Navigation – Plan du site

AccueilPré-publicationArticles sous presseImpact of Climatic and Topographi...

Impact of Climatic and Topographic Factors on Distribution of Sub-tropical and Moist Temperate Forests in Pakistan

L’impact des facteurs climatiques et topographiques sur la répartition des forêts subtropicales et tempérées humides au Pakistan
Naveed Ahmad, Muhammad Irfan Ashraf, Sabeeqa Usman Malik, Ihsan Qadir, Naeem Abbas Malik et Kashif Khan

Résumés

Les facteurs climatiques et topographiques contrôlent la répartition des forêts à travers le monde. Ce travail a étudié les effets de ces facteurs sur la distribution spatiale des forêts subtropicales (broussailles et pins) et tempérées humides du Pakistan. L'étude a utilisé le modèle numérique d'élévation (MNT), des images Sentinel-2 et des données climatiques pour quantifier les impacts des facteurs climatiques et topographiques sur la répartition des forêts. Les données ont été analysées statistiquement à l'aide du coefficient de corrélation (R), de la régression linéaire et d’un arbre de décision. Les résultats ont indiqué six types de forêts. Cette typologie était significativement liée aux facteurs topographiques (altitude) et climatiques. Le coefficient de corrélation (R) indiquait une forte relation positive avec l'altitude (R = 0,92) suivie de la température moyenne annuelle (R = –0,76). De même, les précipitations annuelles indiquent une relation positive avec une valeur R de 0,53. Le modèle de régression linéaire a montré que l'altitude, la saisonnalité des précipitations et la plage de température annuelle étaient fortement significatives avec un R2 global de 0,85. Des arbres de décision ont été développés pour explorer les interactions possibles des facteurs explicatifs afin de déterminer les facteurs impératifs. Les résultats des arbres de décision des deux méthodes de croissance (détection automatique d'interaction du chi carré (CHAID) et arbres de classification et de régression (CRT)) ont montré que l'altitude était le facteur le plus important prédisant un type de forêt particulier. De plus, d'autres facteurs tels que la température du trimestre le plus sec, les précipitations annuelles, la saisonnalité des précipitations et la pente ont été identifiés comme des facteurs importants dans le CRT. La présente étude a conclu que les types de forêts étaient fortement influencés par le climat et la topographie. Cependant, l'altitude était le meilleur facteur explicatif, a une importance relative significative et peut être utilisée pour une stratification forestière détaillée.

Haut de page

Notes de la rédaction

Paper received on August 14, 2019, received in revised form on March 29, 2020, accepted on May 04, 2020.

Texte intégral

1. Introduction

1Natural forests in Pakistan extend over a long range of climatic and topographic conditions starting from the lowest coastal mangroves to the highest sub-alpine forests. Climatic and topographic factors have a great influence on vegetation distribution, composition, density and physiognomy (Ahmed et al., 2006). Moreover, topography integrated with climatic variables can be used as a good predictor in forest modeling (Ashraf et al., 2012; Chen et al., 2007; Luizão et al., 2004). Climatic factors directly influence various attributes and processes such as plant growth, soil moisture and nutrients availability to plant (de Toledo et al., 2012; Xu et al., 2015). The vegetation pattern in mountain areas is controlled by three main topographic factors i.e. elevation, aspect and slope (Titshall et al., 2000; Zhao et al., 2013). These topographic factors also define microclimate of an area and elevation is most important among these factors (Dobrowski, 2011; Xu et al., 2015). Climatic factors (temperature and precipitation) are influenced by elevation which in turn controls vegetation spread at larger spatial scales (Sinha et al., 2018). These factors directly affect ruggedness of terrain, curvature of surface, topographic position and flow of water. Similarly, aspect and slope angle greatly influence evapotranspiration (soil-water balance), air temperature and associated flora (Lookingbill and Urban 2004; Bennie et al., 2008). Resource heterogeneity across the landscape is caused by variation in slope angle, aspect or elevation (Wang and Huang, 2012). The measurement of forest structural attributes (composition, stocking and volume) is important to understand temporal changes and long-term vegetation dynamics (Ediriweera et al., 2016). Remote Sensing (RS) and Geographic Information System (GIS) provided state-of-the-art techniques to quantify vegetation dynamics and its spatial distribution along the topographic and climatic range. Digital Elevation Model (DEM) has extensively been used to characterize horizontally distributed objects for modeling topographic attributes (Wulder et al., 2012). Similarly, spatially explicit information is needed to study forest biomass distribution with respect to topographic gradient and climatic zones (McEwan et al., 2011; Meyer et al., 2013). Such investigation requires detection of fine scale interactions of factors that influence spatial distribution of vegetation. Despite numerous studies connecting environmental factors to above-ground biomass (Holl and Zahawi, 2014; Xu et al., 2015), there are few studies available on the influence of elevation on the vertical distribution of forest biomass (Guisan and Zimmermann, 2000; Hansen, 2000; Miller et al., 2004).

2In Pakistan, few researchers have examined the effects of climate on forests. However, there is no published study that quantifies the effect of topography on vegetation patterns across a wide climatic range. In recent era, the development of GIS and remote sensing (RS) enhanced the efficiency to understand structure and composition of forests in relation to topographic and climatic factors. Conventional forest classification of Pakistan was developed about five decades ago based on generalized field data (Champion et al., 1965). This classification needs to be improved keeping in view advancement and emergence of new technologies e.g. RS, GIS and computer modeling during recent decades. Present study is the first ever attempt to improve forest classification of Pakistan for a confined area of three districts of Punjab and KPK provinces and Islamabad Capital Territory. In order to improve the understanding of forest structure and composition in topographically complex landscape, this study was designed to originate forest stratification based on influencing factors such as elevation, aspect, slope, rainfall, temperature, vegetation indices using GIS and RS. The main objective of the present study was to develop and test a new technique for forest stratification on the basis of topographic (elevation, slope and aspect) and climatic characteristics (temperature and rainfall). The relationship between forest types and biomass was also investigated. To address the above mentioned objectives, the study answered the following questions: (i) how do forest types (floristic composition) change along the topographic extent and (ii) to what extent can topographic and climatic factors predict the occurrence of forest types.

2. Materials and methods

2.1. The Study Area

3Present study explored the impacts of topography and climatic on spatial distribution of sub-tropical and moist temperate forests of Pakistan. The objects of the study consist of four administrative units; District Haripur, Islamabad Capital territory, District Abbottabad and Tehsil Murree of Rawalpindi District (fig. 1). Three forest types and two transitional zones were found in the study area include (i) Sub-tropical Scrub Forests, (ii) Sub-tropical Pine Forests, and (iii) Moist Temperate Forests. District Haripur is situated at latitude 33°44' to 34°22' N and longitude 72°35' to 73°15'E and has two forest types i.e., Sub-tropical Pine (STPF) and Sub-tropical Scrub Forests (include Sub-tropical Thorn Forests (STTF) and Sub-tropical Broadleaved Evergreen Forests (STBF). Generally, STBF occur on gentle slopes at an altitude ranging from 548 to 880 m a.s.l. The principal brush wood species is sanatha (Dodonea viscosa) which has covered all the available space whereas the major broad leaved associates are Phulai (Acacia modesta) and Olea (Olea ferruginea) which occur on the steep rocky slopes. Details of STTF and STBF species are given in Table 1. Sub-tropical Pine Forests (STPF) are generally found on northern and north western aspects with elevation ranging from 1,067 to 1,976 m a.s.l. in Haripur district. However, the tallest trees of Pinus roxburghii were present on the best site quality areas at higher altitudes. Whereas Ghora Gali and Murree (Rawalpindi District), STPF are located in the northeastern tip of province of Punjab. Mostly climate of these STPF falls in sub-tropical continental highlands; receives 1,500 to 1,700 mm of annual precipitation in the form of rain and snowfall, however, precipitation amount vary at different parts of the area. The average maximum and minimum temperature is 18ºC and 6ºC respectively but temperature remains high in the May and June. Moist Temperate Forests (MTF) of District Abottabad is situated between latitude 33°55′ and 34°20′ N and longitude 73°20′ and 63°30′ E. These forests have better biomass stock compared to mixed coniferous forests types. The broadleaved associates are oak (Quercus dilitata and Quercus incana), Kandar (Cornus macropylla) and Amlok (Diosphrus lotus).

Fig. 1 - Location map of the study area.
Fig. 1 – Carte de localisation de la zone d’étude.

Fig. 1 - Location map of the study area.Fig. 1 – Carte de localisation de la zone d’étude.

A: Map of Pakistan; B: Map of administrative units around study area; C: Map of three major forest types; D: Photograph of sub-tropical scrub forest; E: Photograph of sub-tropical pine forest; F: Photograph of moist temperate forest. 1. Administrative units of Pakistan; 2. Study area; 3. Sub-tropical scrub forests; 4. Sub-tropical pine forests ; 5. Moist temperate forests ;
A : carte du Pakistan ; B : carte des unités administratives autour de la zone d’étude ; C : carte des trois principaux types de forêt. D : Photographie d’une forêt d’épineux sub-tropicale ; E : Photographie d’une forêt de pins sub-tropicale ; F : Photographie d’une forêt tempérée humide. 1 : unités administratives du Pakistan ; 2 ; zone d’étude ; 3. forêts d’épineux sub-tropicales ; 4. forêts de pins sub-tropicales ; 5. forêts tempéréeshumides ;

2.2. Methods

4Present study integrated RS estimates and field based measurements to assess forest structure and type variations along climatic and topographic range. Present study used RS and GIS based products to understand spatial distribution of different forest types. The data was extracted from Digital Elevation Model (DEM), WorldClim-Global Climate Data and Sentinel-2 images to quantify the impact of climatic and topographic factors on distribution of forest types. The data collection and analysis were completed in following steps:

2.2.1. Forest Inventory and Departmental Data

5The data regarding species composition and forest types was obtained from forest department officials and departmental management plans (working plans) available in Divisional Forest Offices of districts Haripur, Murree and Abbottabad. Forest types were identified with the help of departmental staff and secondary data derived from departmental history files. According to conventional forest classification of Pakistan, six sub-categories were made including Dry Sub-tropical Thorn Forest, Dry Sub-tropical Thorn Forests, Transitional Zone I (Ecotone I), Sub-tropical Pine Forest, Transitional Zone II (Ecotone II) and Moist Temperate Forests. The elevation range and major species are summarized in Table 1. Further, diameter at breast height (cm) and height (m) were measured in 84 field sampling plots (each plot was 0.1 ha) as these are key variables for biomass estimation. Diameter and height data was further used as inputs in allometric equation to estimate above ground biomass. Allometric equations are actually species specific regression equations developed for biomass estimation. Biomass values were then converted into GIS shape file to evaluate its spatial distribution with respect to climatic and topographic factors.

Tab. 1 - Major forest types of the study area and their attributes.
Tab. 1 – Principaux types de forêts de la zone d’étude et leurs caractéristiques.

Tab. 1 - Major forest types of the study area and their attributes.Tab. 1 – Principaux types de forêts de la zone d’étude et leurs caractéristiques.

STTF: Sub-tropical Thorn Forests; STBF: Sub-tropical Broadleaved Evergreen Forests; STPF: Sub-tropical Pine Forests; MTF: Moist Temperate Forests.
STTF : Forêts d’épineux sub-tropicales ; STBF : Forêts sempervirentes sub-tropicales ; STPF : Forêts de pins subtropicales ; MTF : Forêts humides tempérées.

2.2.2. Delineation of Forest Types and extraction of sample points

6Forest types were recognized and delineated using high-resolution satellite imagery in Google Earth. Before digitalization of polygons, detailed consultation was conducted with forest department staff having field experience in the study area. Visual interpretation was also used to identify vegetation composition. Some unclear areas needed field visits for ground truthing. The delineation of major species boundaries along lower slopes where transitional zones occur and converted into continuous forests was certainly subjective. A systematic grid of 966 point locations covering the study area was created in QGIS 2.18, with a random start and spaced 1,122 m apart (supposed to be suitable for spatial independence). Out of 966 random points, 316 points were overlaid in Scrub Forests of district Haripur and Margallah Hills, 155 points were overlaid in Murree Forests and remaining 494 points in Abbottabad Forests. These random points were then overlaid on Google Earth and various types (mentioned in Table 1) were identified with visual interpretation and also validated by field visits at random locations. These random points are coded from “1” to “6” based on forest types mentioned in Table 1. Points in transitional zones or places with mixed species composition were clearly investigated and classes were assigned to them whereas points on bare lands or water bodies were excluded and not used in forest types prediction.

2.2.3. Computation of topographic and climatic variables

7The SRTM 30 m based DEM was acquired from USGS Earth Explorer (https://earthexplorer.usgs.gov/​) and climatic data was downloaded from WorldClim- Global Climate Data website (http://www.worldclim.org/​bioclim). The information related to topographic factors such as elevation, slope and aspect was derived from DEM. Climatic data was used to develop biologically meaningful variables based on the monthly temperature and rainfall values. A total of 966 random points were imported into ArcGIS 10.3 for further analysis followed overlaid these points on DEM. Spatial analyst tool of ArcGIS 10.3 was employed to obtain values of aspect, elevation and slope. Similar to topographic factors; putative predictors among climatic attributes were selected to predict spatial distribution of forest types. Selections of climatic attributes were based on published literature and source of each variable has been provided in Table 2.

Tab. 2 - Range of topographic and climatic factors used in this study.
Tab. 2 - Gamme des valeurs des facteurs topographiques et climatiques de cette étude.

Tab. 2 - Range of topographic and climatic factors used in this study.Tab. 2 - Gamme des valeurs des facteurs topographiques et climatiques de cette étude.

2.2.4. Sentinel-2 Image Processing

8Sentinel-2 image was acquired from Copernicus Open Access Hub (https://scihub.copernicus.eu/​) for the study area (Islamabad Capital Territory and districts Abbottabad, Rawalpindi and Haripur). Sentinel-2 tile had large spatial coverage with 100 km2 area covering outside the study area. Sentinel-2 was rectified in order to reduce atmospheric effects, cloud cover, aerosols particles before analysis (Roy et al., 2016; Ali et al., 2018). Image was rectified through Sen2Cor-2.3.1 plugin in Sentinel-2 Application Programme (SNAP Tool Box) which was used for correction including cloud detection, scene classification, aerosol optical thickness and water vapors and ultimately resulted in rectified image (with bottom of the atmosphere BoA converted) (Louis et al., 2016; Martins et al., 2017). As image spatial area was large, therefore subset of original image was done for four districts of the study area. Normalized Difference Vegetation Index (NDVI) was computed from rectified image (Equation 1). NDVI was used as a powerful indicator to express spatial distribution of vegetation density (Zhu and Liu, 2015). NDVI was imported into ArcGIS 10.3 and pixel values were extracted against each random point. Extracted data was further used to establish relationship between climatic and topographic attributes and forest vegetation.

2.2.5. Statistical Analysis

9Statistical analysis was conducted to examine the relationship of topographic and climatic factors with the forest types and its attributes. Statistical analysis includes correlation, simple linear regression, stepwise linear regression and decision tree analysis (fig. 2). Statistical tests provide efficient analysis of significant variables and its selection in the final model. Dependent variables include forest type and forest biomass which was used separately against independent variables (climatic and topographic). Significant variables were selected and linear regression model was developed. As far as “Decision Tree” is concerned, it classified cases into various groups and based on independent variables (topographic and climatic), it predicted values of target variable (presence of forest type). Decision tree analysis was important because the dataset contains several correlated predictors that may have complex interactions, the reason why this approach was used to investigate relationships between “Forest Type” and the independent variables (Cutler et al., 2007; Das et al., 2015). The growing methods of Decision Tree include Chi-squared Automatic Interaction Detection (CHAID) and Classification and Regression Trees (DCRT). The former select the independent variable which has strongest relationship/interaction with the dependent variable keeping in view significant difference (categories may merge together in case of non-significant difference between predictors). Later CRT method data were categorized into homogenous segments with respect to predictor variable in SPSS version 21 software (Statistical Package for Social Sciences).

Fig. 2 - The data flow in this study.
Fig. 2 - Le flux de données de cette étude.

Fig. 2 - The data flow in this study.Fig. 2 - Le flux de données de cette étude.

3. Results and discussion

3.1. Sub-Tropical Forest Types

10Results of forest stratification have been shown in Figures 3-6. Four major vegetation types were observed within 567 to 2,079 meters (forest area); Sub-tropical Dry Thorn Forests (STTF) which were found at elevation (< 600 meters) followed by STBF at elevation range 880-1,055 m and third and fourth were Ecotone I and STPF at elevation range of (880-1,055 m) and (1,056-1,974 m) respectively (fig. 3). Out of the four types, STTF covered the lowest area (about 10 %) compared to the rest of the three forest types. Slope of the area range from gentle to steep slopes (0-74.5); STTF and STBF were present in gentle slopes while medium slope was observed for STPF. Some places of Ecotone II occurred at steep slope zone. Vegetation density was highest in STBF in Margallah Hills because it has been declared as National Park, therefore excellent vegetation density was maintained in this area. NDVI values were used as proxy indicator of vegetation density as positive NDVI values showed vegetated areas and negative values showed non-vegetated areas. Higher the NDVI values, higher was vegetation density and vice versa. NDVI map also show highest density (0.82) in Margallah Hills compared to other areas of Haripur where NDVI values are low (0.3-07) as shown in Figure 3. Forest of Haripur are facing anthropogenic disturbances including overgrazing, trees cutting for firewood, lopping, forest fires which have frequently reported by forest officials in these forests. Regarding species composition, STTF consists of Acacia nilotica, Prosopis juliflora, Grewia oppositifolia, Dodonea, Acacia modesta, Adhatoda vasica, Cassia fistula; STBF species include Dodonea viscosa, Acacia modesta, Olea ferruginea, Cappris aphylla, Zizyphus and Adhatoda vesica; Ecotone I comprised of Pinus roxburghii , Acacia modesta, Mallotus philippinensis, Ficus, Berbaris lyceum while Pinus roxburghii, Quercus incana in STPF. Major forest types were STBF and most of area is covered by two major communities were observed (Acacia modesta and Olea ferruginea communities). Acacia modesta community has dominated southern aspects and Olea ferruginea community dominated northern aspects. Next to Ecotone-I, pure STPF was started and Pinus roxburghii was most dominant species with broadleaved associate of Quercus incana at lower depression (fig. 5), however Rhododendron arboretum was also observed at higher elevation towards district Abbottabad, Ecotone II was extended from elevation of 1,975 to 2,152 m and ultimately touches with Galies Reserved Forests. Ecotone II were present at upper northern part in Ghoragali where co-dominant Pinus wallichiana occurs with Pinus roxburghii and major broadleaved associate was Aesculus indica (fig. 7).

Fig. 3 - Thematic maps of Haripur and Margallah Hills.
Fig. 3 – Cartes thématriques de Haripur et Margallah Hills.

Fig. 3 - Thematic maps of Haripur and Margallah Hills.Fig. 3 – Cartes thématriques de Haripur et Margallah Hills.

A: Elevation. 1. 466-790 m; 2. 790-1,050 m; 3. 1,050-1,321 m; 4. 1,321-1,667 m; 5. 1,667-2,433 m; B: Elevation of actual forest areas based on field observation; C: NDVI. 1. -0.346; 2. 0.829; D: NDVI of actual forest areas based on field observation.
A : Altitude : 1. 466-790 m ; 2. 790-1 050 m ; 3. 1 050-1 321 m ; 4. 1 321-1 667 m ; 5. 1 667-2 433 m ; B : Altitude étendues forestières actuelles selon les observations de terrain ; C : NDVI. 1. -0,346 ; 2. 0,829 ; D : NDVI des étendues forestières actuelles selon les observations de terrain.

Fig. 4 - Climatic maps of Haripur and Margallah Hills.
Fig. 4 – Cartes climatiques de Haripur et Margallah Hills.

Fig. 4 - Climatic maps of Haripur and Margallah Hills.Fig. 4 – Cartes climatiques de Haripur et Margallah Hills.

A. Annual mean temperature; 1. 13.92°C; 2. 13.92-16.64°C; 3. 16.64-17.48°C; 4.17.48-18.02°C; 5. 18.02-18.41°C; 6. 18.41-19.22 °C; 7. 19.22-19.76 °C; 8. 19.76-21.08 °C; B. Annual Precipitation; 1. 937-940 mm ; 2. 940-1029 mm ; 3. 1029-1073 mm ; 4. 1073-1105 mm ; 5. 1105-1110 mm ; 6. 1110-1191 mm ; 7. 1191-1265 mm ; 8.1265-1376 mm.
A. Températures moyennes annuelles : 1. 13.92°C ; 2. 13.92-16.64°C ; 3. 16.64-17.48°C ; 4.17.48-18.02°C ; 5. 18.02-18.41°C ; 6. 18.41-19.22 °C ; 7. 19.22-19.76 °C ; 8. 19.76-21.08 °C. B. Précipitations annuelles ; 1. 937-940 mm; 2. 940-1029 mm; 3. 1029-1073 mm; 4. 1073-1105 mm; 5. 1105-1110 mm; 6. 1110-1191 mm; 7. 1191-1265 mm; 8.1265-1376 mm.

11The ground flora was consists of Myrsine africana and Berberis lyceum. Regarding climatic factors, annual mean temperature range from 13.92°C to 21.08°C; most of the area (lower elevation) in Haripur and Margallah Hills has high temperature (17.77-21.08°C) whereas higher elevation areas towards district Abbottabad have lower temperature (13.92°C). Annual precipitation range from 942 mm towards Khanpur area of Haripur while central parts of Haripur and Margallah Hills have 1,147-1,164 mm precipitation; however higher annual precipitation (1,376 mm) was recorded at higher elevation (fig. 4). In this context, maximum rainfall of 161 mm was recorded in the month of August and minimum (10 mm) in month of November whereas the highest daily temperature was recorded in June (reaches to 33°C) while daily temperature become minimum (09°C) in the month of January. Whereas in Murree, annual mean temperature range from 19.63°C in the Ecotone I to 13.18°C towards in STPF and Ecotone II whereas Annual Precipitation varies between 873 to 1,102 mm (fig. 6).

12Results of present study was also in similarity to Amir et al., (2018) conducted study on Sub-tropical Pine forests of the study area for estimation biomass and reported field based elevation range from 939 to 1,873 m. Manhas and Raina, (2018) studied the compositions of forest types along the altitudinal gradient and reported that Sub-tropical Pine forests were found across 950 to 1,900 elevation range with moderate to steep slope. Present study results was consistent with Hussain and Illahi (1991) and Ahmed et al., (2006) who reported that STPF were dominated from 830 to 1,870 m a.s.l.

Fig. 5 - Thematic maps of Murree Tehsil.
Fig. 5 – Cartes thématiques de Murre Tehsil.

Fig. 5 - Thematic maps of Murree Tehsil.Fig. 5 – Cartes thématiques de Murre Tehsil.

A: Elevation. 1. 504-894 m; 2. 894-1,160 m; 3. 1,160-1,449 m; 4. 1,449-1,766 m; 5. 1,766-2,275 m; B: Elevation of actual forest areas based on field observation; C: NDVI of Murree Tehsil. 1. -0.47-0.42; 2. 0.42-0.59; 3. 0.59-0.82; D: NDVI of actual forest areas based on field observation.
A : Altitude. 1. 504-894 m ; 2. 894-1160 m ; 3. 1 160-1 449 m ; 4. 1 449-1 766 m ; 5. 1 766-2 275 m ; B : Altitude des étendues forestières actuelles selon les observations terrain ; C : NDVI for Murree Tehsil. 1. -0.47-0.42 ; 2. 0.42-0.59 ; 3. 0.59-0.82 ; D : NDVI des étendues forestières actuelles selon les observations terrain.

Fig. 6 - Climatic maps of Murree Tehsil.
Fig. 6 – Cartes climatiques de Murree Tehsil.

Fig. 6 - Climatic maps of Murree Tehsil.Fig. 6 – Cartes climatiques de Murree Tehsil.

A: Annual mean temperature; 1. Murree Tehsil boundary; 2. Forest boundary; 3. 16.18-17.09°C; 4. 17.09-18.28°C; 5. 18.28-19.62°C; B: Annual precipitation (mm); 1. 873 ; 2. 873-953 ; 3. 953-1,025 ; 4. 1,025-1,102.
A : Températures moyennes annuelles ; 1. Frontières de Murree Tehsil ; 2. Etendue forestière ; 3. 16.18-17.09°C ; 4. 17.09-18.28°C ; 5. 18.28-19.62°C ; B : Précipitations annuelles ; 1. 873 mm ; 2. 873-953 mm ; 3. 953-1 025 mm ; 4. 1 025-1 102 mm.

3.2. Moist Temperate Forests (MTF)

13Results of MTF stratification have been shown in Figures 7 and 8. These forests extended over elevation range from 1,830 to 2,972 m but most of the tract falls in 2,150-2,910 m range. The lower elevation areas (1,830-2,050 m) were mostly transitional zones between STPF and MTF and were present at the southern parts of the study area (fig. 7). Similarly, slope range from medium to steep slope (up to 72 degrees) throughout MTF range while Ecotone I area has gentle slope comparatively. As the whole of MTF area fall in reserved forests which is highest legal category (all activities are prohibited) therefore, dense vegetation was observed. NDVI map showed that dense vegetation has been extended from southern to northern parts where NDVI was maximum (0.65-0.98) as depicted in Figure 7. Three major plant communities were observed in MFT zones; Pinus wallichiana community; Cedrus deodara community and Abies pindrow-Picea smithiana community. Major associate were Taxus, Quercus dilitata, Quercus semecarpofolia, Acer, Aesculus, Prunus, Ulmus, Fraxinus, Alnus, Cornus macropylla and Rhododendron arboretum. Regarding climatic factors; annual mean temperature range from 13.96°C to 15.73°C was recorded moving northern to southern part (Figure 8). Whereas lower western part Ecotone-II has higher temperature (16.03°C) compared to the rest of the forest area. Mean daily temperature reaches to 12.7°C in the month of June (before monsoon) and temperature decrease to 3°C in the month of January. Similarly, annual precipitation (mm) range from 1,145 to 1,376 mm moving from northern to southern part as shown in Figure 8. Overall, maximum rainfall of 355 mm was observed in August (monsoon) and minimum (17.9 mm) in November after monsoon months. Overall, all four forest types (STTF, STBF, STPF and MTF) and two Ecotones (I and II) have been shown in Figure 10. Results of this study was similar to Ahmad et al., (2006) and Hussain and Illahi (1991) who reported that MTF extended from 1,890 to 2,500 m covering areas between Ayubia to Miandam. Similarly Manhas and Raina, (2018) studied altitudinal gradients and slope conditions for MTF in India and reported the elevation range of 1,400 to 3,500 m with steep slopes.

Fig. 7 - Thematic maps of district Abbottabad.
Fig. 7 – Cartes thématiques du district Abbottabad.

Fig. 7 - Thematic maps of district Abbottabad.Fig. 7 – Cartes thématiques du district Abbottabad.

A: elevation of STPF. 1. 733-1,322 m; 2. 1,322-1,601 m; 3. 1,601-1,830 m. B: NDVI of STPF. 1. 0.04; 2. 0.83. C: elevation (m) of MTF. 1. 1,831-2,022; 2. 2,022-2,228; 3. 2,228-2,465; 4. 2,465-2,972. D: NDVI of MTF. 1. 0.08; 2. 0.98.
A : altitude des forêts d’épineux sub-tropicales. 1. 733-1 322  m ; 2. 1 322-1 601  m ; 3. 1 601-1 830  m. B : NDVI des forêts d’épineux sub-tropicales. 1. 0,04 ; 2. 0,83. C : altitude des forêts humides tempérées. 1. 1 831-2 022 m ; 2. 2 022-2 228 m ; 3. 2 228-2 465 m ; 4. 2 465-2 972 m. D : NDVI des forêts humides tempérées. 1. 0,08 ; 2. 0,98.

Fig. - 8. Climatic maps of district Abbottabad.
Fig. 8 - Cartes climatiques du district Abbottabad.

Fig. - 8. Climatic maps of district Abbottabad.Fig. 8 - Cartes climatiques du district Abbottabad.

A: Annual mean temperature (°C). 1. Moist temperate forests boundary; 2. 13.92-14.0; 3. 14.03-15.82; 4. 15.82-16.46; B: Annual precipitation (mm); 1. 1,145-1,163 ; 2. 1,163-1,205 ; 3. 1,163-1,205 ; 4. 1,205-1,286 ; 5. 1,286-1,376.
A : Températures moyennes annuelles. 1. Frontière de la forêt humide tempérée ; 2. 13.92-14.0; 3. 14.03 -15.82; 4. 15.82-16.46; B: Précipitations annuelles d’Abbottabad. 1. 1 145-1 163; 2. 1 163-1 205; 3. 1 163-1 205; 4. 1 205-1 286; 5. 1 286-1 376.

3.3. Relation of Above Ground Biomass and Environmental Factors

14Correlations between forest biomass and influencing factors (elevation, slope, annual mean temperature and precipitation) are summarized in Table 3. Results showed the highest correlation value of 0.75 for annual precipitation versus biomass followed by elevation with coefficient of correlation value of 0.67 (fig. 9). Annual mean temperature also indicated strong positive relationship with coefficient of correlation value of 0.66 (fig. 9). The relationship of biomass and slope has shown lowest correlation (R2= 0.43) compared to the rest of the factors. Relationship between aspect and biomass was also statistically insignificant (R2= 0.01 and p-value was 0.085). Therefore, even if this study focused on relationship of forest biomass (based on forest types) and environmental factors, but also argued that care should be taken in the context of classification based on topographic characteristics which can’t be used as substitute for species mapping. Further, understanding of habitat distribution of sub-tropical and coniferous forests can be enhanced further if environmental variables are incorporated into their classification. Results of present study were also supported by other similar studies such as Lindenmayer et al., (2000), Dawson et al., (2011) and Hjort et al., (2012) who reported that temperature, precipitation, elevation and slope have strong relationship with biodiversity spread and vegetation patterns. Salinas‐Melgoza et al., (2018) also reported that biomass distribution has non-linear relationship with topographic variables (elevation, wetness index, slope, tangential curvature) and further biomass prediction is also influenced by accessibility factors.

Tab. 3 - Correlation between biomass and environmental Factors.
Tab. 3 - Correlation entre la biomasse et les facteurs environnementaux.

Tab. 3 - Correlation between biomass and environmental Factors.Tab. 3 - Correlation entre la biomasse et les facteurs environnementaux.

Fig. 9 - Scatterplots and regression fitted equations.
Fig. 9 – Nuage de points et équations de régressions affines.

Fig. 9 - Scatterplots and regression fitted equations.Fig. 9 – Nuage de points et équations de régressions affines.

A: Slope versus biomass; B: Elevation versus biomass; C: Annual mean temperature versus biomass; D: Annual precipitation versus biomass.
A : versant versus biomasse ; B : altitude versus biomasse ; C : température moyenne annuelle versus biomasse ; D : précipitations annuelles versus biomasse.

3.4. Relation of Forest Types and Environmental Factors

15Correlation between forest types (sub-topical and moist temperate forests) and environmental factors have been summarized in Table 4. Results showed that most of the factors (ten out of eleven) have strong and significant relationship (significance level was 0.01) with forest types while only aspect has non-significant correlation with forest types points. Correlation of elevation was highest (R2= 0.92) followed by annual mean temperature, temperature of driest quarter, warmest quarter and coldest quarter (R2= 0.76). Whereas, annual precipitation, precipitation of driest quarter and coldest quarter have shown better performance with correlation of 0.53, 0.73 and 0.68 respectively. However, slope and precipitation of warmest quarter have shown low correlation with forest types data with the R2 value of 0.27 and 0.11 respectively.

Tab. 4 - Relationship between Forest Types and Environmental Factors.
Tab. 4 – Relations entre les types de forêts et les facteurs environnementaux.

Tab. 4 - Relationship between Forest Types and Environmental Factors.Tab. 4 – Relations entre les types de forêts et les facteurs environnementaux.

16Similarly, stepwise linear regression model between forest types versus environmental variables (topographic and climatic) showed that elevation, precipitation seasonality and annual temperature range (°C) have been selected in the model as their correlation was strongly significant (p-value as 0.00) as shown in Table 5. Whereas rest of the explanatory variables (aspect, slope, annual mean temperature, temperature seasonality, temperature of warmest, coldest, driest quarters, annual precipitation and precipitation of warmest and driest quarters) were insignificant and therefore removed from final models. During model development, best predictors were designated in five “stepwise selection”; predictors were selected in this order; elevation, precipitation seasonality, temperature seasonality and annual temperature range because at final “fifth model selection”. However, temperature seasonality was removed because it was an insignificant correlation with respect to other explanatory variables. Similarly, overall in stepwise selection, coefficient of correlation was increased from 0.854 to 0.859 in the final model. Results of the present study analysis showed that elevation and climatic variables was the best indicator of presence of particular types of forest in an area (from district Haripur to district Abbottabad).

Tab. 5 - Stepwise linear regression between forest types and environmental factors.
Tab. 5 – Régression linéaire séquentielle entre types de forêts et facteurs environnementaux.

Tab. 5 - Stepwise linear regression between forest types and environmental factors.Tab. 5 – Régression linéaire séquentielle entre types de forêts et facteurs environnementaux.

17Therefore, Sub-tropical Dry Thorn Forest, Sub-tropical Broadleaved Evergreen Forests, Sub-tropical Pine Forests and Moist Temperate Forest and transitional zones between major forest types was recognized using these environmental variables (fig.10). The present study results are in accordance with the Hengl et al., (2018) who provide global map of potential vegetation cover with environmental variables and reported that most important predictors of “vegetation classes” estimation were annual precipitation (total), monthly temperatures while in addition, elevation was important predictor for tree cover modeling. Similarly, species mapping, habitat suitability and forest types are determined by climatic variables including coldest month temperature, driest month precipitation and other similar attributes (San-Miguel-Ayanz, 2016). In this context, environmental covariates are used for potential biomass maps (Erb et al., 2017), global potential natural vegetation (Levavasseur et al., 2012; Tian et al., 2016) and global reforestation map (Griscom et al., 2017).

Fig. 10 - Forest types map.
Fig. 10 – Cartes des types de forêt.

Fig. 10 - Forest types map.Fig. 10 – Cartes des types de forêt.

1. STTF; 2. STBF; 3. Ecotone-I; 4. STPF; 5. Ecotone-II; 6. MTF.
1
. STTF; 2.STBF; 3. Ecotone-I; 4. STPF; 5. Ecotone-II; 6. MTF.

3.5. Decision Trees for Forest Types

18Decision Trees were developed to interpret predictor effects in predicting forest types and to explore possible interactions of predictors. Results of decision trees of both growing methods (CHAID- Chi-squared Automatic Interaction Detection and CRT- Classification and Regression Trees) have been summarized in Figure 11 and 12. Both decision trees established for the all predictors (topographic and climatic) except elevation because during field inventory forest types definition was based on specie composition and random points also were generated based on elevation Therefore, two decision trees were developed for full dataset of all predictors excluding elevation in order to produce un-biased results. Importance of other explanatory factors (except elevation) has been assessed by both growing methods (CHAID and CRT). In CRT, first major split was based on temperature of the driest quarter into six nodes ranging from (≤ 10.4°C) to (15.76°C). Secondly, annual precipitation, precipitation seasonality and slope were the important for forest type’s identification (fig. 11).

Fig. 11 - Decision tree by CRT method for identification important predictors.
Fig. 11 – Arbre de décision par la méthode CRT pour l’identification des facteurs explicatifs importants.

Fig. 11 - Decision tree by CRT method for identification important predictors.Fig. 11 – Arbre de décision par la méthode CRT pour l’identification des facteurs explicatifs importants.

19In contrast, CHAID growing method showed bit different result compared to CRT, six predictors were selected as important for forest type prediction which include precipitation seasonality, temperature seasonality, temperature of the driest quarter, annual precipitation, slope and annual mean temperature (fig. 12). Therefore, the present study found that vegetation pattern and types were strongly based on bioclimatic factors and their interactions. Moreover, assessment of predictors of specific importance for forest type prediction and among all predictors, elevation was the best predictor and has significant relative importance with respect to other explanatory variables. The present study findings were inconsistent with Das et al., (2018) who identified important topographic and climatic factors and based on these factors predicted occurrence of forests.

Fig. 12 - Decision Tree by CHAID method for identification of important variables.
Fig. 12 – Arbre de décision par la méthode CHAID pour l’identification des variables importantes.

Fig. 12 - Decision Tree by CHAID method for identification of important variables.Fig. 12 – Arbre de décision par la méthode CHAID pour l’identification des variables importantes.

4. Conclusions

20Climate is the controlling factor for vegetation at larger scale. However, topography influence vegetation patterns and limits the spatial distribution of forests at local and regional scales. Moreover, topography can be used as a powerful predictor of forest ecosystems which combined with other climatic variables that directly influence various attributes such as plant growth, type of soil and water contents and their mutual interaction. Present study have explored topographic and climatic variables in relation to vegetation spatial distribution and investigated their relative importance for forest type’s prediction. This forest stratification further strengthened the existing classification because it has provided finer spatial extent and wall-to-wall coverage of Sub-tropical and Moist temperate forests. Selection of topographic and climatic strata was based on literature review and field inventory of Sub-tropical and Moist temperate forests. Field inventory showed that species distribution was largely influenced by topographic and climatic factors and vegetation spatial extent was limited by these factors. It was also observed that northern aspects have good vegetation compared to southern aspects. Moreover, slope angle also affect vegetation density and presence of associated species. For example, broadleaved associate (Quercus incana) was found on gentle slopes in Sub-tropical Pine forest. Further, precipitation and temperature were strongly auto-correlated and species composition was changed along these environmental gradients.

21Result of the study have shown that elevation was the most important predictor of particular forest type presence followed by annual mean temperature, temperature of driest quarter, warmest quarter and coldest quarter. Whereas other variables have also secondary importance which include annual precipitation, precipitation of driest quarter and coldest quarter; that can also assist to explore minute micro climate details. Moreover, Decision Trees were also used to explore possible interactions of predictors and to determine important variables. Decision Trees results was also similar to that of the regression models and showed elevation was the most important factor that predicted particular forest type. Moreover, temperature of the driest quarter, annual precipitation, precipitation seasonality, slope, temperature seasonality and annual mean temperature were identified as important variables. Therefore, present study concluded that forest types prediction was strongly based on climatic variables; however elevation was the best topographic predictor with significant relative importance comparatively. Therefore, the spatial mapping based on climatic and topographic factors has provided wall-to-wall species composition details at finer scales. The differences in forest structure along topographic and climatic range were of great significance for better planning and management of natural ecosystems.

*Corresponding Author. Tel : +92 051- 9290019
naveedahmad795@gmail.com (N. Ahmad).

Haut de page

Bibliographie

Ahmed M., Husain T., Sheikh A.H., Hussain S.S., Siddiqui M.F. (2006) - Phytosociology and structure of Himalayan forests from different climatic zones of Pakistan. Pakistan Journal of Botany, 38(2), 361-383.

Ali A., Ullah S., Bushra S., Ahmad N., Ali Asad., Khan M.A. (2018) - Quantifying forest carbon stocks by integrating satellite images and forest inventory data. Austrian Journal of Forest Science, 135 (2), 93–117.

Amir M., Liu X., Ahmad A., Saeed S., Mannan A., Muneer, M.A. (2018) - Patterns of Biomass and Carbon Allocation across Chronosequence of Chir Pine (Pinus roxburghii) Forest in Pakistan: Inventory-Based Estimate. Advances in Meteorology, 2018, 1-8.

DOI: 10.1155/2018/3095891

Ashraf M.I., Bourque C.P.A., MacLean D.A., Erdle T., Meng F.R. (2012) - Using JABOWA-3 for forest growth and yield predictions under diverse forest conditions of Nova Scotia, Canada. The Forestry Chronicle, 88 (6), 708-721.

DOI: 10.5558/tfc2012-137

Bennie J., Huntley B., Wiltshire A., Hill M.O., Baxter, R. (2008) - Slope, aspect and climate: spatially explicit and implicit models of topographic microclimate in chalk grassland. Ecological modelling, 216 (1), 47-59.

DOI: 10.1016/j.ecolmodel.2008.04.010

Busing R.T., White P.S., MacKenzie M.D. (1993) - Gradient analysis of old spruce–fir forests of the Great Smoky Mountains circa 1935. Canadian Journal of Botany, 71 (7), 951-958.

DOI: 10.1139/b93-107

Champion S.H., Seth S.K., Khattak G.M. (1965) - Manual of silviculture for Pakistan. Pakistan Forest Institute, Peshawar, Pakistan.

Chen X.F., Chen J.M., An S.Q. Ju W.M. (2007) - Effects of topography on simulated net primary productivity at landscape scale. Journal of Environmental Management, 85, 585–596.

DOI: 10.1016/j.jenvman.2006.04.026

Cutler D.R., EdwardsJr T.C., Beard K.H., Cutler A., Hess K.T., Gibson J., Lawler J.J. (2007) - Random forests for classification in ecology. Ecology, 88 (11), 2783-2792.

DOI: 10.1890/07-0539.1

Das A., Nagendra H., Anand M., Bunyan M. (2015) - Topographic and bioclimatic determinants of the occurrence of forest and grassland in tropical montane forest-grassland mosaics of the Western Ghats, India. PloS one, 10 (6), 1-19.

DOI: 10.1371/journal.pone.0130566

Dawson T.P., Jackson S.T., House J.I., Prentice I.C., Mace G.M. (2011) - Beyond predictions: biodiversity conservation in a changing climate. Science, 332 (6025), 53-58.

DOI: 10.1126/science.1200303

de Toledo J.J., Magnusson W.E., Castilho C.V., Nascimento H.E. (2012) - Tree mode of death in Central Amazonia: Effects of soil and topography on tree mortality associated with storm disturbances. Forest Ecology and Management, 263, 253-261.

DOI: 10.1016/j.foreco.2011.09.017

Dobrowski S.Z. (2011) - A climatic basis for microrefugia: the influence of terrain on climate. Global change biology, 17 (2), 1022-1035.

DOI: 10.1111/j.1365-2486.2010.02263.x

Ediriweera S., Danaher T., Pathirana S. (2016) - The influence of topographic variation on forest structure in two woody plant communities: A remote sensing approach. Forest systems, 25 (1), 11.

DOI: 10.5424/fs/2016251-07123

Erb K.H, Kastner T., Plutzar C., Bais A.L.S, Carvalhais N., Fetzel T., Gingrich S., Haberl H., Lauk C., Niedertscheider M. (2017) - Unexpectedly large impact of forest management and grazing on global vegetation biomass. Nature, 553 (7686), 73-76.

DOI: 10.1038/nature25138

Griscom B.W., Adams J., Ellis P.W., Houghton R.A., Lomax G., Miteva D.A., Schlesinger W.H., Shoch D., Siikamäki J.V., Smith P., Woodbury P. (2017) - Natural climate solutions. Proceedings of the National Academy of Sciences, 114 (44), 11645-11650.

DOI: 10.1073/pnas.1710465114. Epub 2017 Oct 16.

Guisan A., Zimmermann N.E. (2000) - Predictive habitat distribution models in ecology. Ecological modelling, 135 (2-3), 147-186.

DOI: 10.1016/S0304-3800(00)00354-9

Hansen A.J., Rotella J.J., Kraska M.P., Brown D. (2000) - Spatial patterns of primary productivity in the Greater Yellowstone Ecosystem. Landscape Ecology, 15 (6), 505-522.

DOI: 10.1023/A:1008127228878

Hengl T., Walsh M.G., Sanderman J., Wheeler I., Harrison S.P., Prentice I.C. (2018) - Global mapping of potential natural vegetation: an assessment of machine learning algorithms for estimating land potential. PeerJ, 6, e5457.

DOI: 10.7717/peerj.5457

Hijmans R.J., Cameron S.E., Parra J.L., Jones P.G., Jarvis A. (2005) - Very high resolution interpolated climate surfaces for global land areas. International Journal of Climatology: A Journal of the Royal Meteorological Society, 25 (15), 1965-1978.

DOI: 10.1002/joc.1276

Hjort J., Heikkinen R.K., Luoto M. (2012) - Inclusion of explicit measures of geodiversity improve biodiversity models in a boreal landscape. Biodiversity and Conservation, 21 (13), 3487-3506.

DOI: 10.1007/s10531-012-0376-1

Holl K.D., Zahawi R.A. (2014) - Factors explaining variability in woody above-ground biomass accumulation in restored tropical forest. Forest Ecology and Management, 319, 36-43.

DOI: 10.1016/j.foreco.2014.01.024

Hussain F., Ilahi I. (1991) - Ecology and vegetation of lesser Himalayas Pakistan. Department of Botany, University of Peshawar, 187 p.

Levavasseur G., Vrac M., Roche D.M., Paillard D. (2012) - Statistical modelling of a new global potential vegetation distribution. Environmental Research Letters, 7 (4), 044019.

DOI: 10.1088/1748-9326/7/4/044019

Lindenmayer D.B., Margules C.R., Botkin D.B. (2000) - Indicators of biodiversity for ecologically sustainable forest management. Conservation biology, 14 (4), 941-950.

DOI: 10.1046/j.1523-1739.2000.98533.x

Lookingbill T., Urban D. (2004) - An empirical approach towards improved spatial estimates of soil moisture for vegetation analysis. Landscape Ecology, 19 (4), 417-433.

DOI: 10.1023/B:LAND.0000030451.29571.8b

Louis J., Debaecker V., Pflug B., Main-Knorn M., Bieniarz J., Mueller-Wilm U., Cadau E., Gascon F. (2016) -Sentinel-2 Sen2Cor: L2A Processor for Users. Proceedings Living Planet Symposium 2016, Prague, Czech Republic. ISBN 978-92-9221-305-3 ISSN 1609-042X, Spacebooks Online, 1-8.

DOI : https://elib.dlr.de/107381/

Luizão R.C., Luizão F.J., Paiva R.Q., Monteiro T.F., Sousa L.S., Kruijt B. (2004) - Variation of carbon and nitrogen cycling processes along a topographic gradient in a central Amazonian forest. Global Change Biology, 10 (5), 592-600.

DOI: 10.1111/j.1529-8817.2003.00757.x

Manhas P., Raina A.K. (2018) - Composition, structure and diversity of tree species along an elevational gradient in Dudu forest range, Jammu and Kashmir, India. Journal of Applied and Natural Science, 10 (3), 981-985.

DOI: 10.31018/jans.v10i3.1847

Martins V.S., Barbosa C.C.F., De Carvalho L.A.S., Jorge D.S.F., Lobo F.D.L., Novo E.M.L.D.M. (2017) - Assessment of atmospheric correction methods for Sentinel-2 MSI images applied to Amazon floodplain lakes. Remote Sensing, 9 (4), 322-345.

DOI: 10.3390/rs9040322

McEwan R.W., Lin Y.C., Sun I., Hsieh C.F., Su S.H., Chang L.W., Song M.G.Z., Wang H.H., Hwong J.L., Lin K.C., Yang K.C., Chiang J.M. (2011) - Topographic and biotic regulation of aboveground carbon storage in subtropical broad-leaved forests of Taiwan. Forest Ecology and Management, 262 (9), 1817–1825.

DOI: 10.1016/j.foreco.2011.07.028

Meyer V., Saatchi S.S., Chave J., Dalling J.W., Bohlman S., Fricker G.A., Robinson C., Neumann M., Hubbell S. (2013) - Detecting tropical forest biomass dynamics from repeated airborne Lidar measurements. Biogeosciences, 10 (8), 5421–5438.

DOI: 10.5194/bg-10-5421-2013

Miller J.R., Turner M.G., SmithWick E.A., Dent C.L., Stanley E.H. (2004) - Spatial extrapolation: the science of predicting ecological patterns and processes. AIBS Bulletin, 54 (4), 310-320.

DOI: 10.1641/0006-3568(2004)054[0310:SETSOP]2.0.CO;2

Roy D.P., Li J., Zhang H.K., Yan L. (2016) - Best practices for the reprojection and resampling of Sentinel-2 Multi Spectral Instrument Level 1C data. Remote Sensing Letters, 7 (11), 1023-1032.

DOI: 10.1080/2150704X.2016.1212419

Salinas‐Melgoza M.A., Skutsch M., Lovett J.C. (2018) - Predicting aboveground forest biomass with topographic variables in human‐impacted tropical dry forest landscapes. Ecosphere, 9 (1), e02063.

DOI: 10.1002/ecs2.2063

Sinha S., Badola H.K., Chhetri B., Gaira K.S., Lepcha J., Dhyani P.P. (2018) - Effect of altitude and climate in shaping the forest compositions of Singalila National Park in Khangchendzonga Landscape, Eastern Himalaya, India. Journal of Asia-Pacific Biodiversity, 11 (2), 267-275.

DOI: 10.1016/j.japb.2018.01.012

Tian H., Lu C., Ciais P., Michalak A.M., Canadell J.G., Saikawa E., Huntzinger D.N., Gurney K.R., Sitch S., Zhang B. (2016) - The terrestrial biosphere as a net source of greenhouse gases to the atmosphere. Nature, 531 (7593), 225-228.

DOI: 10.1038/nature16946

Titshall L.W., O'Connor T.G., Morris C.D. (2000) - Effect of long-term exclusion of fire and herbivory on the soils and vegetation of sour grassland. African Journal of Range and Forage Science, 17 (1-3), 70-80.

DOI: https://doi.org/10.2989/10220110009485742

Wang H.C., Huang C.Y. (2012) - Investigating the spatial heterogeneity of a subtropical montane cloud forest plantation with a QuickBird image, International Journal of Remote Sensing, 33 (24), 7868-7885.

DOI: 10.1080/01431161.2012.703346

Wulder M.A., White J.C., Nelson R.F., Næsset E., Ørka H.O., Coops N.C., Hilker T., Bater C. W., Gobakken T. (2012) - Lidar sampling for large-area forest characterization: A review. Remote Sensing of Environment, 121, 196-209.

DOI: 10.1016/j.rse.2012.02.001

Xu Y., Franklin S.B., Wang Q., Shi Z., Luo Y., Lu Z., Zhang J., Qiao X., Jiang M. (2015) - Topographic and biotic factors determine forest biomass spatial distribution in a subtropical mountain moist forest. Forest ecology and management, 357, 95-103.

DOI: 10.1016/j.foreco.2015.08.010

Zhao Z., Ashraf M.I., Meng F.R. (2013) - Model prediction of soil drainage classes over a large area using a limited number of field samples: A case study in the province of Nova Scotia, Canada. Canadian Journal of Soil Science, 93 (1), 73-83.

DOI: 10.4141/cjss2011-095

Zhu X., Liu D. (2015) - Improving forest aboveground biomass estimation using seasonal Landsat NDVI time-series. ISPRS Journal of Photogrammetry and Remote Sensing, 102, 222-231.

DOI: 10.1016/j.isprsjprs.2014.08.014

Version française abrégée

Haut de page

Annexe

Les forêts naturelles du Pakistan occupent un vaste panel de conditions climatiques et topographiques, depuis la mangrove des plaines littorales les plus basses aux zones subalpines. L’intégration de la topographie aux variables climatiques représente un bon facteur explicatif pour la modélisation forestière. Au Pakistan, la classification forestière conventionnelle a été développée voici cinq décennies ; l’accès aux nouvelles technologies permet d’améliorer cette classification. Cette étude porte sur quatre unités administratives : district Haripur, le territoire de la capitale Islamabad, les zones Abbottabad et Tehsil Murree du district Rawalpindi (fig. 1). Pour améliorer la compréhension de la structure et composition de la forêt d’un paysage topographiquement complexe, l’étude propose une stratification forestière basée sur les variables tels l’altitude, l’orientation, la pente, les précipitations, la température, l’indice de végétation en utilisant les SIG et la télédétection (MNT, Données WorldClim-Global Climate et images Sentinel-2). Les questions principales qui sont étudiées ici sont (i) comment les types forestiers (composition floristique) évoluent au long d’un gradient topographique et (ii) comment les facteurs climatiques et topographiques prédisent l’occurrence des types de forêt. L’inventaire forestier et les données départementales permettent de collecter les informations de base (tab. 1) identifiant les types forestiers : forêts d’épineux sub-tropicales ; forêts sempervirentes sub-tropicales ; forêts de pins subtropicales et forêts humides tempérées. Chacun des types forestiers est extrait à partir des images disponibles ; les critères topographiques et les variables climatiques sont extraits des différents MNT et bases de données (tab. 2), et l’indice NDVI est extrait des images Sentinel-2 rectifiées, avant que l’ensemble des données soit traité statistiquement (fig. 2).

Les forêts subtropicales occupent la zone comprise entre 567 et 2 079 m, se subdivisant, en altitudes croissantes, entre la forêt d’épineux, la forêt sempervirente sub-tropicale et la forêt de pins subtropicale (fig. 3-6). La forêt tempérée humide se trouve entre 1 830 et 2 972 m d’altitude (fig. 7-8). Le tableau 3 résume les résultats de corrélation entre la biomasse et les facteurs environnementaux, les relations les plus fortes étant avec les précipitations annuelles et l’altitude (fig. 9). Les relations entre les types forestiers et les facteurs environnementaux montrent que tous ont une valeur significative à l’exception e l’orientation (tab. 4). La régression linéaire séquentielle montre que l’altitude, les précipitations saisonnières et la variation annuelle de la température sont pertinents (tab. 5, fig. 10).

L’interprétation des effets des facteurs explicatifs sur la prédiction des types forestiers, à partir d’arbres de décision (fig . 11-12) montre le rôle de la température des mois les plus secs, celle des précipitations annuelles, de la saisonnalité des précipitations et de la pente.

L’étude conclut que l’altitude est le facteur explicatif le plus important dans la répartition de la typologie forestière dans le secteur d’étude, outre les facteurs climatiques. Les différences de structure forestière au long d’étendues climatiquement et topographiquement contrastées est d’importance pour mieux planifier et gérer les écosystèmes naturels.

Haut de page

Table des illustrations

Titre Fig. 1 - Location map of the study area.Fig. 1 – Carte de localisation de la zone d’étude.
Légende A: Map of Pakistan; B: Map of administrative units around study area; C: Map of three major forest types; D: Photograph of sub-tropical scrub forest; E: Photograph of sub-tropical pine forest; F: Photograph of moist temperate forest. 1. Administrative units of Pakistan; 2. Study area; 3. Sub-tropical scrub forests; 4. Sub-tropical pine forests ; 5. Moist temperate forests ; A : carte du Pakistan ; B : carte des unités administratives autour de la zone d’étude ; C : carte des trois principaux types de forêt. D : Photographie d’une forêt d’épineux sub-tropicale ; E : Photographie d’une forêt de pins sub-tropicale ; F : Photographie d’une forêt tempérée humide. 1 : unités administratives du Pakistan ; 2 ; zone d’étude ; 3. forêts d’épineux sub-tropicales ; 4. forêts de pins sub-tropicales ; 5. forêts tempéréeshumides ;
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-1.jpg
Fichier image/jpeg, 13M
Titre Tab. 1 - Major forest types of the study area and their attributes.Tab. 1 – Principaux types de forêts de la zone d’étude et leurs caractéristiques.
Légende STTF: Sub-tropical Thorn Forests; STBF: Sub-tropical Broadleaved Evergreen Forests; STPF: Sub-tropical Pine Forests; MTF: Moist Temperate Forests.STTF : Forêts d’épineux sub-tropicales ; STBF : Forêts sempervirentes sub-tropicales ; STPF : Forêts de pins subtropicales ; MTF : Forêts humides tempérées.
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-2.jpg
Fichier image/jpeg, 938k
Titre Tab. 2 - Range of topographic and climatic factors used in this study.Tab. 2 - Gamme des valeurs des facteurs topographiques et climatiques de cette étude.
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-3.jpg
Fichier image/jpeg, 749k
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-4.jpg
Fichier image/jpeg, 40k
Titre Fig. 2 - The data flow in this study.Fig. 2 - Le flux de données de cette étude.
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-5.jpg
Fichier image/jpeg, 308k
Titre Fig. 3 - Thematic maps of Haripur and Margallah Hills.Fig. 3 – Cartes thématriques de Haripur et Margallah Hills.
Légende A: Elevation. 1. 466-790 m; 2. 790-1,050 m; 3. 1,050-1,321 m; 4. 1,321-1,667 m; 5. 1,667-2,433 m; B: Elevation of actual forest areas based on field observation; C: NDVI. 1. -0.346; 2. 0.829; D: NDVI of actual forest areas based on field observation.A : Altitude : 1. 466-790 m ; 2. 790-1 050 m ; 3. 1 050-1 321 m ; 4. 1 321-1 667 m ; 5. 1 667-2 433 m ; B : Altitude étendues forestières actuelles selon les observations de terrain ; C : NDVI. 1. -0,346 ; 2. 0,829 ; D : NDVI des étendues forestières actuelles selon les observations de terrain.
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-6.jpg
Fichier image/jpeg, 5,9M
Titre Fig. 4 - Climatic maps of Haripur and Margallah Hills.Fig. 4 – Cartes climatiques de Haripur et Margallah Hills.
Légende A. Annual mean temperature; 1. 13.92°C; 2. 13.92-16.64°C; 3. 16.64-17.48°C; 4.17.48-18.02°C; 5. 18.02-18.41°C; 6. 18.41-19.22 °C; 7. 19.22-19.76 °C; 8. 19.76-21.08 °C; B. Annual Precipitation; 1. 937-940 mm ; 2. 940-1029 mm ; 3. 1029-1073 mm ; 4. 1073-1105 mm ; 5. 1105-1110 mm ; 6. 1110-1191 mm ; 7. 1191-1265 mm ; 8.1265-1376 mm.A. Températures moyennes annuelles : 1. 13.92°C ; 2. 13.92-16.64°C ; 3. 16.64-17.48°C ; 4.17.48-18.02°C ; 5. 18.02-18.41°C ; 6. 18.41-19.22 °C ; 7. 19.22-19.76 °C ; 8. 19.76-21.08 °C. B. Précipitations annuelles ; 1. 937-940 mm; 2. 940-1029 mm; 3. 1029-1073 mm; 4. 1073-1105 mm; 5. 1105-1110 mm; 6. 1110-1191 mm; 7. 1191-1265 mm; 8.1265-1376 mm.
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-7.jpg
Fichier image/jpeg, 2,6M
Titre Fig. 5 - Thematic maps of Murree Tehsil.Fig. 5 – Cartes thématiques de Murre Tehsil.
Légende A: Elevation. 1. 504-894 m; 2. 894-1,160 m; 3. 1,160-1,449 m; 4. 1,449-1,766 m; 5. 1,766-2,275 m; B: Elevation of actual forest areas based on field observation; C: NDVI of Murree Tehsil. 1. -0.47-0.42; 2. 0.42-0.59; 3. 0.59-0.82; D: NDVI of actual forest areas based on field observation.A : Altitude. 1. 504-894 m ; 2. 894-1160 m ; 3. 1 160-1 449 m ; 4. 1 449-1 766 m ; 5. 1 766-2 275 m ; B : Altitude des étendues forestières actuelles selon les observations terrain ; C : NDVI for Murree Tehsil. 1. -0.47-0.42 ; 2. 0.42-0.59 ; 3. 0.59-0.82 ; D : NDVI des étendues forestières actuelles selon les observations terrain.
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-8.jpg
Fichier image/jpeg, 6,5M
Titre Fig. 6 - Climatic maps of Murree Tehsil.Fig. 6 – Cartes climatiques de Murree Tehsil.
Légende A: Annual mean temperature; 1. Murree Tehsil boundary; 2. Forest boundary; 3. 16.18-17.09°C; 4. 17.09-18.28°C; 5. 18.28-19.62°C; B: Annual precipitation (mm); 1. 873 ; 2. 873-953 ; 3. 953-1,025 ; 4. 1,025-1,102.A : Températures moyennes annuelles ; 1. Frontières de Murree Tehsil ; 2. Etendue forestière ; 3. 16.18-17.09°C ; 4. 17.09-18.28°C ; 5. 18.28-19.62°C ; B : Précipitations annuelles ; 1. 873 mm ; 2. 873-953 mm ; 3. 953-1 025 mm ; 4. 1 025-1 102 mm.
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-9.jpg
Fichier image/jpeg, 3,6M
Titre Fig. 7 - Thematic maps of district Abbottabad.Fig. 7 – Cartes thématiques du district Abbottabad.
Légende A: elevation of STPF. 1. 733-1,322 m; 2. 1,322-1,601 m; 3. 1,601-1,830 m. B: NDVI of STPF. 1. 0.04; 2. 0.83. C: elevation (m) of MTF. 1. 1,831-2,022; 2. 2,022-2,228; 3. 2,228-2,465; 4. 2,465-2,972. D: NDVI of MTF. 1. 0.08; 2. 0.98.A : altitude des forêts d’épineux sub-tropicales. 1. 733-1 322  m ; 2. 1 322-1 601  m ; 3. 1 601-1 830  m. B : NDVI des forêts d’épineux sub-tropicales. 1. 0,04 ; 2. 0,83. C : altitude des forêts humides tempérées. 1. 1 831-2 022 m ; 2. 2 022-2 228 m ; 3. 2 228-2 465 m ; 4. 2 465-2 972 m. D : NDVI des forêts humides tempérées. 1. 0,08 ; 2. 0,98.
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-10.jpg
Fichier image/jpeg, 5,9M
Titre Fig. - 8. Climatic maps of district Abbottabad.Fig. 8 - Cartes climatiques du district Abbottabad.
Légende A: Annual mean temperature (°C). 1. Moist temperate forests boundary; 2. 13.92-14.0; 3. 14.03-15.82; 4. 15.82-16.46; B: Annual precipitation (mm); 1. 1,145-1,163 ; 2. 1,163-1,205 ; 3. 1,163-1,205 ; 4. 1,205-1,286 ; 5. 1,286-1,376.A : Températures moyennes annuelles. 1. Frontière de la forêt humide tempérée ; 2. 13.92-14.0; 3. 14.03 -15.82; 4. 15.82-16.46; B: Précipitations annuelles d’Abbottabad. 1. 1 145-1 163; 2. 1 163-1 205; 3. 1 163-1 205; 4. 1 205-1 286; 5. 1 286-1 376.
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-11.jpg
Fichier image/jpeg, 1,9M
Titre Tab. 3 - Correlation between biomass and environmental Factors.Tab. 3 - Correlation entre la biomasse et les facteurs environnementaux.
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-12.jpg
Fichier image/jpeg, 463k
Titre Fig. 9 - Scatterplots and regression fitted equations.Fig. 9 – Nuage de points et équations de régressions affines.
Légende A: Slope versus biomass; B: Elevation versus biomass; C: Annual mean temperature versus biomass; D: Annual precipitation versus biomass.A : versant versus biomasse ; B : altitude versus biomasse ; C : température moyenne annuelle versus biomasse ; D : précipitations annuelles versus biomasse.
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-13.jpg
Fichier image/jpeg, 380k
Titre Tab. 4 - Relationship between Forest Types and Environmental Factors.Tab. 4 – Relations entre les types de forêts et les facteurs environnementaux.
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-14.jpg
Fichier image/jpeg, 858k
Titre Tab. 5 - Stepwise linear regression between forest types and environmental factors.Tab. 5 – Régression linéaire séquentielle entre types de forêts et facteurs environnementaux.
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-15.jpg
Fichier image/jpeg, 838k
Titre Fig. 10 - Forest types map.Fig. 10 – Cartes des types de forêt.
Légende 1. STTF; 2. STBF; 3. Ecotone-I; 4. STPF; 5. Ecotone-II; 6. MTF.1. STTF; 2.STBF; 3. Ecotone-I; 4. STPF; 5. Ecotone-II; 6. MTF.
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-16.jpg
Fichier image/jpeg, 3,0M
Titre Fig. 12 - Decision Tree by CHAID method for identification of important variables.Fig. 12 – Arbre de décision par la méthode CHAID pour l’identification des variables importantes.
URL http://journals.openedition.org/geomorphologie/docannexe/image/14564/img-17.jpg
Fichier image/jpeg, 402k
Haut de page

Pour citer cet article

Référence électronique

Naveed Ahmad, Muhammad Irfan Ashraf, Sabeeqa Usman Malik, Ihsan Qadir, Naeem Abbas Malik et Kashif Khan, « Impact of Climatic and Topographic Factors on Distribution of Sub-tropical and Moist Temperate Forests in Pakistan », Géomorphologie : relief, processus, environnement [En ligne], Articles sous presse, mis en ligne le 25 mai 2020, consulté le 01 octobre 2020. URL : http://journals.openedition.org/geomorphologie/14564

Haut de page

Auteurs

Naveed Ahmad

Department of Forestry and Range Management, PMAS Arid Agriculture University, Rawalpindi, 46 300 Pakistan.

Muhammad Irfan Ashraf

Department of Forestry and Range Management, PMAS Arid Agriculture University, Rawalpindi, 46 300 Pakistan.

Sabeeqa Usman Malik

Department of Forestry and Range Management, PMAS Arid Agriculture University, Rawalpindi, 46 300 Pakistan.

Ihsan Qadir

Department of Forestry and Range Management, Bahauddin Zakariya University, Multan, 60 800 Pakistan.

Naeem Abbas Malik

Institute of Geo-Information and Earth Observation, PMAS Arid Agriculture University, Rawalpindi, 46 300 Pakistan.

Kashif Khan

Institute of Geographic Information System, School of Civil and Environmental Engineering, National University of Sciences and Technology, Islamabad, 44000 Pakistan.

Haut de page

Droits d’auteur

© Groupe français de géomorphologie

Haut de page
  • Logo CNRS - Institut des sciences humaines et sociales
  • Logo Groupe français de géomorphologie
  • OpenEdition Journals
Search OpenEdition Search

You will be redirected to OpenEdition Search