Introduction

The climate change and the exacerbation of drought stress in recent decade have caused extensive damage to forests worldwide. However, the trend of forests drought inducing decline and its consequences are still little understood and require further investigation (Martin et al. 2015). Drought and heat stresses lead to physiological disorders in plants due to the drought stress and associations of other mediating factors including pest outbreak and disease attacks and fire (Allen et al. 2010). Almost a large part of the world’s forests mainly in the Mediterranean areas, Alps, Oceania and the Pacific have been affected by this phenomenon. Therefore, assessment of variations in forest cover and density due to climate change and socioeconomic factors is essential for the proper management of forest ecosystems (Wiens 2009). Climate change can be severely effective on the forest sustainability and its benefits (Breck et al. 2018). Usually, in most cases, the lack of soil moisture is the main cause of forest trees decline and, consequently, the prevalence of pests and diseases (Smith et al. 2015). According to Zhang et al.’s (2017) report on climate variables, the mortality (tree death) probability decreased with the increasing mean annual temperature and previous summer mean temperature. Remote sensing has enormous potential for precise mapping of tree mortality and early detection and would enable better monitoring, conservation and effective forest management intervention (Cobb et al. 2017) in a rapidly changing climate. By contrast, the mortality probability increased with the increasing previous winter mean minimum temperature and annual heat–moisture index. Generally, identification of factors effective on forests decline including the species exposed to decline, decline intensity, dead trees age, assessment of the effect of drought on the secondary factors of decline (the prevalence of pests and diseases) and monitoring decline over time have attracted much attention from forest ecosystem researchers. Martin et al. (2015) investigated the annual decline of Fagus sylvatica forests in the Europe. The results showed that about 33% of beech trees and 40% of the forest floor grasses declined over the past 50 years due to drought and lack of moisture along with the prevalence of pests and diseases, most of which was due to the severe drought in 1995. According to Metcalfe and Bradford (2008), in the tropics, the trees with a diameter at breast height more than 10 cm were commonly affected by drought, and the trees of Elaeocarpaceae family were generally more susceptible to drought with a higher decline rate detected in 2005 (Metcalfe and Bradford 2008). In Iran, the forest decline is mostly occurred in the Zagros oak forests. Quercus persica is the dominant species in the Zagros vegetative zone. According to the estimates reported by the Forests, Rangelands and Watershed Management Organization, the forest decline phenomenon is spread over nine provinces in Zagros, and based on the findings of Heshmati et al. (2018), dust storms, improper and heavy grazing, forest clearance and over-harvesting of groundwater are among the aggravating factors in this regard. Also according to Park and Lee (2016) report, the results also revealed that the largest cause of deforestation was the establishment of settlements followed by conversion to cropland and grassland. However, climate change is the main affecting factor of Zagros forests. According to the results of Mohammadi and Taghavi (2005), climate change is based on the reduction in precipitation and the increase in air temperature and evapotranspiration compared with the long-term average. Behmanesh et al. (2015) stated that these changes caused drought and heat stresses, which ultimately exacerbates the plant physiological disorders, increased prevalence of pests and diseases and fire. In the event of continued drought and the current inefficient management of natural resources, the elimination of forests, especially in semiarid areas, is very likely. The purpose of this research was to determine the annual trend of forest decline intensity and the factors exacerbating, using remote sensing technique together with field measurements during 6 years in Zagros forests. This study has been carried out in Kalehzard forest, Kermanshah, Iran, in 2012–2017.

Materials and methods

Study area

Kermanshah province has 520,000 ha of forest land with the dominant species of oak (Quercuse brantii). This research was performed in a part of Kermanshah forests (Kalehzard Village), across 707 ha area. The study area was located 120 km southwest, Kermanshah (Fig. 1). The average annual precipitation and temperature are 450 mm and 15.2 °C, respectively. Winters are cold enough for temperatures to drop to below zero for 90 days on average in December, January, and February; however, summers are hot to cool and dry, indicating a semiarid region. The mean slope of the whole area is 25%. The land-use change is observed in the altitude range between 1427 and 1606 m above sea level. The highest and lowest altitude of the study area is 1427 and 1935 m, respectively. About 90% of this site is located on the marl formations which is susceptible to erosion. Field observations indicated severe degradation and conversion of forest to arable lands. The dominant species of the study forest is Quercus brantii Lindl with seedling and coppice growth. The tree and shrub species associated with oak included Acer monspessulanum, Crataegus aronia, Pistacia atlantica, Amygdalus orientaalis, Daphena mucronata and Cerasus spp.

Fig. 1
figure 1

Geographic location of the study area in Zagros forest in the Kalehzard, Kermanshah Province of Iran

Full size image

Methodology

To do this research, the satellite images with the resolution Pan 2/5 IRS related to the previous years (2005–2017 in the peak growth season) were prepared. Then, geometric and atmospheric corrections of satellite images were performed. Eventually, the satellite images were georeferenced using the field survey and determining the latitude and longitude of control points. The distance and spatial distribution of oak decline were determined using spatial analyst and weighted evidence method considering the location of trees decline and the ground layers including red and near-red infrared bands, slope and aspect used in modeling. In the zoning map of forest decline, the red colors indicate the areas with maximum likelihood of decline. Then, in the decline zone, the study area was divided into the northern and southern slopes. Then, 20 plots of 50*50 m were established in the decline area and satellite images. Inside the plots, the crown cover, number of trees by species and number of dead trees were recorded in two directions (N and S). Each year, after identifying the dead trees, the precise location of these trees was recorded by GPS. The precise coordinates of dead trees were recorded every year during 2011–2017, and these points were checked with satellite images. Based on these points, the distribution map of forest decline was obtained within the study forest area (Fig. 2). To analyze the oak decline trend with climate change in the region, the statistics of the Eslamabad-e Gharb Synoptic Station, which is adjacent to the research site, were investigated.

Fig. 2
figure 2

Spatial die-off forest tree in the study site using Scatter plot software

Full size image

Results and discussion

Climate effects on forest decline

The synoptic meteorology station of Eslamabad-e Gharb, located 30 km west of the study area, covers a significant part of the forests in Kermanshah Province and represents the climatic conditions of the study area. Therefore, the statistical data of a 31-year period are presented in Table 1. According to the data, the changes in precipitation pattern and temperature of 7 years compared with the long-term average are as follows:

Table 1 Meteorological data of the synoptic station of Eslamabad-e Gharb
Full size table

  • During the study years (2012–2017), rainfall decreased by about 23 mm;

  • The winter precipitation decreased by about 37 mm;

  • The average annual temperature increased by about 0.14 °C, which often occurred in the hot season of each year with being higher in August than in other months (in some cases more than 1 °C);

  • The average number of long-term frost days in the area was reduced by about 10 days (Table 1).

According to Soltani et al. (2016), the climate change trend in Zagros forests based on temperature, precipitation, number of snowy and frost days showed that over the past 15 years (1995–2010), the number of hot days increased as much as 26 days per year. In addition, the number of days with a rainfall more than 20 mm has decreased. The reduced number of frost days as well as the reduced ratio of snowfall to total annual rainfall is reported in Zagros forests including our study site. The findings reported by Irannezhad et al. (2017) showed that the reduced snowfall relative to total annual rainfall has been considerable over the last century. Generally, these changes will reduce the forest regeneration and exacerbate their decline. It is more acute in the southern and eastern slopes due to the lack of soil moisture, accelerating the decline of upper stratum trees by reducing the crown cover, shadow and the albedo coefficient (Redmond and Kelsey 2018). Perhaps one of the causes of the Zagros forests decline could be attributed to the fluctuations of rainfall and temperature indices, which is evident in the Eslamabad-e Gharb station and generally the Zagros area. The reduced winter rainfall and rising summer temperature are the most important indicators of climate fluctuations, causing reduced available water for forest trees, deficiencies in tree growth and germination and eventually forest decline, which lead to the conversion of forest ecosystem to the rangelands (Verhoeven et al. 2018).

Overall, the average annual precipitation and temperature adversely had changed by the effect of recent drought leading to forest mortality. Therefore, as a source of needed moisture, the oak root system is immune from evaporation, for it is saved in the depth of soil layer with higher soil organic carbon content. In contrast, soil disturbance factors such as tillage practice, wild fire occurrence, heavy animal grazing and surface soil erosion cause decline soil quality, particularly soil organic carbon promoting evaporation phenomenon and soil moisture lost.

Forest decline zonation

To do this research, 97 forest points (patches) within the site were field measured accurately and their coordinates were recorded using GPS (Table 2). Then, the coordinates were transferred to the IRS satellite image (2016) and processed using the NV software (version 4.5). In the next stage, the decline points were evaluated and the frequency of healthy and dead trees was calculated in the 2500 m2 plots. The spatial distribution of decline is shown in Fig. 2, prepared using the scatter plot diagram.

Table 2 Geographic location and height of the dead trees in the village of Kalehzard
Full size table

Furthermore, the results of the normality test for the field data recorded by GPS (the number of dead trees) showed that the data were normally distributed (Fig. 3a). Also, zonation map of forest decline is shown in Fig. 3b. The distance and spatial distribution of oak decline were determined using spatial analyst and weighted evidence method across the study area. In other words, considering the location of trees decline and the ground layers including red and near-red infrared bands, the slope and aspect were used in modeling. In the zoning map of forest decline, the red colors indicate the areas with maximum likelihood of decline.

Fig. 3
figure 3

Field data recorded by GPS, a normality analysis of dead trees in the study site, b forest decline zonation in the study site

Full size image

Forest decline is related to prolonged drought in the Zagros region (Fig. 4).

Fig. 4
figure 4

Die-back forest phenomenon in related to drought hit in Zagros forest, Kermanshah, Iran, in 2006, 2011, 2014 and 2017 on satellite image

Full size image

Clinical signs of dead trees

The field signs of decline were divided into three classes including partial crown cover decline, full decline and decline caused by diseases and pests (Table 2 and Fig. 5). Generally, the decline of coppice was more than that of seedlings. Moreover, in the early stages of the decline, the upper crown cover was collapsed downward. In advanced stages, the effects of pest and fungal disease are gradually observed as a secondary factor (Safaee et al. 2017a). For instance, the trees are subjected to pest infestation (wood borer beetles) and charcoal diseases at this stage. Field surveys showed that the wood borer beetles and the causal agent of charcoal oak disease (Biscogniauxia mediterranea) are widely distributed in Zagros as important factors in increasing forest trees decline. Although a number of researchers believe that pests and diseases are the main cause of decline in these forests, a large number studies of reported pests and diseases as secondary causes of forest trees mortality (decline). Safaee et al. (2017a) demonstrated that oak charcoal disease caused by Biscogniauxia mediterranea and Obolarina persica is the most important disease in Zagros forests. Field observations have shown that successive drought stress in recent years has resulted in predisposing Zagros forest trees to losses caused by the disease and wood boring insects (Safaee et al. 2017b). It has been indicated that B. mediterranea is a weak pathogen which acts in an aggressive manner under drought stress and high abnormal temperature (Vannini et al. 2009). Therefore, it is concluded that the drought stress can play a major role in oak decline in Zagros forests.

Fig. 5
figure 5

Severe dead oak tree due to land-use change and tillage practice in understory of forest (a–c), the whole dead of seedling oak (d), infected by Biscogniauxia mediterranea disease (e) and Biscogniauxia mediterranea pest (f)

Full size image

Changes in the crown cover of forest trees

The ANOVA of tree cover variables is presented in Table 3. The results demonstrated that during the study period (2012–2017), changes in the crown cover were divided into three significant periods. Accordingly, the tree crown cover in the first period (2012), the second period (2013–2015) and the third period (2016–2017) were about 35, 30 and 28%, respectively, which had significant differences (P < 0.05). In other words, during the study period, the trees crown cover was reduced about 7%. As already presented in Table 2, the decrease in the crown cover of forest trees in the southern slope was significantly higher. In other words, a significant part of change in the trees crown cover has occurred on the southern slopes. However, these changes also varied year by year that this difference is dropped in the last 3 years (2015–2017), so that the change in the southern slope is about 7% less than that in northern slope.

Table 3 Statistical ANOVA output for the assessed variables during 6 years (2011–2018) in Zagros forest
Full size table

Changes in forest tree density

The changes in the density of forest trees in the northern slope for the three groups were 100% (2012–2013), about 94% (2014) and about 87% (2016–2017), showing a decreased density of 13% from the beginning of the study period to the end. This trend has been much more severe in the southern slope, divided into five separate periods including 82% (2013), 77% (2014), 72% (2015), 66% (2016) and 61% (2016–2017; Table 1). Given the above trend, except for the last 2 years, forest tree density often significantly reduced as the study period progressed, and the tree density decreased by totally 20%. It also could be expressed by the year-to-year results of the two dominant southern and northern slopes. Evaluation of the relationship between forest trees density and decline rate using Landsat satellite images and their periodic comparison indicated that decline is inversely proportional to altitude at sea level, so that it is more intense at low altitudes, and vice versa, it is sharply reduced with increasing altitude, especially in slopes where the moisture is higher (Weisberg et al. 2016). Human factors and unprincipled agricultural activities exacerbate the forest decline through reducing the biodiversity and disturbing lands including disturbing the soil surface and releasing organic carbon, leading to (causing) global warming (MariaIordan et al. 2018).

The trend of forest trees decline

According to Table 3, the decline rate during the period 2012–2017 in the northern and southern slopes was 18.8% and 31.1%, respectively, indicating a higher decline rate in the southern slope based on ANOVA results (Table 4). Also, comparing the decline rate of each year in the two northern and southern aspects indicates the annual change. The comparison of the mean of data showed significant differences between the northern and southern slopes as follows:

Table 4 Statistical t test output for the assessed variables during 6 years (2011–2018) between northern and southern slope aspect in Zagros forest
Full size table
  1. A.

    Northern aspect slope: Accordingly, in the first 3 years of 2012–2014 in one group (about 5% every year, a total of 15%), while in the next 2 years, from 2013 to 2016, the decline rate is significantly lower (in total 3.8%). Eventually, in the last year (2017), the decline rate was reduced to its lowest level (less than 1%).

  2. B.

    Southern aspect slope: The decline rate differed significantly (P < 0.05) year by year. The highest and lowest decline rate was obtained to be 8.8% (2014) and 1.1% (2017), respectively.

The higher decline phenomenon and its exacerbation in the first 3 years (2012–2014) are associated with reduced rainfall and increased air temperature. In addition, according to field observations, the decline intensity was higher in the margins of landslides, plowed areas, fire areas and the margin of farms and waterways. Zagros forests in Kermanshah province have decreased by about 38% during 1955–2017 mainly due to population growth, increase in agricultural machinery and ineffective official control as well as land prices enhancement (Gheitury et al. 2019). According to the findings of Fang et al. (2018), although the intensity of decline was proportional to the intensity of drought, the intensity of tree decline due to various factors such as forest degradation factors caused the exacerbation of decline, whose important consequence was the occurrence of dust phenomena. On the other hand, the role of topographic factors should not be ignored, so that the decline is usually higher at low altitudes and southern slopes. According to the findings reported by Malekian et al. (2009) in Zagros forests in Ilam province, the forests near the village with northern and western slopes and a relatively deeper soil are cultivated up to a slope gradient of 30%, resulting in severe decline in vegetation cover and regeneration of forest trees as well as soil erosion. Were et al. (2013) investigated vegetation changes in the Kenyan forests over a 40-year period using the Landsat images of MSS, TM and ETM+ and stated that one percent of the forest area was reduced annually, while the area of agricultural land and residential areas increased by 6% and 16%, respectively. In Europe, the role of topographic factors in the decline rate and the damage caused by pests and diseases showed that the most damages of wood borer beetles to the pine forests were observed in the altitude range of 800 to 100 m in the southern and western slopes (Nelson et al. 2007). Moreover, the relationship between density and decline intensity of conifers using satellite images during 25 years showed that topographic and climatic factors played a significant role in forest decline. Accordingly, the intensity of decline was higher at low altitudes and in years with less rainfall or drought (Van Gunst et al. 2016).

The factors exacerbating the Zagros forests decline

The field observations of this study showed that forest decline was higher in the following conditions:

  • The landslide margin, especially in the southern slopes, with an average decline rate up to 20% higher;

  • The margin of dry land farming and the illegal cultivated areas;

  • In these areas, due to the reduction in albedo coefficient, the poor soil organic matter and surface erosion, the loss of soil moisture content is higher;

  • Areas where forest trees have become weakened due to the grazing, tillage practice and fire, or as fragmented forest areas.

Although the intensity of decline was proportional to the intensity of drought, the intensity of tree decline due to various factors such as forest degradation factors caused the exacerbation of decline (Fang et al. 2018), whose important consequence was the occurrence of dust phenomena. In order to predict the spatial distribution of forest degradation and detect the factors affecting forest degradation, Arekhi (2012) studied a forest area in the north of Ilam province using MSS and TM satellite data. The results showed that during the 31 years (1976–2007) about 20,000 hectares of these forests area were reduced, indicating severe degradation in areas close to the forest and non-forest boundaries. In addition, slope, distance from population centers and roads have an inverse relationship with degradation and as the altitude increases, the degradation rate decreases. According to the findings Malekian et al. (2009) in Zagros forests in Ilam province, the forests near the village with northern and western slopes and a relatively deeper soil are cultivated up to a slope gradient of 30%, resulting in severe decline in vegetation cover and regeneration of forest trees as well as soil erosion. Were et al. (2013) investigated vegetation changes in the Kenyan forests over a 40-year period using the Landsat images of MSS, TM and ETM+ and stated that 1% of the forest area was reduced annually, while the area of agricultural land and residential areas increased by 6% and 16%, respectively. In Europe, the role of topographic factors in the decline rate and the damage caused by pests and diseases showed that the most damages of wood borer beetles to the pine forests were observed in the altitude range of 800 to 100 m in the southern and western slopes (Nelson et al. 2007).

Conclusion

According to the results of this research, the decline trend is proportional to drought occurrence, so that based on climate indices, changes in rainfall, temperature and number of frost days have occurred in recent years. This process has led to the emergence of forest decline phenomena, the severity of which is a function of topographic factors and human activities. In general, the average decline during the 6 years studied (2012–2017) was 4.2% per year. Moreover, the decline rate is a function of the dominant slope direction. This research showed that decline in the two dominant northern and southern slope aspects was about 19 and 31%, respectively, which is significantly higher in the southern slope direction. Also, in 2017, the decline rate in the northern and southern slopes directions decreased by 0.3 and 1%, respectively. This can be attributed to ecological balance with the current capacity of the forest habitat. In addition to topographic factors, soil erosion and human activities also exacerbate forest decline, which is more severe in the margin of areas where landslide has occurred and the margin of rain-fed farming and plowed areas. Therefore, although the Zagros forests are faced with the decline caused by climatic conditions, topography and human factors also exacerbate it. In general, due to the persistence of drought stress and physiological weakness, the conditions for the onset of pest and disease are also provided, which is obviously wood borer beetles and charcoal diseases. Finally, human factors and improper management in various forms such as converting forest into arable lands and construction activities (roads, mines, pipelines and various facilities) worsen the forest decline because of disrupting the soil profile, increased organic carbon emissions, soil erosion and raised bare soil surface promoting rapid soil moisture lost.