of these lands is for timber pro-
this apparent stabilization of forested wetland duction (Wigley and Sweeney 1993). Ifexist-
area, a much higher proportion of woodland ing bottomland hardwood forests are to re-
Neotropical migrant species was in decline in
main a viable resource for Neotropical migra-
the eastern United States over the period
tory birds, management options that minimize
1982-1991 than during 1966-1979 (Peterjohn
negative effects on breeding birds, but are ac-
and Sauer 1994). Thus, some aspect of the
ceptable to the forest products industry and
quality, not just quantity, of the existing bot-
tomland hardwood forest may be a factor in private landowners, need to be identified and
their use eneouraged.
From a silvicultural perspective, clearcut-
‘ USDA Forest Service Southern Research Station, ting is the favored means of harvesting these
Center for Forested Wetlands Research, 2730 Savan- forests (Clatterbuek and Meadows 1993,
UnaSh^DUHASwyDF.oA,reCFshotarreSlseetrsvtSioencr,ev-iSScCaevaS2on9un4t1ah4he,rRniUvSeRAre,.sePa.Or.chBoStxat7i0o0n,, MsoenasdfoowrsitsanadppSetaalntiusrtfha1t99i7t)i.s Athmeomnegthtohde rtehaa-t
New Ellenton, SC 29809, USA. best promotes regeneration ofshade-intolerant
^Corresponding author; e-mail: [email protected] species such as oaks (Quercus spp.; Clatter-
314
Harrison and Kilgo • BIRD RESPONSE TO HARVEST PRACTICES 315
buck and Meadows 1993). The abundance of were higher in harvested areas (Baker and
f
shrub-successional birds, several of which are Lacki 1997). In Canadian boreal forest, seven
of conservation concern (Rich et al. 2004), as of eight bird species lost from clearcuts were
I
' well as total bird abundance, can be as great retained at low abundance levels in partial
or greater in clearcuts (Hurst and Bourland cuts (Norton and Hannon 1997). In hardwood
I
1996) and other early successional habitats forests ofWest Virginia, avian abundance and
I (Buffington et al. 1997) as in mature bottom- nesting success were comparable for most
'
land hardwood forest. However, species rich- bird species tested among control areas, two-
i
ness and diversity are highest in mature bot- age harvest, and clearcut treatments 15 years
tomland hardwoods (Hurst and Bourland after harvest (Duguay et al. 2001). Both Nor-
1996, Buffington et al. 1997), and the impacts ton and Hannon (1997) and Duguay et al.
ofclearcutting on most species that preferma- (2001) concluded that the two-age method
ture forest-interior conditions are negative was a viable conservation alternative. How-
(Hurst and Bourland 1996, Baker and Lacki ever, from a timber management perspective,
1997). Thus, we need alternatives to clearcut- shelterwood methods can be difficult to im-
ting that retain the advantages for forest re- plement because of the critical importance of
I
generation but that are less damaging to bird choosing the appropriate establishmentcutting
II species that inhabit mature forest. intensity (Meadows and Stanturf 1997).
Several studies have been conducted to A third alternative, representing a hybrid of
I
evaluate the impacts on birds of one such al- group selection and clearcutting methods, is
I ternative, group-selection harvest. In this patch-retention harvest, in which residual
!
method, small groups of mature trees are trees and snags are retained in small patches
cleared from a stand at regular spacing inter- that mimic or actually represent remnants of
vals (Meadows and Stanturf 1997). Moorman the original forest. Patch-retention harvesting
and Guynn (2001) concluded that when ade- is appealing to forest managers, because leav-
!
quate mature forest was left unharvested, the ing patches of uncut forest is operationally
abundance of most breeding forest-interior easier (for equipment such as feller-bunchers)
species was not impacted by group-selection than attempting to retain a high basal area of
harvest. Similarly, Moorman et al. (2002) de- more evenly distributed trees (Tittler et al.
termined that the productivity of a represen- 2001; J. P. Martin, MeadWestvaco Corpora-
tative forest-interior species, the Hooded War- tion, pers. comm.). Additionally, the incidence
bler (Wilsonia cithnci), was minimally affect- of windthrow may be lower for trees in patch-
ed. Kilgo et al. (1999) concluded that group- es compared with isolated residual trees. Re-
selection harvest gaps benefited many species ported impacts to forest bird communities
of migrating forest-interior birds because they have been less severe in patch-retention har-
I
used the early successional patches during vests than in clearcuts within boreal forests of
passage. However, when applied in the strict- Canada (Schieck et al. 2000, Tittler et al.
,,
est sense (gap size no greater than 0.5 ha), this 2001) and aspen forests in Minnesota (Merrill
technique usually favors regeneration of et al. 1998). However, this method has not
j stands that are dominated by low-value, been widely tested.
shade-tolerant trees because of limitations on Our objective was to compare the effects
I light availability (Meadows and Stanturf on bird species composition of retaining
( 1997). patches of bottomland hardwood forest within
I Two-age harvest prescriptions, in which a a clearcut with performing a traditional clear-
I
predetermined quantity of basal area is re- cut or leaving the Idlest intact (unharvested
[ tained in an even distribution across the site control). In particular, we addressed whether
I
I (a modification of the shelterwood method), the responses of individual species and/or avi-
have also been evaluated (Baker and Lacki an habitat-use groups differed between a
1997, Norton and Hannon 1997, Duguay et al. patch-retention cut and a clearcut area, and
2001 ). In Kentucky, such prescriptions did not whether bird species composition changed in
alleviate negative effects ofclcarcuttifig on the these areas from preharvest to 2 years post-
abundance of certain forest-interior birds, but harvest. In effect, we sought to determifie
some indices of bird community structure whether bird species composition of the
316 THE WILSON BULLETIN • Vol. J16, No. 4, December2004
patch-retention area more closely resembled We surveyed breeding birds along line tran-
that of the unharvested control, particularly sects (Bibby et al. 2000) for 1 year preharvest
with respect to birds that depend on mature (1999) and 2 years postharvest (2000-2001).
forest-interior habitat. We arranged transects such that each treat-
ment area was completely covered without
METHODS
duplication of coverage, assuming a 50-m de-
The study was conducted on a 350-ha area tection zone on each side of the line. We con-
located in the Lower Coastal Plain of south- ducted three to four counts each year between
eastern South Carolina on the floodplain ofthe 15 May and June. Single-observer surveys
1 1
fourth-order, blackwater Coosawhatchie River began around 06:00 EST and continued until
(Burke et al. 2003), a relatively small, anas- all treatments had been surveyed, usually
tomosing stream that drains an area of ap- around 1 1:00. During each survey, the ob-
proximately 1,000 km^. Topography in the server proceeded along the transect, stopping
low-relief (<2 m) floodplain is characterized only to record detections. The observer
by a network of slightly elevated hummocks mapped locations of all birds encountered by
and scour channels (Burke et al. 2003). Water sight and/or sound and estimated the perpen-
tupelo {Nyssa aquatica), swamp tupelo {N. dicular distance (0—50 m, to the nearest 10 m)
sylvatica var. hiflora), sweetgum {Liquidam- from the transect line to detected individuals.
har styraciflua), bald cypress (Taxodium dis- T1o99a7c)count for the potentially confounding ef-
tichum), laurel oak {Quercus laurifoUa), and fect of time of day, we varied the order in
red maple {Acer rubrum) dominate the plant which the treatments were surveyed.
communities of the floodplain (Burke et al. We estimated species richness using the
2000). Where present, the understory (in un- software Estimates, ver. 6.Obi (Colwell
disturbed forest) consists of widely scattered Based on detection data, this program
.
patches of Vaccmium spp., Sabal minor, and provides values for several species-richness
Anindinaria gigantecr, very little under- or e1s9t9i8m)ators. We present the first-order jack-
midstory structure exists. knife estimates because they are robust and
Three 1 1- to 13-ha treatment areas were es- have performed well in other studies (as cited
tablished in the fall of 1999: an uncut control, in Nichols et al. 1998, Hellmann and Fowler
a patch-retention area, and a clearcut. This 1999). We obtained density estimates using
size range approximated an operational har- program Distance, ver. 3.5 (Thomas et al.
vest. The three treatment areas were arranged We used grouped data (10-m intervals)
.
linearly, parallel to the direction ofwater flow. stratified by treatment and year. Although they
They were approximately equidistant (300— were not independent, we treated each visit to
400 m) from the main body of the Coosa- a given treatment within a season as a “rep-
whatchie River and were similar with respect licate.” Upon determination of the most ap-
to soils, hydrology, and preharvest plant com- propriate model for the detection function
munities (Burke et al. 2000, Eisenbies and (uniform, half-normal, or hazard rate) using
Hughes 2000, Murray et al. 2000). A 100-m likelihood ratio tests. Distance provides an es-
forested buffer separated the patch-retention timate of density and error (Buckland et al.
from the clearcut area, with a somewhat nar- 1993). Because Distance bases each density
rower and more irregular buffer between the estimate on a unique detection function, the
uncut control and patch-retention areas. In the estimates can be compared among sites with
patch-retention treatment, three “patches” of differences in detectability. We tested for an-
two sizes (two 0.20 ha and one 0.61 ha) were nual differences in estimated density within
left uncut, one each in an area representative each treatment area by determining whether
of a convex, concave, or flat landform. Thus, the 95% confidence intervals overlapped; we
1.01 ha offorest was retained within the 13.1- accepted as different those confidence inter-
ha patch-retention treatment area. Using the vals that did not overlap (Hodges and Kre-
average basal area for the study site (46 m7 mentz 1996). We compared density estimates
ha; Burke et al. 2003), the residual basal area for all birds combined, for individual species
of the patch-retention treatment was 3.6 mV with at least 25 observations, and for four avi-
ha. an habitat-use groups: forest interior (I), in-
—
Harrison and Kilgo • BIRD RESPONSE TO HARVEST PRACTICES 317
terior-edge (I-E), field-edge (F-E), and edge- Speciesrichness
shrub (E-S) (Whitcomb et al. 1981). 35 r - —- -- -
Logistical constraints prevented us from
replicating our treatment units in an experi-
mental manner; harvesting multiple units on
the study area was not feasible or desirable
for the landowner. We used a before-after,
control impact (BACI) design (Johnson 2002),
in which we sampled both before and afterthe
harvest treatments on both control and treat-
ment areas. This design was used in an at- Uncutcontrol Clearcut Patchretention
tempt to minimize the effect of variables un- Treatment
related to the treatments. Nevertheless, our re-
sults must be vie—wed with caution, since
lacking replication we do not know whether
Density
they would be applicable on other sites.
60 T
RESULTS
50
Prior to the harvest (1999), the estimated
species richness of breeding birds was lower
in the control area than in the patch-retention
area (Fig. 1). However, we could not make
statistical comparisons between the estimate
from the clearcut area in 1999 with the others,
because it had a variance of 0 (the first-order
jackknife estimator uses the number ofunique Uncutcontrol Clearcut Patchretention
species recorded on multiple visits; since we Treatment
detected two unique species on each visit, the FIG. 1. Speciesrichnessanddensity (mean ± 95%
variance of these identical values was 0). Es- confidence intervals) for breeding birds in three har-
timated total density did not differ among the vest treatment areas in a bottomland hardwood forest
in South Carolina, 1999-2001. No confidence interval
three treatment areas prior to harvest (Fig. 1). is given for species richness in the clearcut area in
Following harvest, total density was similar 1999 because the variance was 0 (see text for expla-
and species richness actually increased in the nation).
uncut control area. However, both measures
declined immediately after harvest (2000) in
both the clearcut and patch-retention areas tected postharvest were recorded in both post-
(Fig. 1). The decrease was especially large in harvest seasons (Tables 1 and 2), whereas in
the clearcut, where estimated species richness the patch-retention area, 11 of the 21 (52%)
declined from 25.0 to 9.3, and estimated den- species detected postharvest were recorded in
sity fell from 33.0 to 2.8 pairs/10 ha. In the both seasons.
second postharvest year (2001), density in- In the patch-retention area, 35 of 43 (81%)
creased in both the clearcut (Fig. 1 ) and patch- birds observed in 2()()() were recorded within
retention (nonsignilicantly) areas. Species retained forest patches, but in 2001, only 37
richness rebounded in the patch-retention ol' 89 (42%) observations occurred in retained
area, rising from 15.3 to 25.0, but not in the forest patches. Although the total number tif
clearcut. F^'rom preharvest to 2 years posthar- birds observed in the patch-retention area es-
vest in the clearcut, 15 species of the forest- sentially doubled, the number ol' birds ob-
interior or interior-edge groups disappeared, served within the retained patches remained
whereas 5 species of these groups disappeared about the same.
in the patch-retention area ( fables 1 and 2). We estimated densities for 1 1 species (Ta-
During the postharvest period, species com- ble 1 ). Iiuli\idual species appeared to respond
position also changed in both treatments. In to the clearcut treatment in different ways.
the clearcut, only 2 of (189F) species de- T hree of the species had disappeared after
1 1 1 1
—
^ ^^ 1^ ^ ^^ H^^ ^ ^^
318 THE WILSON BULLETIN • Vol. 116, No. 4, December2004
ir-n 'It qo oin ON in iqn mq
bottomland dor—i NdO doimn1 dNO dorri-1 rdirn dOd001 cOnN qrd- din dOn1 <dqn OdiNn1; NrqOl <0dN0 <(qNN mdI" (4.36-11.3) 5.25 (2.21-12.5)
a '—" '—"
in
areas 0d0 iro(Nn- N(ON qNNiOOn1 r9r40n fqodOo1 d'itn oO0di0n1 dNfOO oqo(^N1 ©m mmdq1 qO(N d00 qOd1 0
d (dN 't d d dr'' 0d0 d
treatment
value.
patch-retention1t9h9e9 mdq qONqNO1 dOn dqd1 in Odn1‘ iNonO dqOqN1 'It qdr7'' in dmdO’-)1H (rqNi d0(qqN01 0qi0n Ndo7(NO N(qON idd7n 0
from
in (N
and different do1 O(NN qd1 o dqin 2.37 (1.37-4.10)
d (dN
clearcut,
(4.81-11.0)
P
significantly in 0;Z0Tn dqin m dq (N dqin in 00
control, © dNO m1 ©fS (ON1 d't 0O01 d(N (ON1 o dNO m1 © o 0
d(N d d d d(N
— —
uncut are ' ^ ' '
sipenciesareas) qOm7N <dmr^dqN"1 dN0O0 OO0qdN0On1 0Nq0O q('70^Nt0 dN0O0 O<dmq^N1 d0r-0 N'dd7G^Ot) in 'Oddq—^t1 dq 0(<dNONN0ON1 dq (dqqdN1 o 0.29
treatment 'i—n' (0.01-10.3)
sbeilerctded (within NidqiOnn1 0N0O d0dqS0' N0O0 OdOdoqnN1 dq Oqdmiqn1 0N0O (dqsiiNnn O'tN dd0rqs0-1 m NddiqPOn OOONn oqdm^1' d0G0) dNd(^ON1 o 0
of '—
entries
interval)]Bold-faced d'dqN^tO1 (N dOd'^I1t dq'It d((qONN1 dqNO dON0(NnO01 'OtN dom'NdtO1^dq(N qodqirn-1 CN dOmd'i—n11 oqd O(dq7NN^dr- d<qdmq^N1 o 0
— NO
' ^ ^ ' '
c(onf9ide5nce%1999-2001. (dO7N) OONN odNqO1 (N ddii7nn d'qt rqd1—N1 md't i0'dtn01 'r-t- 0q0Nmd00O1 (N ddGmin)1 0d0 dqi7mqn 'qdt 'ditn1 o 0
q.
ha
Carolina,
[pairs/10
Densities Siounth mOwIN Ic^O M(4cOU>x sCo^X E^C^J'Ss r(O-N ^SO o^ r- ai x2: trichas) (27)cyanea)
o a
species
1. forest Bunting
"S s o 2 (Geothlypis (Passerina
cd I Q —
CQ O ^
TABLE hardwood O U o Edge-shrub Indigo
oi
Harrison and Kilgo • BIRD RESPONSE TO HARVEST PRACTICES 319
I
TABLE 2. Number ofbird observations in uncut control, clearcut, and patch-retention treatment areas in a
bottomland hardwood forest in South Carolina, 1999-2001. Included are all species not analyzed using program
Distance (Thomas et al. 1998).
Uncutcontrol Clearcut Patchretention
Species 1999 2000 2001 1999 2000 2001 1999 2000 2001
Eorest-interior species
Barred Owl {Strix varia) 0 0 1 0 0 0 0 0 0
Pileated Woodpecker {Dryocopospileatus) 2 1 5 0 0 0 0 0 1
White-breasted Nuthatch {Sitta carolinensis) 0 11 6 1 0 0 2 0 2
Pine Warbler (Dendroicapinus) 0 0 0 2 0 0 1 0 0
American Redstart (Setophaga ruticilla) 0 0 1 0 0 0 0 0 0
Ovenbird (Seiurus aurocapilla) 0 0 1 0 0 0 0 0 0
Kentucky Warbler (Oporomisformosus) 0 0 3 0 0 0 0 0 0
Hooded Warbler (Wilsonia citrina) 0 1 0 0 0 0 0 0 0
Interior-edge species
Yellow-billed Cuckoo (Coccyzus americanus) 4 0 2 3 0 0 3 0 0
Downy Woodpecker (Picoidespubescens) 3 7 8 2 0 0 0 3 0
Eastern Wood-Pewee {Contopus virens) 0 1 0 0 0 0 0 4 1
White-eyed Vireo {Vireo griseus) 1 2 0 0 1 0 1 0 0
Yellow-throated Vireo {Vireoflavifrons) 3 4 4 4 0 0 0 0 1
Blue Jay {Cyanocitta cristata) 0 0 0 2 0 0 1 0 0
Prothonotary Warbler {Protonotaria citrea) 1 5 2 5 0 0 2 0 0
Summer Tanager {Piranga rubra) 0 2 5 1 0 0 0 2 1
Northern Cardinal {Cardinalis cardinalis) 5 1 1 4 2 3 1 1 4
Edge-shrub species
Red-headed Woodpecker {Melanerpes erythrocephalus) 0 2 0 0 0 0 0 0
1
Field-edge species
Killdeer {Charadrius vociferus) 0 0 0 0 0 0 0 2 0
American Crow {Corvus brachyrhynchos) 3 2 2 2 0 0 3 0 1
Yellow-breasted Chat {Icteria virens) 0 0 1 0 0 0 0 0 0
Eastern Towhee {Pipilo erythrophthalmus) 0 0 1 0 0 0 0 0 0
Field Sparrow {Spizellapusilla) 0 0 0 0 0 0 0 0 1
Blue Grosbeak {Passerina caerulea) 0 0 0 0 0 2 0 0 0
Wetland species
Wood Duck {Aix sponsa) 0 0 0 0 0 0 0 0
1
the hrst postharvest year, and seven (e.g., Aca- year and accounted for 70% of the total ob-
dian Flycatcher, Empidonax virescens; Red- servations.
eyed Vireo, Vireo o/ivaceus; and Northern Pa- In the patch-retention treatment, as in the
rula, Parula americanw, all Neotropical mi- clearcut, the abundance of most species we
grants) had disappeared by the second post- analyzed appeared to decline in the first post-
harvest year. Only two species detected more harvest year, but only Acadian Flycatcher was
than once before the harvest, Carolina Wren not observed at all. During the second post-
(Thryothorus ludovicianus) and Blue-gray harvest season, however, abundance of most
Gnatcatcher {Polioptila caerulca), used the species stabilized or rebounded slightly: for 9
clearcLit in the second postharvest year. Final- of I I species, densities in 2001 were not sig-
ly. Common Yellowthroat {Geotidypis tricluis) nificantly less than those in 1999. Common
jI and Indigo Bunting (Passerina cycinca), Yellowthroat and Indigo Bunting, absent dur-
which were essentially absent preharvest ing the preharvest and the first postharvest
j (only one observation of Indigo Bunting in seasons, accounted for 40% of the total ob-
^ 1999) and in the first postharvest .season, col- servations in this treatment during the second
I onized the clearcut in the second postharvest postharvest year.
320 THE WILSON BULLETIN • Vol. 116, No. 4, December 2004
Uncutcontrol second postharvest season. The density of the
edge-shrub group increased in both of these
treatment areas in the second postharvest year,
1999 due primarily to the abundance ofIndigo Bun-
22000010 tings (Table 1).
DISCUSSION
Retention ofpatches offorest within an oth-
erwise clearcut area appears to enhance post-
Habitat-usegroup harvest forest bird diversity compared to that
in clearcuts without such patches (Merrill et
al. 1998, Schieck et al. 2000). We observed
greater species richness and overall bird den-
Clearcut
sities in the patch-retention treatment area
than in the clearcut in both 2000 and 2001. A
high percentage of the total bird detections in
1999
2000 the patch-retention area were recorded inside
2001 retained patches. By the second postharvest
year, the estimated density of the forest-inte-
rior group in the patch-retention area had re-
covered to approximately half that in the un-
cut control, whereas in the clearcut, no indi-
Habitat-usegroup
viduals of this group were detected. These
hndings corroborate the results of studies
from other regions. For example, in aspen for-
Patchretention ests of northern Minnesota, Merrill et al.
45
40 (1998) found high bird diversity in 0.54-ha
35 patches within clearcuts that averaged 14.6 ha.
Similarly, the overwhelming majority ofbirds
detected in our postharvest patch-retention
area (excluding Common Yellowthroat and
Indigo Bunting) occurred inside the patches.
Merrill et al. (1998) noted that birds using re-
tained patches did not necessarily nest there
Habitat-usegroup but used them for foraging, singing, or other
PIG. 2. Densities of avian habitat-use groups activities. In boreal forest of Alberta, Canada,
(mean ± 95% confidence intervals) in three harvest avian community composition in harvested
treatment areas in a bottomland hardwood forest in sites was most similar to that of unharvested,
South Carolina, 1999-2001. I = Porest Interior, I-E = old growth sites when retained trees and snags
Interior-Edge, P-E = Pield-Edge, and E-S = Edge- included large trees and were clumped togeth-
Shrub. er (Schieck et al. 2000). Such clumps may
have resembled the original forest by preserv-
Density estimates within each of the four ing some of its structure and microclimates
habitat-use groups did not differ over the 3 (Schieck et al. 2000).
years in the uncut control area (Fig. 2). How- The harvested portions of both the clearcut
ever, in both the clearcut and patch-retention and patch-retention areas had been colonized
areas, the density of the forest-interior group by early successional species (e.g.. Common
declined in the hrst postharvest year. In the Yellowthroat, Indigo Bunting) by the second
second postharvest year, density of forest-in- postharvest year. Their appearance in the sec-
terior birds rebounded in the patch-retention ond year after harvest was not unexpected, as
area, but not in the clearcut. Density of the at least one growing season is required for es-
forest-interior group was also greater in the tablishment of early successional grasses and
patch-retention area than in the clearcut in the forbs. Moorman and Guynn (2001) also re-
Harrison and Kilgo • BIRD RESPONSE TO HARVEST PRACTICES 321
ported these species in small group-selection mined, and this issue should be the focus of
cuts in bottomland hardwoods during the sec- future research.
ond postharvest year. Migratory songbirds ACKNOWLEDGMENTS
may discover potential future breeding habitat
during the season-ending dispersal phase and We thank MeadWestvaco Corporation for use ofits
return there the following year (Brewer and land, forimplementing and maintainingthetreatments,
Harrison 1975, Morton 1992). and for other assistance in this study. In particular we
We suggest that patch-retention timber har- appreciate the contributions ofJ. R Martin, who con-
ceived the design of the harvest treatments and pro-
vest may be more desirable than clearcutting vided invaluable logistical assistance. We thank D. De
when landowners are interested in maintaining Steven and A. A. Davis for providing helpful com-
songbird habitat, and that its impacts on bird ments on early drafts ofthe manuscript. J. D. Lanham,
populations warrant further investigation. C. E. Moorman, H. Li, and three anonymous referees
Patch-retention harvesting offers some of the awlasso isumppproorvteeddtbhye mthaenuUscSrDipAt EwoirtehsttheSierrvreivcieewSso.utChAerHn
silvicultural advantages of clearcutting, but Research Station. JC—K was supported by the U.S. De-
maintains elements of the preharvest stand partment of Energy Savannah River Operations Of-
—
that are apparently attractive to songbirds. Our fice through the USDA Eorest Service Savannah
findings indicate that, to some degree, the re- River and the USDA Eorest Service Southern Re-
sidual patches offorest continue to be used by search Station under Interagency Agreement DE-
AI09-00SR22188.
forest birds, while the surrounding clearcut
portion provides suitable habitat for edge- LITERATURE CITED
shrub and other early successional species.
That the densities ofmany birds did not dif- Baker, M. D. and M. J. Lacki. 1997. Short-term
changes in bird communities in response to sil-
fer in the patch-retention area before and after vicultural prescriptions. Eorest Ecology and Man-
harvest does not necessarily indicate that the agement 96:27-36.
habitat quality was similar (Van Horne 1983). Bibby, C. j., N. D. Burgess, D. A. Hill, and S. H.
Site fidelity among individual birds may ex- Mustoe. 2000. Bird census techniques, 2nd ed.
plain some use of the patches after harvest Academic Press, London, United Kingdom.
Brewer, R. and K. G. Harrison. 1975. The time of
(though not the increased use in the second
postharvest year), even if the quality of the Buckhlaabnitda,t sSe.leTc,tioDn.bRy.biArndds.erIsboisn,1 1K7.:5P2.1-B5u2r2n.ha.m,
patches was poor. Whether retained forest AND J. L. Laake. 1993. Distance sampling: esti-
patches have negative impacts on avian pop- mating abundance of biological populations.
ulation dynamics is unclear, as our study did Chapman and Hall, London, United Kingdom.
not address this question. Field-forest edges Buffington, J. M., J. C. Kilgo, R. A. Sargent, K. V.
Miller, and B. R. Chapman. 1997. Comparison
may be ecological traps for birds by concen- ofbreeding bird communities in bottomland hard-
trating nesting activity but also attracting nest wood forests ofdifferent successional stages. Wil-
predators (Gates and Gysel 1978). Similarly, son Bulletin 109:314-319.
the patches may provide perches for Brown- Burke, M. K., S. L. King, M. H. Eisenbies, and D.
headed Cowbirds (Molothnis ater), facilitat- Gartner. 2()()(). Vegetation and soils. Pages 23-
ing parasitism of nearby nests. Thus our re- 2st8udiyn:TahreepCoortosoanwhtahtecdheiveelbooptmteonmtlaonfdaecreofseyrsetnecme
tention patches, with their high edgc:area ra- wetland (M. K. Burke and M. II. Idsenbies. Eds.).
tios, could possibly have functioned as pop- General Technical Report SRS-38. U.S. Eorest
ulation sinks. However, studies in which the Service, Southern Research Station. Ashe\ille.
effect of variable tree retention (two-age type North Carolina.
harvests) on nesting success has been exam- Burki;, M. K.. S. E. King, D. Gartmr, and M. H.
Ejsi;nbii;s. 2003. Vegetation, soil, and Hooding re-
ined have generally found little or no evidence lationships in a blackwatcr tloodplain forest. Wet-
for such a phenomenon (Tittler and Hannon lands 23:988 1002.
2()()(), Duguay et al. 2001, Stuart-Smith and C'l aliiRBICK. W. K. AND .1. S. Ml ADows. 1993. Re-
Hayes 2003), and cowbird parasitism rates in generating oaks in the bottomlatuls. Ikiges 184
southeastern forests are low (Kilgo and Moor- 195 in Oak regeneration: serious problems, prac-
man 2003). Nevertheless, avian producti\ity tical recommendations (I). L. Eoftis and ('. Ei.
McCiee, EaIs.). General Eechnical Report .SI>84.
in residual forest patches such as those in our U..S. Eorest Ser\ice. Southern Research Station.
patch-retention treatment has yet to be deter- Ashe\ille, North ('arolina.
322 THE WILSON BULLETIN • Vol. 116, No. 4, December2004
Colwell, R. K. 1997. Estimates: statisticalestimation bird diversity on northern Minnesota aspen clear-
of species richness and shared species from sam- cuts. Conservation Biology 12:190-199.
ples, ver. 6.0b1. Online at <http://viceroy.eeb. Moorman, C. E. and D. C. Guynn, Jr. 2001. Effects
uconn.edu/estimates/> (accessed 27 September of group-selection opening size on breeding bird
2004). habitat use in a bottomland forest. Ecological Ap-
Duguay, J. R, P. B. Wood, and J. V. Nichols. 2001. plications 11:1680-1691.
Songbird abundance and avian nest survival rates Moorman, C. E., D. C. Guynn, Jr., and J. C. Kilgo.
in forests fragmented by different silvicultural 2002. HoodedWarblernesting success adjacentto
treatments. Conservation Biology 15:1405-1415. group-selection and clearcut edges in a southeast-
Eisenbies, M. H. andW. B. Hughes. 2000. Hydrology. ern bottomland forest. Condor 104:366-377.
Pages 10-13 in The Coosawhatchie bottomland Morton, M. L. 1992. Effects ofsex and birth data on
ecosystem study: a report on the development of pre-migration biology, migration schedules,return
a reference wetland (M. K. Burke and M. H. Ei- rates and natal dispersal in the Mountain White-
senbies, Eds.). General Technical Report SRS-38. crowned Sparrow. Condor 94:117-133.
U.S. Forest Service, Southern Research Station, Murray, L. A., B. Eppinette, and J. H. Thorp. 2000.
Asheville, North Carolina. Geomorphologyandsoil survey. Pages7-9inThe
Gates, J. E. and L. W. Gysel. 1978. Avian nest dis- Coosawhatchie bottomland ecosystem study: are-
persion and fledgling success in field-forest eco- port on the development of a reference wetland
tones. Ecology 59:871-883. (M. K. Burke and M. H. Eisenbies, Eds.). General
Hellmann, j. j. and G. W. Fowler. 1999. Bias, pre- Technical Report SRS-38. U.S. Forest Service,
cision, and accuracy of four measures of species
Southern Research Station, Asheville, North Car-
richness. Ecological Applications 9:824-834.
olina.
Hodgmfeoisrg,ersaMtts.orFoy.fabdnridefefDed.rienGng.tbKwirireddmtehcsnotmmzau.lnoni1g9t9i6et.hseNeinAotlrrtioappamiracihaaaln Nichollosc,k,J.AND.D,JT..RB.ouSlaiuneire.r,19J.98E..EHsitniemsa,tiKn.gHr.atePsolo-f
local species extinction, colonization, and turn-
River, Georgia. Wilson Bulletin 108:496-506.
Hunter, W. C., D. N. Pashley, and R. E. F. Escano. over in animal communities. Ecological Applica-
tions 8:1213-1225.
1993. Neotropical migratory landbird species and
theirhabitats ofspecial concern within the South- Norton, M. R. and S. J. Hannon. 1997. Songbird re-
east region. Pages 159-171 in Status and man- sponse to partial-cut logging in the boreal mixed-
agement of Neotropical migratory birds (D. M. wood forest of Alberta. Canadian Journal ofFor-
FRienpcohrtanRdMP-.2W2.9.StaUn.geSl., FEdosr.e)s.tGeSneerrvailceT,ecRhnoicckayl Pashlesety,ReDs.eaNr.chan27d:4W4.-5C3..Barrow. 1993. Effects of
Mountain Research Station, Fort Collins, Colora- land use practices on Neotropical migratory birds
in bottomland hardwood forests. Pages 315-320
do.
Hurst, G. A. and T. R. Bourland. 1996. Breeding in Status and management of Neotropical migra-
birds on bottomland hardwood regeneration areas tory birds (D. M. Finch and P. W. Stangel, Eds.).
on the Delta National Forest. Journal ofField Or- General Technical Report RM-229. U.S. Forest
nithology 67:181-187. Service, Rocky Mountain Research Station, Fort
Johnson, D. H. 2002. The importance ofreplication in Collins, Colorado.
wildliferesearch. Journal ofWildlifeManagement Peterjohn, B. G. and J. R. Sauer. 1994. Population
66:919-932. trends of woodland birds from the North Ameri-
Kilgo, j. C., K. V. Miller, and W. P. Smith. 1999. can Breeding Bird Survey. Wildlife Society Bul-
Effects of group-selection timber harvest in bot- letin 22:155-164.
tomland hardwoods on fall migrant birds. Journal Rich, T. D., C. J. Beardmore, H. Berlanga, P. J.
of Field Ornithology 70:404-413. Blancher, M. S. W. Bradstreet, G. S. Butcher,
Kilgo, J. C. and C. E. Moorman. 2003. Patterns of D. W. Demerest et al. 2004. Partners In Flight
cowbirdparasitism in the southern AtlanticCoast- North American landbird conservation plan. Cor-
al Plain and Piedmont. Wilson Bulletin 115:277- nell LaboratoryofOrnithology, Ithaca, NewYork.
284. ScHiECK, J., K. Stuart-Smith, and M. Norton. 2000.
Kilgo, J. C., R. A. Sargent, B. R. Chapman, and K. Bird communities are affectedby amountanddis-
V. Miller. 1998. Effect of stand width and adja- persion of vegetation retained in mixedwood bo-
cent habitat on breeding bird communities in bot- real forestharvestareas. ForestEcologyandMan-
tomland hardwoods. Journal ofWildlife Manage- agement 126:239-254.
ment 62:72-83. Shepard, J. P, S. J. Brady, N. D. Cost, and C. G.
Meadows, J. S. and J. A. Stanturf. 1997. Silvicul- Storrs. 1998. Classification and inventory. Pages
tural systems for southern bottomland hardwood 1-28 in Southern forested wetlands: ecology and
forests. Forest Ecology and Management 90:127- management (M. G. Messina and W. H. Conner,
140. Eds.). Lewis Publishers, Boca Raton, Florida.
Merrill, S. B.,F. J. Cuthbert,andG.Oehlert. 1998. Stuart-Smith, A. K. andJ. P. Hayes. 2003. Influence
Residual patches and their contribution to forest- of residual tree density on predation of artificial
Harrison and Kilgo • BIRD RESPONSE TO HARVEST PRACTICES 323
and natural songbird nests. Eorest Ecology and Van Horne, B. 1983. Density as a misleading indi-
Management 183:159-176. cator of habitat quality. Journal ofWildlife Man-
Thomas, L., J. L. Laake, J. E Derry, S. T. Buckland, agement 47:893-901.
D. L. Borchers, D. R. Anderson, K. P. Burnham Whitcomb, R. E, C. S. Robbins, J. E Lynch, B. L.
ET AL. 1998. Distance 3.5, release 6. Research Whitcomb, M. K. Klimkiewicz, andD. Bystrak.
Unit for Wildlife Population Assessment, Univer- 1981. Effects offorest fragmentation on avifauna
sity of St. Andrews, United Kingdom. Online at on the eastern deciduous forest. Pages 125-292 in
<http://www.ruwpa.st-and.ac.uk/distance/> (ac- Forest island dynamics in man-dominated land-
cessed 27 September 2004). scapes (R. L. Burgess and D. M. Sharpe, Eds.).
Tittler, R. and S. j. Hannon. 2000. Nest predation Springer-Verlag, New York.
in and adjacent to cutblocks with variable tree re- WiGLEY, T. B. AND J. M. SwEENEY. 1993. Cooperative
tention. Forest Ecology and Management 136: partnerships and the role of private landowners.
147-157. Pages 39-44 in Status and management of Neo-
Tittler, R., S. J. Hannon, and M. R. Norton. 2001. tropical migratory birds (D. M. Finch and P. W.
Residual tree retention ameliorates short-term ef- Stangel, Eds.). GeneralTechnicalReportRM-229.
fects of clear-cutting on some boreal songbirds. U.S. Forest Service, Rocky Mountain Research
Ecological Applications 11:1656-1666. Station, Fort Collins, Colorado.