Oak Forest Types Research Articles

Predicting the potential habitat of oaks with data mining models and the R system

Oak forests are essential for the ecosystems of many countries, particularly when they are used in vegetal restoration. Therefore, models for predicting the potential habitat of oaks can be a valuable tool for work in the environment. In accordance with this objective, the building and comparison of data mining models are presented for the prediction of potential habitats for the oak forest type in Mediterranean areas (southern Spain), with conclusions applicable to other regions. Thirty-one environmental input variables were measured and six base models for supervised classification problems were selected: linear and quadratic discriminant analysis, logistic regression, classification trees, neural networks and support vector machines. Three ensemble methods, based on the combination of classification tree models fitted from samples and sets of variables generated from the original data set were also evaluated: bagging, random forests and boosting. The available data set was randomly split into three parts: training set (50%), validation set (25%), and test set (25%). The analysis of the accuracy, the sensitivity, the specificity, together with the area under the ROC curve for the test set reveal that the best models for our oak data set are those of bagging and random forests. All of these models can be fitted by free R programs which use the libraries and functions described in this paper. Furthermore, the methodology used in this study will allow researchers to determine the potential distribution of oaks in other kinds of areas.

A Long-term Study of Seedling Regeneration for an Oak Forest Restoration in Cleveland Metroparks Brecksville Reservation, Ohio

The Cleveland Metroparks Brecksville Reservation initiated an oak forest restoration in 1990, using thinning and burning to encourage regeneration of oak forest species. In 96 quadrats (81 treatment and 15 control), understory woody species were inventoried from 1992 to 2002, and canopy opening was quantified. Six years into the study, the entire reservation became heavily populated with white-tailed deer (Odocoileus virginianus) and gypsy moths (Lymantria dispar), complicating the restoration efforts. During data analysis, understory woody species were classified into oak forest type and nonoak forest type, and three distinct phases became evident (phase 1: 1992–1994 treatment effects only, phase 2: 1995–1999 period of invasion by deer and gypsy moths, and phase 3: 2000–2002 post-invasion period). Both oak forest and nonoak forest species increased in the treatment area over the control area. During phase 1, burning encouraged oak seedlings, and thinning reduced competition from nonoak species, suggesting a temporarily successful restoration attempt. The intense gypsy moth browse in phase 2 reduced the number of oak seedlings, creating conditions favorable to nonoak species. The quantity and diversity of seedlings increased as gypsy moths opened the canopy. Deer browsed most species, oak and nonoak, even when deer populations decreased and more species were available. Original treatment effects may have been continuing in phase 3; however, additional years of study are needed.

울산 무제치 습원의 식생: 현존식생도와 오리나무 개체군

중간습원 무제치 습원의 현존식생도와 오리나무 개체군의 공간적 분포양식을 규명하였으며, 그 보존을 위한 생태학적 정보를 발굴하였다. 자늪과 웅늪의 1:100 실측지형도 상에서 총3,036개의 격자(5m <TEX>$\times$</TEX> 5m)를 이용하여 1996년 상세 현장조사가 이루어졌다. 현존식생도는 쇠털골형, 진퍼리새형, 오리나무형, 억새형, 삼림형 그리고 개방나지 등의 6 가지 식생범례로 제작되었으며, 습원 고유의 진퍼리새형과 비고유의 억새형이 가장 넓은 면적을 차지하고 있었다. 오리나무 1,491개체에 대한 매목조사와 티센망도(thiessen polygons) 분석으로부터 오리나무 개체군의 발달은 습원의 훼손(토사 및 산화 등)으로부터 매우 급속히 진행되고 있는 것으로 나타났다. 웅늪은 자늪에 비하여 보다 온전한 구조로 유지되고 있는 중간습원으로 평가되었으나, 자늪과 웅늪 모두 다년생 목본 식물종에 의한 산지 참나무림과 오리나무 우점의 산지 습생림의 발달에 의한 진퍼리새군강으로 대표되는 중간습원의 입지환경 변화와 그 소멸이 급격히 진행되고 있었다. 결국 진퍼리새군강의 중간습원 보존은 임도와 산불에 따른 토양의 부영양화에 대한 응급조치가 필요함이 지적되었다. Actual vegetation map drown with the scale 1 to 100 and Alnus japonica population in Ja-neup and Woong-neup of the Mujechi moor were described in order to monitor long-termly and preserve permanently, where is a very rare Molinietea moor and a legally protected area. A total of 3036 plots of 5m <TEX>$\times$</TEX> 5m were surveyed during summer 1996. Thiessen polygons of 1491 alder trees were derived from the plot data. Actual vegetation map was illustrated by 6 cover types such as needle spike-rush type, moor-grass type, alder-moor type, eulalia type, oak forest type, and exposed site. Molinia grasses native to the moor and Miscanthus grasses alien to the moor are reciprocally dominant. The area of Molinia grasslands was rapidly in decline and alder population size was dramatically in increase in the moor, particularly in Woong-neup. In Molinietea moor preservation more attention should be focused on the regulation of a nutrient rich soil from forest road and fire.

Technical Note—Amounts of Down Woody Materials for Mixed-Oak Forests in Kentucky, Virginia, Tennessee, and North Carolina

Abstract Fallen or down dead wood is a key element in healthy forest ecosystems. Although the amount of down wood and shrubs can provide critical information to forest resource managers for assessing fire fuel build up, data on biomass of down woody materials (DWM) are not readily accessible using existing databases. We summarized data collected by the USDA Forest Service's Forest Inventory and Analysis (FIA) program into biomass for mixed-oak forest types in Kentucky, Virginia, Tennessee, and North Carolina to obtain a baseline average for Appalachian mixed-oak forests. We sampled a subset of 76 plots from 16 oak forest types and computed biomass for each DWM component, using slightly different equations for each. Biomass of DWM components was summarized using simple statistics. The mean of all DWM components combined was 32 Mg/ha, and the median was 29 Mg/ha. Over half the mean DWM was duff and litter (18 Mg/ha combined); coarse woody material (CWM) was less than 15% of mean DWM; and fine woody material (FWM) (7 Mg/ha) was almost twice CWM. The range of mass for individual components was quite large. These estimates compare favorably to those found in studies for other forest types. South. J. Appl. For. 28(2):113–117.

Oak dune forests in Denmark and their ecology

Floristic and structural data of oak dune forests in western Jutland (Kaergård and Blåbjerg) were collected in 1998 and compared with previous studies of Danish oligotrophic oak forests. A hierarchical divisive clustering resulted in four main clusters, containing two coastal oak vegetation types, Quercus robur– Carex arenaria and Q. robur– Pleurozium schreberi communities, both restricted to the oak dune forest areas in western Jutland, and two more widespread oak forest types, Q. robur– Vaccinium myrtillus and Q. robur– Holcus mollis communities. The low-growing Q. robur– C. arenaria shrub and forest vegetation occurs primarily in the outer dunes, with the tallest trees on the leeward sides of the dunes, while a Q. robur– P. schreberi community is found in the inner parts of the dune system. The similarity to a considerable number of the acidophilous vascular plant species closely relates the Danish oak dune forest communities to the Pinus sylvestris or P. sylvestris– Q. robur vegetation described in neighbouring countries, suggesting that P. sylvestris probably had a prominent role in the Danish oak dune forests before human exploitation caused its extinction in the 18th century. The low-growing oak shrub in the outer dunes, aged between 52 and 85 years, generally seems to have remained unchanged in floristic composition and structure since 1964 when the previous study on the shrub was performed. In the dune valleys, oak stand-level dieback is widespread and was latest recorded in 1985. Such retrogressive processes are probably due to a combination of low winter temperatures and other adverse climatic conditions. In the central parts of the oak dune forest, the oak trees, aged between 55 and 73 years, grow to a higher stature, but also here stand-level dieback is widespread leading to retrogression. Only in the innermost parts, a regular forest cover seems to prevail with trees of 68–104 years of age and the development of the field vegetation cover progresses into a typical forest community. This is exemplified by the decrease of light-demanding species and the establishment of true forest species like Oxalis acetosella and Dryopteris dilatata. The structure and poor-species diversity make the oak dune forest unique both in Denmark and on an international scale. We recommend further studies of the remnants of the original, and in the past probably more widespread dune forests in Denmark, and drawing up of a detailed management and conservation plan. The possibility of permitting the native and now reintroduced P. sylvestris to re-colonise the oak dune forest area and be part of the natural vegetation succession processes should be seriously considered.

Nitrification in Sludge‐Amended Michigan Forest Soils

AbstractNet nitrification following liquid sludge application to four Michigan forest types was studied by aerobically incubating intact cores containing the forest floor and upper 10 cm of mineral soil. Significant net nitrification occurred in cores receiving surface applications of anaerobically digested municipal sludge (22.3 g N m−2). Core NO3‐N contents at 8 wk were 4.7, 5.3, 4.9, and 2.0 g m−2 for aspen (Populus spp.), northern hardwoods, oak (Quercus spp.), and pine (Pinus spp.) forest types, respectively. Net nitrification did not occur during 8‐wk incubations of untreated control cores or cores receiving surface applications of sludge sterilized by freeze‐drying (24.5 g N m−2) or liming (25.7 g N m−2), suggesting that the nitrifiers responsible for net nitrification in cores receiving the anaerobically digested sludge were added with the sludge. Nitrate content at 8 wk in cores from an oak field site receiving anaerobically digested sludge (43.0 g N m−2) 33 mo prior to incubation was 72% of that in oak site soil cores treated immediately prior to incubation (22.3 kg N m−2), indicating that nitrifying populations introduced with sludge 33 mo prior were still viable. Nitrate content at 8 wk in oak cores was 77% lower when sludge was incorporated. Nitrate contents in incubated soil cores could not be used to directly predict NO3 concentrations in soil leachate and groundwater beneath sludge‐treated plots at the four field sites. Results indicate that nitrification and potential for NO3 leaching following sludge application to acid forest soils are influenced by sludge type, forest type, and history of prior sludge applications.

Laboratory Investigation of Leachate Chemistry from Six Appalachian Forest Floor Types Subjected to Simulated Acid Rain

AbstractIn order to determine the role of the forest floor in neutralization of strong acidity in acid rain, simulated acid rain at pH 4.0 was applied under laboratory conditions to forest floor samples from six Appalachian forest types. Effects of forest type, storm size, and repeated storm applications on leachate chemistry were investigated. Forest floor leachate from a hemlock forest [Tsuga canadensis (L.) Carr.] showed significantly lower pH (mean pH 3.9) than other leachate types due to high concentrations of organic acids. Cove hardwood forest found in an area with calcareous soil parent material produced forest floor leachate with a significantly higher pH (mean pH 6.5) than other forest types. Forest floor leachates from mixed oak (Quercus spp.), northern hardwood, red pine (Pinus resinosa Ait.), and white pine (P. strobus L.) forests were intermediate in pH (mean pH 4.1–4.4) due to the acidifying influence of organic acid leaching balanced against the neutralizing effect of H+ exchange for Ca2+ in the forest floor. Weak acidity concentrations obtained by Gran plots were well correlated with 400‐nm wavelength transmittance and anion deficits in the leachates. Organic anions dominated leachates from hemlock and mixed oak forest types. Increasing storm sizes caused reduced ionic concentrations, especially of H+, Mg2+, and Ca2+, but three repeated storms produced no consistent significant reductions in loads or concentrations.

Microhabitats in a Forest‐Floor Small Mammal Fauna

This study examines microhabitat configurations for 3 small mammal species occupying second—growth mesic forest on Walker Branch Watershed in eastern Tennessee. Small mammal populations were sampled in oak—hickory, chestnut oak and pine forest types, with three 0.36—ha live—trapping grids per forest type. There were 255 total captures for 4 forest—floor species during 9,696 trap nights between 1 May and 16 August 1973. These captures included 2% Blarina brevicauda, 66% Peromyscus leucopus, 5% Ochrotomys nuttalli and 27% Tamias striatus. Blarina was captured most frequently in the pine forest type, Peromyscus and Tamias most frequently in the oak—hickory, and Ochrotomys exclusively in the pine. Twenty—nine variables describing proximate habitat structure were measured at each sample mammal capture site. There were significant overall species differences on 18 of these variables. A subset of variables was identified for each of the 3 most frequently observed species, consisting of only the variables of particular significance for distinguishing the capture sites of that species from those of the other species collectively. Discriminant analysis of these few variables both described the average microhabitat configuration, for a species and indicated the relative importance of each variable for characterizing or describing that configuration, subject to the assumptions of the analysis. The observed differences between species microhabitat configuration are consistent with both the composition by forest type of each species sample and the literature information for each species. Microhabitat segregation among these species is discussed with reference to recent theories of habitat selection, ecological, specialization and resource partitioning in equilibrium faunas.

Some Effects of Carbaryl on Mites, Collembola, and Calosoma sp. in an Oak Type Forest

An aerial application of carbaryl (Sevin-4-oil)® at rate of one lb AI/acre was studied for effects on mites, Collembola, and Calosoma beetles in an oak forest type. Pre- and post-spray populations were compared, and except for mites a short term reduction resulted, but population densities appeared normal by the following spring.