relating land-use history and climate to the dendroecology ... prinus 1998.pdf · relating land-use...

Relating land-use history and climate to the

dendroecology of a 326-year-old Quercus prinus

talus slope forest

Charles M. Ruffner and Marc D. Abrams

Abstract: Dendroecology and land-use history were used to investigate the ecological history of a 326-year-old Quercus

prinus L. forest. Quercus prinus, Castanea dentata (Marsh.) Borkh., and Pinus rigida Mill. dominated this talus slope prior to

European settlement based on witness tree records. Oak species have exhibited continuous recruitment over three centuries

probably in response to periodic fire and wind disturbances. While the stand escaped the direct impacts of timber cutting and

the charcoal iron industry, the indirect effects of these land-use practices increased growth and recruitment. Different criteria

were used for understory versus overstory trees to improve our detection of growth releases. Overall, major disturbances

occurred approximately every 40 and 31 years before and after European settlement, respectively. This century, old-growth

Q. prinus experienced marked growth increases coupled with high recruitment following the introduction of the chestnut

blight (Cryphonectria parasitica (Murr.) Barr.) to the stand in 1909. Tree growth was also highly correlated with temperature

and Palmer drought severity indices between 1895 and 1995. Climatic fluctuations in the 1820s–1830s and 1920s reduced

radial growth and recruitment resulting in stem exclusion stages following regeneration pulses. Relating land-use history and

climatic data to the dendroecology of this forest improved our understanding of its historical development.

Résumé: Les auteurs ont eu recours à la dendroécologie et à l’historique de l’utilisation des terres pour examiner l’histoire

écologique d’une forêt de Quercus prinus L. vieille de 326 ans. Selon les chronologies des arbres témoins, cette pente de talus

était dominée par le Q. prinus, le Castanea dentata (Marsh.) Borkh. et le Pinus rigida Mill. avant la colonisation européenne.

Les espèces de chêne montraient un recrutement continu durant trois siècles, probablement en réaction aux feux périodiques et

aux perturbations causées par le vent. Bien que le peuplement ait échappé à l’impact direct de la coupe du bois et de la

production de charbon pour l’industrie sidérurgique, les effets indirects de ces pratiques d’utilisation des terres ont favorisé la

croissance et le recrutement. Les auteurs ont utilisé différents critères chez les arbres du sous-étage par opposition à ceux de

l’étage dominant pour mieux détecter les augmentations de croissance reliées aux dégagements. En général, les perturbations

se sont produites à tous les 40 et 31 ans avant et après la colonisation européenne, respectivement. Au cours du présent siècle,

les vieux Q. prinus ont connu une augmentation marquée de leur accroissement, couplée à un recrutement élevé, suite à

l’introduction de la brûlure du châtaignier dans le peuplement en 1909. Entre 1895 et 1995, l’accroissement des arbres était

aussi étroitement corrélé avec la température et avec l’indice de sévérité de la sécheresse de Palmer. Les fluctuations du climat

dans les années 1820 à 1830 et 1920 ont réduit l’accroissement radial et le recrutement, ce qui a engendré des stades

d’exclusion des tiges suite à des poussées de régénération. Dans cette forêt, le fait de mettre l’historique de l’utilisation des

terres et les données climatiques en relation avec la dendroécologie a amélioré la compréhension de son développement passé.

[Traduit par la Rédaction]

Introduction

Much of the eastern deciduous forest experienced widespreadlogging, grazing, insect–pathogen introductions, and altereddisturbance regimes following European settlement (Williams1989; Abrams 1992; Whitney 1994). Recent dendroecologicalstudies of surviving old-growth forests have greatly increasedour knowledge of historic disturbance regimes (Foster 1988;Abrams et al. 1995; Nowacki and Abrams 1997), species com-position, distribution, and community structure (Mikan et al.1994; Nowacki and Abrams 1994; Stahle and Chaney 1994;Abrams et al. 1998), and successional trends (Runkle 1981;

Abrams and Downs 1990). Nonetheless, much work needs tobe completed to identify other old-growth sites to assess thelong-term effects of anthropogenic activities such as fire ex-clusion, introduced insects and pathogens, and differing land-use histories (Stephenson et al. 1993; White and White 1996).

Land-use history has been used to characterize and explainvegetation change from pre-European settlement to presentconditions (Raup 1966; Cronon 1983; Foster 1992; Orwig andAbrams 1994). These processes and subsequent vegetationchanges have provided an historical perspective towards futuremanagement and restoration of human disturbed ecosystems(Foster et al. 1992; Foster et al. 1996). In addition to land-usehistory, climatic variation has had a profound effect on vege-tation dynamics. Recent dendroclimatological studies in thenortheast have reported climatic variations influencing recruit-ment patterns and growth suppression–release events (Cookand Jacoby 1977; Abrams and Orwig 1995; D’Arrigo et al.1996; Abrams et al. 1997). The coupling of land-use historyand climate data should provide an improved understanding ofecological processes.

Received May 29, 1997. Accepted December 2, 1997.

C.M. Ruffner 1 and M.D. Abrams. The Pennsylvania StateUniversity, School of Forest Resources, 203 Forest ResourcesLab, University Park, PA 16802-4704, U.S.A.

1 Author to whom all correspondence should be addressed.e-mail: [email protected]

Can. J. For. Res. 28: 347–358 (1998)

347

© 1998 NRC Canada

X97-220.CHPTue May 12 13:03:38 1998

Color profile: DisabledComposite Default screen

Estimating disturbance frequencies in closed forests hasbeen based on canopy ascension dates and major and moderaterelease from suppression criteria for understory trees (Lorimerand Frelich 1989). This model is most valid for suppressedunderstory trees and may not be appropriate for overstory in-dividuals. Therefore, Nowacki and Abrams (1997) developeda criterion to be used for overstory trees using more liberalrelease criteria. However, we feel a better methodology in-cludes both these techniques in which Lorimer and Frelich’s(1989) criteria are used to assign canopy accession dates andreleases in understory trees, while Nowacki and Abrams’(1997) model identifies overstory disturbances.

The present study integrates dendroecology, land-use his-tory, and climatic data to investigate the pre-European settle-ment composition, structure, and disturbance regime of anold-growth chestnut oak (Quercus prinus L.) talus slope forestin the Ridge and Valley of central Pennsylvania. We utilize anew approach for estimating disturbances by using differentrelease criteria for understory versus overstory trees. Our ob-jectives were to (1) document the historical landscape by ana-lyzing witness tree distributions and Native Americanoccupation as well as stand characteristics such as age struc-ture and disturbance regimes and (2) assess the impact of land-use history of the surrounding landscape and regional climateon tree growth and recruitment in an old-growth remnant.

Study area

This study was conducted within a 21.5-ha chestnut oak stand occur-ring within the 400-ha Detwiler Run watershed (Fig. 1) in the SevenMountains region of the Ridge and Valley province of central Penn-sylvania (UTM 18: E267970/ N4511540; elev. 600 m) (Fenneman1938). The synclinal topography is typified by long-parallel sand-

stone capped ridges (500–670 m elevation) separated by a steep ra-vine slope. Large talus areas are found on the upper side slopes, whilecolluvial material forms the lower slopes and stream bed. Soils withinthe study area are in the Hazleton–Dekalb association (Typic Dystro-chrepts) and are characterized as deep, very steep, well-drained soilsformed in residuum and colluvium from acidic sandstone (Braker1981). The climate of the area is dry continental having average dailytemperatures range from –5° to 7°C (December–March) and 26° to28°C (June–August). Annual precipitation averages 93.4 cm and thefrost-free season lasts 170 days (April 27 – October 14) (Braker1981).

Forests of the region were originally classified by Braun (1950) asoak–chestnut, although they are now considered mixed-oak, follow-ing the elimination of overstory chestnut (Castanea dentata (Marsh.)Borkh.) earlier this century by the introduced chestnut blight(Cryphonectria parasitica (Murr.) Barr.) (Keever 1953; Stephenson1986). The area is noted for the 145-ha old-growth hemlock – whitepine community composing the Detwiler Run Natural Area in thelower ravine (Braun 1950); however, Braun (1950) also identified thechestnut oak community on the south-facing talus slope. The stand ischaracterized by low-branching, twisted, sparse-crowned individualstypical of old-growth forests (Fig. 2) (Witherow 1908; Stahle andChaney 1994).

Methods

Pre-European landscape and land-use historyImpacts of anthropogenic activity were compiled from a variety ofsources. Ownership history was traced from land treaties with NativeAmericans and original warrants issued by the Pennsylvania LandOffice during European settlement (ca. 1778 at nearby McAlevy’sFort). Warrant surveys represent a tract of land as surveyed at the timeof first settlement and are made up of several bearings and distanceslinking property corners, often referenced as blazed witness trees,posts, or stone corners (Munger 1991). After a tract was surveyed, a

Fig. 1. Map of Detwiler Run watershed in Huntingdon and Centre counties, Pennsylvania. The old-growth Quercus prinus talus slope forest is

outlined in the middle of the figure with the dark, broken line, and transects within are represented by three broken lines.

Can. J. For. Res. Vol. 28, 1998348

© 1998 NRC Canada

X97-220.CHPTue May 12 13:03:53 1998


warrant map was produced illustrating the configuration of the prop-erty, including boundary line descriptions, referenced corners, andother geographically significant features such as springs, streams, andNative American villages or trails (Abrams and Ruffner 1995). Mostof the warrants in the Seven Mountains region were patented (initialdeeds issued) to landowners who eventually assimilated several war-rants into huge tracts of valuable timberland.

Documents including original warrants, deeds to charcoal ironcompanies and lumber operations, and the transfer of the tracts to thePennsylvania State Bureau of Forestry were searched for details suchas timber volumes, species, railroad grade locations, and corner trees.Many of these transfers included, for tax purposes, survey drawingsoutlining warrant boundaries and cultural features such as railroadgrades and sawmill locations. These survey drawings, often referredto as connected drafts, coupled with deed references, provided thebasis for the witness tree analysis and documentation of land-usehistory events. Deeds and tax records are available from the Recorderof Deeds and Tax Assessment Offices, Centre County and Hunting-don County Courthouses in Bellefonte and Huntingdon, Pennsylva-nia, respectively. Connected drafts were acquired courtesy of theGreenwood Furnace State Park and Edward Heary, a local landsurveyor.

Witness tree – topographic relationships were characterized usingcontingency table analysis, a method that tests for independence be-tween topographic position and the presence–absence of a species

(Strahler 1978). This test is performed by calculating the likelihood-ratio χ2 statistic, G2, and comparing this value to the appropriatequantile of the χ2 distribution (Agresti 1996). Standardized residualswere calculated following Haberman’s method (1973) for contin-gency tables revealing significance. Residual values quantify a spe-cies preference (positive) or avoidance (negative) of a particulartopographic position (Whitney 1990). Connected drafts were over-layed on USGS 7.5′ topographic maps, and corner trees were talliedby presence–absence and topographic position forming a 2 × 3 con-tingency table. Because of the extreme topography of the SevenMountains region and the relatively low number of witness trees tal-lied, only three topographic variables were used: ridge, side slopes,and valley sites. This was done so that adequate numbers of treescould be placed into each topographic position and G2 statistics couldbe calculated.

Vegetation samplingFollowing ground reconaissance to locate the stand, stand structuredata were collected in June 1996, on twenty 0.02-ha overstory fixed-area plots located approximately 50 m apart on three transects ar-ranged along the contour within the 21.5-ha stand. Within each plot,overstory data included DBH (1.37 m) and crown class by species forall trees (DBH ≥ 8 cm) living or dead. For each species a relativeimportance value was calculated by summing the relative density(no. of trees/ha), relative frequency (distribution), and relative domi-nance (basal area/ha) and dividing by three (Cottam and Curtis 1956).While standing snags were tallied and measured, these data were notincluded in importance value calculations. In addition, within eachplot three or four trees were cored at breast height (1.37 m) for agedeterminations and radial growth analysis. Trees representing all spe-cies on the site were selected for coring to provide several cores foreach diameter class (i.e., several individuals of each species per di-ameter class). Nested circular plots (20 and 5 m2) were used to collectsapling (tree species <8 cm DBH, >1.4 m tall) and seedling (tree spe-cies ≤1.4 m tall) densities by species, respectively. Other stand de-scriptors such as aspect, slope, and elevation also were recorded.

Fig. 2. Old-growth chestnut oak stand showing 122-cm (DBH)

chestnut oak with characteristic form, and high amount of fallen

woody material and extreme talus conditions of site.

Species Ridge Side slope Valley

Quercus alba 6 9 35

Quercus prinus 21 23 1

Quercus rubra 3 7 1

Quercus velutina 5 3 2

Pinus spp. 26 10 18

Pinus rigida 7 0 1

Pinus strobus 3 2 6

Castanea dentata 17 12 1

Betula spp. 6 3 1

Carya spp. 3 9 5

Tsuga canadensis 0 3 5

Other species 3 19 24

No. of trees 118 100 96

Note: Other species include A. rubrum, A. saccharum,

Cornus spp., Fagus grandifolia, Fraxinus spp., Juglans

nigra, Liriodendron tulipifera, Magnolia acuminata, Nyssa

sylvatica, Ostrya virginiana, Tilia americana, and Ulmus

spp. Of these only A. rubrum and L. tulipifera represented

more than 5% frequency on side slopes. All others

represented less than 1% across topographic positions.

Table 1.Species frequency (%) of presettlement

forests and number of witness trees in each

topographic position in the Seven Mountains region

(26 000 ha), central Pennsylvania.

Ruffner and Abrams 349

© 1998 NRC Canada

X97-220.CHPTue May 12 13:04:09 1998


Radial growth analysisIndividual cores (n = 94) were taken to the laboratory for drying,mounting, sanding, and cross dating (Phipps 1985). After cross-dating the individual trees using skeleton plots of signature years(Stokes and Smiley 1996), the chronology was compared with thenearby Alan Seeger Natural Area master chronology as a furthercross-dating reference (Nowacki and Abrams 1994). Annual growthincrements were measured to the nearest 0.01 mm with a tree-ringmeasuring device and recorded using the MACDENDRO microcom-puter program (Regents Instruments Inc., Quebec, Canada). A stand-ardized ring width index for each tree was created by dividing yearlymeasured growth values by expected values obtained from fitting alinear regression to measured values or the mean growth increment(Fritts and Swetnam 1989). Standardization removes the age-relatedgrowth trend and converts all series to the same relative variance,giving each equal weight when averaged together into the masterchronology (Fritts 1976). Cores from the 29 oldest chestnut oak treeswere used to construct a master chronology spanning 326 years.

Different criteria were used to improve our detection of growthreleases for trees in understory and overstory canopy positions. Majorand moderate release criteria developed by Lorimer and Frelich(1989) were applied to understory individuals, while more liberalrelease criteria developed by Nowacki and Abrams (1997) were usedfor overstory individuals.

Canopy accession dates for all trees were determined using thefollowing three methods developed by Lorimer and Frelich (1989).First, early growth rates were measured to determine whether treeswere initially growing fast enough (radial growth rate ≥1.4 mm/year),indicating a high probability of gap origin. If so, the canopy accessiondate is the decade of the innermost ring (Lorimer and Frelich 1989).Second, each core was examined for periods of suppression and re-lease based on major and moderate release criteria (cf. Lorimer andFrelich 1989). They defined a major sustained release as a ≥100%average growth increase lasting at least 15 years, and a moderatetemporary release as a >50% average growth increase lasting 10 to15 years for trees in the understory. These major and moderatereleases corresponded to canopy accession dates (if the first criterionwas not met) and distinguished significant disturbance events fromclimatic or thinning factors (Lorimer and Frelich 1989). Twenty-sixof the 29 oldest chestnut oaks used in chronology development wereassigned canopy accession dates using these major and moderate re-lease events criteria. Two others exhibited growth rates higher than1.4 mm/year and were assigned accession dates in accordance withthe first technique. The last core did not fit either criteria and a modi-fied technique was used. Although the ambiguous growth of this in-dividual masked detectable releases, its growth reflected a gradualincrease lasting nearly 100 years. An “ambiguous zone” was identi-fied as extending from the earliest ring in a period of increasinggrowth to the point at which 80% of the maximum growth rate wasachieved (Lorimer and Frelich 1989). The decade within the ambigu-ous zone having median growth was identified as the likely canopyaccession date (cf. Lorimer and Frelich 1989).

Following canopy accession, overstory release criteria developedby Nowacki and Abrams (1997) identified disturbances in matureoverstory individuals exhibiting ≥25% growth increases lasting10 years. This method is based on the observation that mature treesresponded to disturbances with smaller growth releases often lasting

Fig. 3. Significant associations (p < 0.01) of representative witness tree species on various landforms within the Seven Mountains region of

central Pennsylvania. Positive and negative values indicate a preference or avoidance, respectively, for the landform.

Fig. 4. Percent canopy class distribution for species in the Detwiler

Run chestnut oak talus slope forest, central Pennsylvania. OT,

overtopped; INT, intermediate; CODOM, codominant; DOM,

dominant. Species abbreviations are QURU, Quercus rubra; PIST,

Pinus strobus; PIRI, Pinus rigida; NYSY, Nyssa sylvatica; BELE,

Betula lenta; ACRU, Acer rubrum; and QUPR, Quercus prinus.

Can. J. For. Res. Vol. 28, 1998350

© 1998 NRC Canada

X97-220.CHPTue May 12 13:04:27 1998


less than 15 years. Thus, Lorimer and Frelich’s 50–100% growthincrease, lasting 15 years would not adequately detect all overstorydisturbances in mature individuals. By convention then, trees ac-cessed to the canopy either by (1) originating in a gap and meetinghigh growth requirement or by (2) major and moderate release events.Following canopy accession, less stringent overstory releases are re-ported. Annual and mean seasonal (spring, March 1 – May 31; sum-mer, June 1 – August 31) climate variables (Palmer drought severityindex (PDSI), precipitation, and temperature) from 1895 to 1995 werecorrelated (Pearson’s) with tree growth (National Oceanic and At-mospheric Administration 1996).

Results

Pre-European landscapePre-European settlement forests on ridge sites and side slopesin the Seven Mountains region were dominated by Pinus

spp.(pine), chestnut oak, and chestnut (Table 1). Species citedless often on these sites included Pinus rigida Mill. (pitch

pine), Betula spp. (birch), Quercus alba L. (white oak), Caryaspp. (hickory), Quercus rubra L. (red oak), and Quercus velu-tina Lam. (black oak). Valley sites were dominated by whiteoak, pine, and Pinus strobus L. (eastern white pine), followedby hickory, and Tsuga canadensis (L.) Carrière (eastern hem-lock) (Table 1). Analysis of species–site relationships withstandardized residuals illustrates the increased distribution ofchestnut, pine, pitch pine, and chestnut oak on more xeric ridgesites, while white oak exhibited a clear preference for valleyfloor sites (Fig. 3).

Land-use historyPost-European land-use history in the study area essentiallybegan with the Albany Union Purchase of 1754, which trans-ferred a huge tract of land from the Iroquois to the Pennsylva-nia Proprietary government (Table 2). This treaty fraudulentlyclaimed all lands south of a line drawn northwest from themouth of Penns Creek on the Susquehanna River to the Lake

Date Ownership Comments

1672 Six Nations Confederacy Defeat of Susquehannock Nation and development of Covenant Chain forging alliance

between Iroquois and English against Pennsylvania Native peoples, particularly the

Lenape (Delawares), Shawnees, and Susquehannocks (Snow 1994)

1754 Albany Purchase Transfer of lands to Pennsylvania Proprietary Government comprising all “lands within the

province north of the Kittatinny Hills, west of the Susquehanna River, and south of a line

drawn ‘northwest and by west’ from a point a mile above the mouth of Penns Creek.”

(Kent 1974)

1794 Mathew Simpson 400 acres warranted Feb. 25, 1794, surveyed June 21, 1794, unpatented and remained

unseated until 1842

1842 James Gilliam Huntingdon County Treasurer Andrew Hurst deeded said warrant to James Gilliam, of

Philadelphia, Pa., possibly a land speculator

1844 James Duncan Simpson warrant transferred to James Duncan, also of Philadelphia

1845 Stephen Kurtz and

Adam Detwiler

Conglomeration of three 400-acre (1200 acres total) warrants into “Kurtz and Detwiler

Lands” with first mention of “sawmill erected thereon”

1858 John Barr, Jr. Further addition of James Fulton warrant (400 acres) to above acreage (total 1600 acres);

continued to maintain sawmill (Huntingdon (Hu) county deedbook O-2, page 37)

1866 Freedom Iron & Steel Co. Purchase of above tracts in addition to landholdings representing over 65 000 acres centered

on Greenwood Furnace (Hu deedbook U-2, page 197)

1871 Logan Iron & Steel Co. Managed timberlands to provide charcoal for three iron furnaces, two at Greenwood furnace

and one at Burnham Furnace (Fagley 1993); reserved all surface and subsurface mineral

rights for all landholdings when properties sold to Kulp Lumber Co.; in addition, a

15-acre tract in the southwest corner of said Simpson warrant was not sold (Hu deedbook

B-3, page 208)

1902 Kulp Lumber Co. Railroad lumbering operation primarily cutting pitch pine for mine prop timber trade, leased

study area lands to Reichley Bros. and Beidlelheimer lumbering operations (Kline 1971)

(Hu deedbook U-4, page 558)

1903 Commonwealth of Pennsylvania Purchase of timberlands under the Forest Reserve Act (Hu deedbook X-4, page 454)

1906 Reichley Brothers 151-acre tract of Simpson and Fulton warrants lying in Centre County transferred from Kulp

Lumber (Centre (Ce) county deedbook 95, page 113)

1906 Commonwealth of Pennsylvania 151-acre tract complete with “reserved standing timber” transferred from Reichley Bros.;

agreement to set aside all timber “measuring in diameter from the exterior of the bark

12 inches or less...shall be allowed to stand and remain uninjured upon the said land. In

addition to the foregoing, the said party of the second part (Commonwealth of PA) agree

to leave standing upon the tract of land...three hundred sound and healthy white pine trees

and three hundred sound and healthy hemlock trees, each of which from the exterior of the

bark shall measure not less than 15 inches;” this being the reservation of the Natural Area

previously described by E.L. Braun (Ce deedbook 96, page 334; Ce Miscellaneous book

Ms O, page 435)

Table 2.Land ownership history of the Detwiler Run watershed area, Rothrock State Forest, Huntingdon County, central Pennsylvania.


© 1998 NRC Canada

X97-220.CHPTue May 12 13:04:29 1998


Erie shoreline. Justly dissatisfied, the Iroquois demanded andwere granted a new treaty placing the northern boundary as aline drawn west from a point 1 mi (1 mi = 1.6 km) north ofPenns Creek to the ridge of the Allegheny Mountains, thensouth along the ridgeline of the Allegheny Mountains (Munger1991). This treaty dated October 23, 1758, surrendered alllands north of this line back to the Iroquois, while lands to thesouth, including the study area, were opened to settlement.

The study stand falls within two adjacent warrants, theMathew Simpson and James Fulton warrants issued in Febru-ary 1794. It appears that these two warrants were not settleduntil they were purchased by partners Kurtz and Detwiler aspart of a 1200-acre (1 acre = 0.40 ha) tract on which theyerected a sawmill and cut timber from the lowlands usinghorse-drawn skidding (Table 2). They held the lands untilJohn Barr, Jr., purchased the 1200-acre tract in 1858 andadded the James Fulton warrant of 400 acres. Until its sale in1866, Barr apparently maintained the sawmill on the property.Nearby in Greenwood Furnace, the Freedom Iron and SteelCompany was buying timberlands throughout the SevenMountains region to produce charcoal for its iron furnaces.

In 1866, Freedom Iron and Steel purchased the four war-rants owned by Barr, including the study area warrants. Failingin 1871, creditors seized Freedom Iron and Steel, selling alllands, furnaces, and capital to the newly formed Logan Ironand Steel Company (Fagley 1993). Using a 20- to 30-year

rotation, Logan managed over 65 000 acres of chestnut, mixedoak, and hickory forests to produce charcoal for three furnaces,two at Greenwood Furnace and another at Burnham in nearbyMifflin County. There was probably enough local wood tocontinue charging the two furnaces at Greenwood, althoughthere was not enough to keep their larger and more importantBurnham furnace viable (P. Fagley, personal communication).By the late 1890s, locals claimed that one “could stand on topof the stack and not see a green thing taller than a lilac bushas far as the eye could see” (Fagley 1993). Although warrantsof the study area lay at the extreme periphery of Logan’s hold-ings and were never cut for charcoal, adjacent warrants werecut for these purposes (P. Fagley, personal communication).Starting in 1893, the company began selling off large tracts ofland to lumber companies (Fagley 1993).

The Kulp Lumber Company purchased the lands surround-ing Detwiler Run in 1902 as part of a large network of loggingrailroads throughout the Seven Mountains. The ReichleyBrothers and Beidleheimer Logging used existing tramwaysowned by Kulp to log Reichley lands in the headwaters areaof Detwiler Run, while Beidleheimer cut lands near the mouthof the watershed (Kline 1971). Interestingly, the two tramwayswere never linked; a break between the two straddled the Centre–Huntingdon County line, and subsequently the Detwiler RunNatural Area was never cut. Reichley, however, did constructa skid path angling up the side of Thickhead Mountain to

Fig. 5. Diameter distribution of major species for Detwiler Run talus slope forest, central Pennsylvania. Species abbreviations are QURU,

Quercus rubra; PIST, Pinus strobus; PIRI, Pinus rigida; NYSY, Nyssa sylvatica; BELE, Betula lenta; ACRU, Acer rubrum; and QUPR,

Quercus prinus.

Species

Dominance

(m2⋅ha–1)

Rel. dominance

(m2⋅ha–1)

Density

(no./ha)

Rel. density

(no./ha) Frequency Rel. frequency RIV*

Quercus prinus 25.24 61.1 198 33.6 20 22.0 38.9

Acer rubrum 2.8 6.8 138 23.4 18 19.8 16.7

Betula lenta 3.43 8.3 80 13.6 16 17.6 13.2

Nyssa sylvatica 2.2 5.3 95 16.1 13 14.3 11.9

Quercus rubra 2.7 6.5 38 6.5 13 14.3 9.1

Pinus strobus 3.95 9.6 30 5.1 7 7.7 7.4

Pinus rigida 1 2.4 10 1.7 4 4.4 2.8

Totals 41.32 100.0 589 100.0 91 100.0 100

*Relative importance values = (relative dominance + relative density + relative frequency)/3.

Table 3.Relative importance values (RIV) for species in Detwiler Run talus slope forest, central Pennsylvania.

Can. J. For. Res. Vol. 28, 1998352

© 1998 NRC Canada

X97-220.CHPTue May 12 13:04:46 1998


remove pitch pine timber within 1000 m of the study stand.Despite the regional history of charcoal production and lumberoperations, the extreme talus conditions of the site preventedexploitation of the timber resources. By 1906 all tracts weresold to the Commonwealth of Pennsylvania under the ForestReserves Act of 1897 and assimilated into the present-dayRothrock State Forest.

Current vegetationThe talus slope forest overstory is dominated by chestnut oak,while black birch (Betula lenta L.), red maple (Acer ru-brum L.), and blackgum (Nyssa sylvatica Marsh.) make upmuch of the intermediate and overtopped classes (Fig. 4). Thestand contains 589 trees⋅ha–1 and 41.32 m2⋅ha–1 of basal area.Chestnut oak represents 61% of the total basal area and ischaracterized by fairly well distributed large trees (Table 3).Despite their relatively low basal area, other species of moder-ate importance in the stand are red maple, black birch, andblack gum. Typical of an old-growth forest, the diameter dis-tribution typifies an inverse-J shaped curve (Smith 1986), withlarge diameter classes dominated by chestnut oak and whitepine, while medium diameter classes are made up of chestnutoak, sweet birch, pitch pine, red oak, and red maple (Fig. 5).Pole-sized individuals are primarily red maple and black gum.Saplings were dominated by red maple, with 175 ha–1, whilethe seedling class was dominated by chestnut oak and red oakwith 1600 and 500 ha–1, respectively (data not shown).

DendroecologyThe stand is uneven aged with chestnut oak present across allage-classes and several different age cohorts (Figs. 6 and 7A).Chestnut oak recruited nearly continuously from 1670 to 1950,excluding the period from 1750 to 1790, and peaked between1790 and 1830. Many species regenerated between 1900 and1920, coinciding with the 1909 outbreak of chestnut blight (Penn-sylvania 1914). While one 238-year-old red maple was found,most red maple recruitment occurred between 1900 and 1920.

Radial growth analysis of the 29 oldest chestnut oak iden-tified suppression–release events during the last three centu-ries (Fig. 7B). Reductions in growth from 1680 to 1690 and1710 to 1720 are followed by rapid growth increases after 1690and 1720, respectively. Overstory and moderate temporary re-leases occurred between 1730 and 1750, concurrent withrecruitment of several chestnut oak, white pine, and one redmaple. Between 1750 and 1790 the stand exhibited much re-duced growth as well as low recruitment. The decades between1790 and 1820 exhibited a pulse of understory and overstorygrowth releases, with several chestnut oaks attaining canopystatus (Figs. 7A and 7B). Growth continued to increasethrough the early 1800s until 1810, after which a reductionoccurred. Chestnut oak recruitment was high during this periodof increased growth from 1790 to 1810, peaking in the 1830s.Despite the reduction in growth from 1810 to 1822, canopyrecruitment was high, during which 6 of the 29 oldest chestnutoak attained canopy status. While no major or moderategrowth releases occurred during the 1830s, several chestnutoak, sweet birch, and pitch pine recruited from 1830 to 1850.Multiple releases occurred between 1840 and 1860, duringwhich time 3 of the 29 oldest chestnut oak accessed to thecanopy (Fig. 8). Growth generally increased between 1860and 1960, although sporadic reductions occurred such as from1870 to 1875. In addition, recruitment was limited from 1860to 1900.

During this century, insect–pathogen attacks and recurringdrought have impacted old-growth chestnut oak in variousways. Peak releases occurred from 1900 to 1910, correspond-ing to the 1908–1909 outbreak of chestnut blight in this region(Pennsylvania 1914). Despite the generally high growth ratesof chestnut oak this century, droughts of 1909–1910 and themid-1960s severely reduced growth in the stand. Followingpeak growth during the 1970s, a profound growth decrease isnoted in the chestnut oak chronology. Only the prolongeddrought of the mid-1960s limited regeneration in the stand,however. Correlation analysis indicates an influence of his-toric climate variables on tree growth (Table 4). Seasonal Pal-mer drought severity indices for the current and previous yearexhibited significant (p < 0.01) negative influence on treegrowth. In addition, seasonal precipitation was significant (p <0.05), while seasonal temperature had a slightly negative in-fluence on tree growth.

Fig. 6. Decade of canopy accession for 29 chestnut oak trees used

for standardized ring width chronology for Detwiler Run talus

slope forest, central Pennsylvania.

Variable Year t (n = 100) Year t – 1(n = 99)

PDSI

Spring 0.395** 0.340**

Summer 0.295** 0.274**

Annual 0.365** 0.344**

Precipitation

Spring 0.247* 0.219*

Summer 0.118 0.089

Annual 0.241* 0.240*

Temperature

Spring –0.069** –0.111

Summer –0.087 –0.054

Annual 0.002 –0.042

Note: *p < 0.05 = 0.205, **p < 0.01 = 0.267.

Table 4.Pearson correlations between the standard ring

width chronology and the mean spring (March 1 –

May 31), mean summer (June 1 – August 31), and mean

annual values of the Palmer drought severity index

(PDSI), precipitation, and temperature in the current

year t and the previous year t – 1, 1895–1995.


© 1998 NRC Canada

X97-220.CHPTue May 12 13:04:51 1998


Discussion

Regional palynological studies suggest that by 9000 years ago,forests of the region were dominated by oak species (Watts1979). Local forests of the Seven Mountains region weredominated by oak, chestnut, pine, black gum, and birch ac-cording to pollen analysis conducted at Bear Meadows 1 km tothe northwest of the study area (Kovar 1964). Chestnut oak,chestnut, and pine dominated ridgetop and slope presettlementforests of the Seven Mountains region, while white oak, whitepine, and hemlock inhabited valleys and riparian areas. Otherregional witness tree studies have reported similar findings(Nowacki and Abrams 1992; Abrams and Ruffner 1995).

The pre-European landscape was dominated by peoples ofthe Susquehannock, Lenape (Delaware) and Shawnee Nations(Kent 1974; Kent 1993; Wallace and Hunter 1981). An inves-tigation of the Pennsylvania Archaeological Site Survey

(PASS) files revealed 12 Native American occupation sites inStone Valley, including a palisaded village along StandingStone creek, into which Detwiler Run flows. Several “Indianpaths” mapped in the 1760s (Scull 1770) crossed the SevenMountains, linking the Lenape villages of Standing Stone(Huntingdon), Kishocoquillas (Lewistown), and Bald Eagle’sNest (Milesburg) (Wallace 1965). Archeological site densitiessuggest Native populations in the region were low, but NativePeoples had a lasting impact on local vegetation through agri-cultural clearing, hunting and gathering, collection of fuelwood, and widespread use of fire for driving game, clearingunderbrush, improving browse, etc. (Table 5) (Day 1953;Chapman et al. 1982; Pyne 1983; DeVivo 1991). In 1775,Philip Vickers Fithian, a Presbyterian circuit rider noted“...large open plains, cleared either by Indians or accidentalfire. Hundreds of acres are covered with fine grass...” in PennsValley (to the north) and Kishocoquillas Valley (to the south)(Fithian 1934; Losensky 1961). Indeed, many travelers andsurveyors noted old Indian fields, Indian towns, and orchardsin contemporary journals and maps (Miller 1996; Wallace andHunter 1981). These lowlands comprised lands upon whichmany Europeans established farms and frontier settlements.

Forests of the region experienced fire from various sources,including Native American and European land clearing, slashfires, and lightning. Escaped fires from both Native and settlerland clearance in the valley would have reached the oak forestson the slopes and ridgetops. Although no fire records exist forthe area, slash fires from logging and nearby charcoal produc-tion likely occurred in these forests based on contemporaryaccounts of the late 1800s and regeneration of fire-adaptedoak, birch, and pitch pine during this period (Rothrock 1894;Little and Garrett 1990; Abrams and Nowacki 1992; Whitney1994). The forest districts composing the Seven Mountainsarea reported a total of 204 lightning strikes, resulting in over65 detected fires between 1912 and 1913 (Pennsylvania De-partment of Forestry 1912–1915). Spiral seams indicatinglightning damage were observed in several individuals duringfield sampling. In addition, many sample plots contained soilcharcoal and fire scars. Several of these scars were very deepwithin the stem, evidence of a long history of fire in the stand(Fig. 9).

Post-European settlement land uses included agriculturalclearing and timber cutting for lumber and the charcoal ironindustry. Settlers probably employed clearing and burning foragricultural lands (Table 5) (Cronon 1983; Whitney 1994).While the talus slope was certainly not completely cleared, theSeeger farm at the mouth of the watershed was settled duringthe early 1800s. Growth releases during the early settlementperiod of 1790–1820 coincided with canopy ascension of oaksand increased regeneration of chestnut oak, black birch, andpitch pine. Between 1790 and 1873, 15 sawmills operated inthe immediate area of Jackson Township (Africa 1883). More-over, many growth releases occurred after a sawmill operatedat the mouth of the watershed from 1840 to 1860 (Table 5).Large-scale logging was noted during the 1840s–1850s in thenearby Alan Seeger Natural Area (Nowacki and Abrams1994). After 1804 local forests were cut on a 20- to 30-yearrotation for the charcoal iron furnaces at Juniata Forge, Green-wood Furnace, and Monroe Furnace (Table 5) (Lytle 1878).

The disturbance regime of the stand increased afterEuropean settlement. From radial growth analysis, it is evident

Fig. 7. Age–diameter and radial growth relationships of cored trees

in an old-growth chestnut oak talus slope forest, central

Pennsylvania. (A) Age–diameter graph of 81 cored trees. (B) Mean

ring width index of 29 oldest chestnut oak. Values across graph are

number of oaks included in master chrolonology.

Can. J. For. Res. Vol. 28, 1998354

© 1998 NRC Canada

X97-220.CHPTue May 12 13:05:03 1998


that presettlement overstory major disturbances impacted thearea an average of every 40 years, longer than the 27-yearinterval identified by Nowacki and Abrams (1997) for nearbyTussey Ridge. In contrast, the postsettlement disturbance in-terval averaged 31 years, as opposed to a 40-year postsettle-ment interval for Tussey Ridge (Nowacki and Abrams 1997).Postsettlement land uses such as selective logging, charcoal-ing, and subsequent slash fires apparently intensified the dis-turbance regime, affecting residual tree growth andrecruitment patterns.

Climate fluctuations apparently influenced residual growthand canopy accession of the early 1800s. Reduced growth dur-

ing the 1810s and 1830s may be tied to decade-scale fluctua-tions in North Atlantic climate (e.g., Jacoby and D’Arrigo1992; D’Arrigo et al. 1996). Low chestnut oak growth on thissite appears linked to reduced temperatures after the Tamboraeruption, which resulted in 1816 being a year without a sum-mer. An abrupt increase in radial growth, high oak recruitment,and overstory releases lasting through the 1850s conformed toa warming trend in the 1820s–1830s (D’Arrigo et al. 1992).Low regeneration during the periods from 1750 to 1790 and1870 to 1890 suggests stem exclusion stages after regenerationpulses in the 1740s and 1850s (cf. Abrams et al. 1997b).

In 1906, the Commonwealth of Pennsylvania purchased

Fig. 8. Releases per decade in old-growth Detwiler Run chestnut oak talus slope forest. Major (≥100% growth increase lasting 15 years) and

moderate (>50% growth increase lasting 10–15 years) releases in understory individuals were detected using the method of Lorimer and

Frelich (1989). Overstory (≥25% growth increase lasting 10 years) releases were detected using the method of Nowacki and Abrams (1997).

Land-use practice Period Impact on chestnut oak stand

Native American occupation Before 1760 Periodic fires reduced fire-intolerant species; hunting pressure reduced deer herbivory

Native American migration 1760–1788 Cessation of fires promotes secondary succession on cleared areas

European settlement After 1788

Selective logging 1845–1866 Removal of overstory trees results in release of residual trees and understory initiation;

windthrow in stand increased surrounding gaps

Iron industry charcoaling 1804–1902 20- to 30-year cutting rotation perpetuated pitch pine and oak species

Slash fires 1866–1906 Reduction of fire-intolerant species; selection for sprouting species

Logging of pitch pine 1902–1906 Removal of pitch pine seed stock

Designation as Natural area 1906 Implementation of fire-cessation policies

Chestnut blight introduction 1909–1920 Removal of overstory chestnut releases residual trees, regeneration cohort of diverse species

Lightning 1912–1915 204 strikes in Seven Mountains region, resulting in 65 detected fires; historic impact

unknown

Periodic drought 1895–present Severe droughts (1921–1930 and 1960s) reduced growth of overstory individuals and

fostered recruitment of drought-tolerant oaks

Insect defoliations 1967–1968 Oak leaftier and oak leafroller defoliations of red oak resulted in growth increases in

old-growth chestnut oak

Oak decline 1980–present Gypsy moth defoliations coupled with extended droughts reduced vigor of old-growth

chestnut oak and resulted in loss of some overstory chestnut oak after attack by Armillaria

root rot, a secondary killing agent

Table 5.Land-use practices, time periods, and impacts on chestnut oak stand in Detwiler Run, central Pennsylvania.


© 1998 NRC Canada

X97-220.CHPTue May 12 13:05:12 1998


Detwiler Run from a logging firm that was cutting pitch pineto the east of the stand. Cutting of overstory pitch pine mightbe responsible for multiple releases of the early 1890s; how-ever, most individuals exhibited reduced growth consistentwith the prolonged drought of the late 1890s. Increased growthand multiple releases between 1900 and 1920 correspond withthe introduction of the chestnut blight fungus to the region.Joseph T. Rothrock inspected forests west of the study standon June 21, 1910, and reported “a considerable body of deadstanding chestnut in this reserve” that had previously beenreported in 1909 as “good thrifty white oak, chestnut, and rockoak (chestnut oak)” (Rothrock 1912; Morton 1910). With theremoval of canopy chestnut, a flush of regeneration occurredfrom chestnut oak, red oak, red maple, and black gum. Foreststhroughout the oak–chestnut region (Braun 1950) experiencedspecies composition and forest structure changes following theloss of overstory chestnut (Keever 1953; Stephenson 1986;Abrams et al. 1997).

Variable growth of the 1920s can be attributed to changingclimatic conditions, particularly the severe droughts of1921–1923, 1925, and 1930. In contrast, high growth from1931 to 1950 is correlated with increased precipitation andmoderate temperatures. Old-growth chestnut oak trees experi-enced overstory releases after reduced growth due to the early

1960s drought and subsequent defoliations of red oak indi-viduals by the oak leaftier (Croesia semipurpurana Kearfott)and oak leaf roller (Archips semiferanus Walker) (J. Quimby,personal communication). In addition, gypsy moth (Lymantriadispar L.) defoliations in 1981, 1989, and 1991 resulted inreduced diameter growth of overstory chestnut oak. Secondaryeffects of Armillaria root rot may have resulted in the loss ofseveral overstory individuals. Defoliations also increased re-cruitment of red maple, birch, and black gum (Allen and Bow-ersox 1989; Hix et al. 1991). The increased competition fromthese less commercially valuable species coupled with the im-pact of white-tailed deer (Odocoileus virginianus Boddaert)browsing often results in low oak seedling recruitment (Mar-quis et al. 1976). A complex interaction of site conditions,recurring drought, and multiple gypsy moth defoliations ap-pears to be responsible for oak decline in the Appalachianregion (Wargo et al. 1983). Decreasing radial growth of chest-nut oak in Detwiler Run after 1970 appears to conform to theseoak decline conditions.

Conclusions

By integrating dendroecology, land-use history, and climatedata, we have gained an improved historical perspective on thedevelopment of the old-growth chestnut oak forest comparedwith studies integrating fewer of these types of data. Usingdifferent understory and overstory release criteria also pro-vides a better estimate of the rate of overstory disturbancesaffecting canopy accession, growth releases, and recruitmentpatterns throughout the stand’s history rather than using a fixedcriterion for all canopy classes. We suggest that this method beutilized in future historical ecology studies to enhance ourknowledge of the disturbance regimes, before and after Euro-pean settlement.

During the stand’s history, recruitment of oak has beenfairly continuous except for two stem exclusion stages follow-ing substantial regeneration pulses and coinciding with cli-matic perturbations. This century, old-growth chestnut oakresponded to the loss of overstory chestnut and several defo-liations of associate oak species, enhancing new recruitmentand growth of residual overstory individuals. However, de-creasing growth after 1960 may be tied to oak decline, com-mon throughout the Appalachian region. Comparingpre-European settlement to present-day conditions, chestnutoak and red oak have increased somewhat, probably as a con-sequence of the loss of overstory chestnut. In addition, redmaple, black gum, and sweet birch have increased in impor-tance, probably in response to fire exclusion (cf. Mikan et al.1994). While we believe that the stand is experiencing a re-duction in oak importance, the droughty talus conditions andpossibility of future fire at Detwiler Run will probably main-tain a stand composed of oak, red maple, black birch, and blackgum.

Acknowledgements

The authors thank Paul Fagley, Historian at Greenwood Fur-nace State Park, and Edward Heary for contributing connecteddrafts for witness tree analysis as well as Todd A. Morgan for hisefforts during field data collection and core preparation. Thisproject was conducted with the support of the Pennsylvania

Fig. 9. Chestnut oak (56 cm DBH) with deep fire scar on talus

slope forest in Detwiler Run area.

Can. J. For. Res. Vol. 28, 1998356

© 1998 NRC Canada

X97-220.CHPTue May 12 13:05:24 1998


Bureau of Forestry. Pennsylvania Archaeological Site Surveyfiles are available through the Pennsylvania Historical andMuseum Commission, Harrisburg.

References

Abrams, M.D. 1992. Fire and the development of oak forests. Bio-Science, 42: 346–353.

Abrams, M.D., and Downs, J.A. 1990. Successional replacement ofold-growth white oak by mixed-mesophytic hardwoods in south-western Pennsylvania. Can. J. For. Res. 20: 1864–1870.

Abrams, M.D., and Nowacki, G.J. 1992. Historical variation in fire,oak recruitment and post-logging accelerated succession in centralPennsylvania. Bull. Torrey Bot. Club, 119: 19–25.

Abrams, M.D., and Ruffner, C.M. 1995. Physiographic analysis ofwitness tree distribution (1765–1798) and present forest coverthrough north-central Pennsylvania. Can. J. For. Res.25: 659–668.

Abrams, M.D., Orwig, D.A., and DeMeo, T.E. 1995. Dendroecologyand successional dynamics of a pre-settlement-origin white pine –mixed oak forest in the southern Appalachians, USA. J. Ecol.83: 122–133.

Abrams, M.D., Orwig, D.A., and Dockry, M.J. 1997. Dendroecology,climatic responses and successional status of two contrasting old-growth oak forests in the Blue Ridge Mountains. Can. J. For. Res.27: 994–1002.

Abrams, M.D., Ruffner, C.M., and DeMeo, T.E. 1998. Dendroecol-ogy and species co-existence in an old-growth Quercus–Acer–Tilia talus slope forest in the central Appalachians, USA.For. Ecol. Manag. In press.

Africa, J.S. 1883. History of Huntingdon and Blair counties.L.H. Everts, Philadelphia.

Agresti, A. 1996. An introduction to categorical data. John Wiley &Sons, New York.

Allen, D., and Bowersox, T. 1989. Regeneration in oak stands follow-ing gypsy moth defoliations. In Seventh Central Hardwood ForestConference Proceedings, Edited by G. Rink and C.A. Budelsky.USDA For. Serv. Gen. Tech. Rep. NC-132.

Braker, W.L. 1981. Soil survey of Centre County, Pennsylvania.USDA Soil Conservation Service, Washington, D.C.

Braun, E.L. 1950. Deciduous forests of eastern North America.MacMillan, New York.

Chapman, J., Delcourt, P.A., Cridlebaugh, P.A., Shea, A.B., and Del-court, H.R. 1982. Man–land interaction: 10 000 years of Ameri-can Indian impact on native ecosystems in the lower LittleTennessee River Valley, eastern Tennessee. South. Archaeol.1: 115–121.

Cook, E.R., and Jacoby, G.C., Jr. 1977. Tree-ring drought relation-ships in the Hudson Valley, New York. Science (Washington,D.C.), 198: 399–401.

Cottam, G., and Curtis, J.T. 1956. The use of distance measures inphytosociological sampling. Ecology, 37: 451–460.

Cronon, W. 1983. Changes in the land: Indians, colonists and theecology of New England. Hill and Wang, New York.

D’Arrigo, R.D., Jacoby, G.C., and Cook, E.R. 1992. Impact of recentNorth Atlantic anomalies on surrounding land areas based on den-droclimatic evidence. Geophys. Res. Lett. 19: 2321–2324.

D’Arrigo, R.D., Cook, E.R., and Jacoby, G.C. 1996. Annual todecadal-scale variations in northwest Atlantic sector temperaturesinferred from Labrador tree rings. Can. J. For. Res. 26: 143–148.

Day, G.M. 1953. The Indian as an ecological factor. Ecology,34: 329–346.

DeVivo, M.S. 1991. Indian use of fire and land clearance in the south-ern Appalachians. In Fire and the environment; ecological andcultural perspectives. Edited by S.C. Nodvin, and T.A. Waldrop.USDA For. Serv. Gen. Tech. Rep. SE-69.

Fagley, P.T. 1993. Greenwood Furnace, Huntingdon County, PA:The rise and fall of a Juniata Valley Iron Industry. In Canal Historyand Technology Proceedings. Edited by L.E. Metz. Canal Historyand Technology Press, Easton, Pa. Vol. 12. pp. 163–195.

Fenneman, P.V. 1938. Physiography of eastern United States.McGraw-Hill, New York.

Fithian, P.V. 1934. Journals of Philip V. Fithian, 1775–1776. Editedby R.G. Albion and L. Dodson. Princeton University Press,Princeton, N.J.

Foster, D.R. 1988. Disturbance history, community organization andvegetation dynamics of the old-growth Pisgah Forest, south-western New Hampshire, USA. J. Ecol. 76: 105–134.

Foster, D.R. 1992. Land-use history (1730–1990) and vegetation dy-namics in central New England, USA. J. Ecol. 80: 753–772.

Foster, D.R., Zebryk, T., Schoonmaker, P., and Lezberg, A. 1992.Post settlement history of human land-use and vegetation dynam-ics of a Tsuga canadensis woodlot in central New England.J. Ecol. 80: 773–786.

Foster, D.R., Orwig, D.A., and McLachlan, J.S. 1996. Ecological andconservation insights from reconstructive studies of temperateold-growth forests. Trends Ecol. Evol. 11: 419–424.

Fritts, H.C. 1976. Tree rings and climate. Academic Press, London.Fritts, H.C., and Swetnam, T.W. 1989. Dendroecology: a tool for

evaluating variations in past and present forest environments.Adv. Ecol. Res. 19: 111–188.

Haberman, S.J. 1973. The analysis of residuals in cross-classifiedtables. Biometrics, 29: 205–220.

Hix, D.M., Fosbroke, D.E., Hicks, R.R., Jr., and Gottshalk, K.W.1991. Development of regeneration following gypsy moth defo-liation of Appalachian Plateau and Ridge and Valley hardwoodstands. In Eighth Central Hardwood Forest Conference. Edited byL.H. McCormick and K.W. Gottshalk. USDA For. Serv. Gen.Tech. Rep. NE-148.

Jacoby, G.C., and D’Arrigo, R.D. 1992. Spatial patterns of tree-growth anomalies for the North American boreal treeline in theearly 1800s, including the year 1816. In The year without a sum-mer? World climate in 1816. Edited by C.R. Harrington. CanadianMuseum of Nature, Ottawa. pp. 255–265.

Keever, C. 1953. Present composition of some stands of the formeroak–chestnut forest in the southern Blue Ridge Mountains. Ecol-ogy, 34: 44–54.

Kent, B.C. 1993. Susquehanna’s Indians. Commonwealth of Pennsyl-vania Historical and Museum Commission, Harrisburg, Pa.

Kent, D.H. 1974. History of Pennsylvania purchases from the Indians.Garland Publishing, New York.

Kline, B.F.G., Jr. 1971. Pitch pine and prop timber. Logging railroadera of lumbering in Pennsylvania No. 1. Lycoming Printing, Wil-liamsport, Pa.

Kovar, A.J. 1964. Pollen analysis of the Bear Meadows bog of centralPennsylvania. Pa. Acad. Sci. 38: 16–24.

Little, S., and Garrett, P.W. 1990. Pitch pine. In Silvics of NorthAmerica. Edited by R.M. Burns and B. Honkala. USDA For. Serv.Agric. Agric. Handb. 654.

Lorimer, C.G., and Frelich, L.E. 1989. A method for estimating can-opy disturbance frequency and intensity in dense temperate for-ests. Can. J. For. Res. 19: 651–663.

Losensky, B.J., III. 1961. The great plains of central Pennsylvania.M.S. thesis, Pennsylvanian State University, University Park.

Lytle, M.S. 1878. History of Huntingdon County in the state of Penn-sylvania. W.H. Roy Publisher, Lancaster, Pa.

Marquis, D.A., Eckert, P.L., and Roach, B.A. 1976. Acorn weevils,rodents, and deer all contribute to oak regeneration difficulties inPennsylvania. USDA For. Serv. Res. Pap. NE-356.

Mikan, C.J., Orwig, D.A., and Abrams, M.D. 1994. Age structure andsuccessional dynamics of a pre-settlement-origin chestnut oak for-est in the Pennsylvania Piedmont. Bull. Torrey Bot. Club,121: 13–23.


© 1998 NRC Canada

X97-220.CHPTue May 12 13:05:26 1998


Miller, E.L. 1996. Journal of an unknown Englishman travelingthrough central Pennsylvania, 1794. Transcribed and edited byE.L. Miller from Pennsylvania State Archives ManuscriptCollection, Harrisburg, Pa.

Morton, T.R. 1910. Huntingdon County Reserve, Barre Division1908–09. In Report of the Pennsylvania Department of Forestry.Edited by C.E. Aughinbaugh. Pennsylvania Department of For-estry, Harrisburg.

Munger, D.B. 1991. Pennsylvania land records: a history and guidefor research. Scholarly Resources, Wilmington, Del.

National Oceanic and Atmospheric Administration. 1996. Climate visu-alization website: URL (http://www.ncdc.noaa.gov/onlineprod/).

Nowacki, G.J., and Abrams, M.D. 1992. Community, edaphic, andhistorical analysis of mixed oak forests of the Ridge and Valleyprovince in central Pennsylvania. Can. J. For. Res. 22: 790–800.

Nowacki, G.J., and Abrams, M.D. 1994. Forest composition, struc-ture, and disturbance history of the Alan Seeger Natural Area,Huntingdon County, Pennsylvania. Bull. Torrey Bot. Club,121: 277–291.

Nowacki, G.J., and Abrams, M.D. 1997. The use of ring-width aver-aging for reconstructing disturbance histories of oak forests. Ecol.Monogr. 67: 225–249.

Orwig, D.A., and Abrams, M.D. 1994. Land-use history(1720–1992), composition, and dynamics of oak–pine forestswithin the Piedmont and coastal plain of northern Virginia. Can. J.For. Res. 24: 2141–2149.

Pennsylvania Chestnut Tree Blight Commission. 1914. Final reportof the Pennsylvania Chestnut Tree Blight Commission. Wm. Stan-ley Ray, Harrisburg, Pa.

Pennsylvania Department of Forestry. 1912–1915. Reports of thePennsylvania Department of Forestry. Harrisburg Publishing Co.,Harrisburg, Pa.

Phipps, R.L. 1985. Collecting, preparing, cross-dating, and measur-ing tree increment cores. US Geological Survey, Water ResourceInvestigation Report 85-4148.

Pyne, S.J. 1983. Indian fires. Nat. Hist. 2: 6–11.

Raup, H.M. 1966. The view from John Sanderson’s farm: a perspec-tive for the use of land. For. Hist. 10: 1–11.

Rothrock, J.T. 1894. Forests of Pennsylvania. Proc. Am. Philos. Soc.33: 114–133.

Rothrock, J.T. 1912. Reports of examinations of property and inspec-tions of forest practice, with recommendations thereon, made byDr. J.T. Rothrock during the years 1910–1911. In Report of thePennsylvania Department of Forestry. Edited by C.E. Aughin-baugh. Pennsylvania Department of Forestry, Harrisburg, Pa.

Runkle, J.R. 1981. Gap regeneration in some old-growth forests of theeastern United States. Ecology, 62: 1041–1051.

Scull, W. 1770. A map of Pennsylvania; exhibiting not only the im-proved parts of that province but also its extensive frontiers. Com-monwealth of Pennsylvania Historical and Museum Commission,Harrisburg, Pa.

Smith, D.M. 1986. The practice of silviculture. John Wiley & Sons,New York.

Snow, D.R. 1994. The Iroquois. Blackwell Publishers, Cambridge,Mass.

Stahle, D.W., and Chaney, P.L. 1994. A predictive model for thelocation of ancient forests. Nat. Areas J. 14: 151–158.

Stephenson, S.L. 1986. Changes in a former chestnut dominated for-est after a half century of succession. Am. Midl. Nat.116: 173–179.

Stephenson, S.L., Ash, A.N., and Stauffer, D.F. 1993. Appalachianoak forests. In Biodiversity of the southeastern Unites States: up-land terrestrial communities. Edited by W.H. Martin, S.G. Boyce,and A.C. Echternacht. John Wiley & Sons, New York.

Stokes, M.A., and Smiley, T.L. 1996. An introduction to tree-ringdating. University of Arizona Press, Tucson. [Reprint of1968 edition.]

Strahler, A.H. 1978. Binary discriminant analysis: a new method forinvestigating species–environment relationships. Ecology,59: 108–116.

Wallace, P.A.W. 1965. Indian paths of Pennsylvania. Commonwealthof Pennsylvania Historical and Museum Commission.

Wallace, P.A.W., and Hunter, W.A. 1981. Indians in Pennsylvania.2nd revised ed. Commonwealth of Pennsylvania Historical andMuseum Commission, Harrisburg, Pa.

Wargo, P.M., Houston, D.R., and LaMadeleine, L.A. 1983. Oak de-cline. USDA For. Serv. For. Insect Dis. Leafl. 165.

Watts, W.A. 1979. Late Quaternary vegetation of central Appalachiaand the New Jersey Coastal Plain. Ecol. Monogr. 49: 427–469.

White, P.S., and White, R.D. 1996. Old-growth oak and oak hickoryforests. In Eastern old-growth forests: prospects for rediscoveryand recovery. Edited by M.B. Davis. Island Press, Washington,D.C.

Whitney, G.G. 1990. The history and status of the hemlock–hardwoodforests of the Allegheny Plateau. J. Ecol. 78: 443–458.

Whitney, G.G. 1994. From coastal wilderness to fruited plain: a his-tory of environmental change in temperate North America, 1500to the present. Cambridge University Press, New York.

Williams, M. 1989. Americans and their forests: a historical geogra-phy. Cambridge University Press. New York.

Witherow, J.L. 1908. A silvical study of the rock oak (Quercusprinus), western Franklin and eastern Fulton counties. In Reportof the Pennsylvania Department of Forestry for the year 1907.Harrisburg Publishing Company, Harrisburg, Pa.

Can. J. For. Res. Vol. 28, 1998358

© 1998 NRC Canada

X97-220.CHPTue May 12 13:05:27 1998
