Oak trees decline; a sign of climate variability impacts in the west of Iran

Oak trees decline; a sign of climate variability impacts in the west of Iran

Attarod P¹*, Sadeghi S.M.M¹, Pypker T.G², Bayramzadeh V³

1. Department of Forestry and Forest Economics, Faculty of Natural Resources, University of Tehran, Karaj, Iran.

2. Department of Natural Resource Sciences, Faculty of Science, Thompson Rivers University, Kamloops, Canada.

3. Department of Wood Science, Faculty of Agriculture and Natural Resources, Karaj Branch, Islamic Azad University, Karaj, Iran.

* Corresponding author’s E-mail: attarod@ut.ac.ir

(Received: May. 06. 2017 Accepted: Sep. 24. 2017)

The Persian oak, Quercus brantii, trees in the Zagros region of Western Iran have been in decline since 2000. The decline is assumed to be highly connected with changes in meteorological parameters. Our objectives were to quantify the long-term trends in meteorological parameters and reference evapotranspiration (ET₀) in the Zagros region and estimate ecohydrological parameters highly affected by climate variability and related to rainfall interception (I) process (i.e., canopy storage capacity (S), the ratio of mean evaporation rate from the wet canopy to the mean rainfall intensity (E/R), and the free throughfall coefficient (p)). Long-term (1961-2010) changes in air temperature (T), precipitation (P), and wind speed (WS) were obtained from six synoptic meteorological stations located in the region. Throughfall (TF) was measured using the sixteen rain gauges randomly located underneath the crown of the five individual trees. P was measured using rain gauges fixed in an open space nearby to the oak trees .I was computed as the difference between P and TF. From 2000 to 2010, meteorological parameters and ET₀ changed slightly; T, WS, and ET₀ increased (+0.6 ˚C, +0.4 m.s^-1, +0.25 mm.day^-1, respectively), while P decreased (-60 mm).When climate patterns between 1961-2010 were analyzed for 6 synoptic weather stations, P significantly decreased significantly at one station, whereas T significantly increased at two stations and significantly decreased at another. I was estimated to be 40% and 25% within in-leaf and leaf-less periods, respectively. During the in-leaf period, the mean values of S, E/R, and p were roughly estimated to be 1 mm, 0.22, and 0.23, respectively. Our results indicate that the Zagros region is getting warmer and oak trees will indisputably experience reduction in the available water because of increased evaporative loss.

INTRODUCTION

Decline/dieback of trees is a condition characterized by episodes of premature, progressive loss of trees or stand vigor without clear evidence of physical injury or attack by a primary disease or pest. Several biotic and abiotic factors have been considered important in tree health studies, such as climate conditions, drought, storms, heat (Drobyshev et al. 2008; Bolte et al. 2010), insect damage (Moraal & Hil­szczanski 2000), disease

outbreaks (Mistretta 2002), or human interventions induced influences such as cli­mate change, air pollution and fires (Signell et al. 2005; Kabrick et al. 2008). Over the last few years, millions of oak (Quercus spp.) in the world have been affected by a complex disease known as oak decline, oak dieback or oak mortality, depending on the area and the particular case taken into consideration (González Alonso 2008).

The tree decline has heavily affected oak species in many countries e.g., Hungary, Austria, Slovakia, Turkey as well as Iran (Hämmerli & Stadler 1989; Freer-Smith & Read 1995; Führer 1998; Thomas & Büttner 1998; Davari & Askari 2005). Determining the reason for oak trees decline is challenging because there are numerous interacting factors that affect oak health (González Alonso 2008). Persian oak,Q. brantii var. persica, in the west of Iran have been in decline since 2000 (Beyranvand et al. 2016). Iranian forest managers believe that no single cause is responsible for the decline of oak trees (Attarod et al. 2016). Forest degradation, frequent droughts, overgrazing, suspended particles and dusts created from adjacent locations, coupled with changes in meteorological parameters in recent years, are assumed to reduce the ability of trees to deal with environmental stresses. One of the predominant hypotheses in decline of the oak trees in Iran is variations in meteorological parameters in recent years (Attarod et al. 2016). Research suggest that climatic variability and climate change may have a significant impact on hydrological parameters; namely runoff, evapotranspiration (ET), soil moisture, interception (I), and ground water (e.g., Remrova & Cislerova 2010; Wang et al. 2013; Murray 2014). The main objectives of our study were (i) to examine the long-term trends of the meteorological parameters and reference evapotranspiration (ET₀) in the Zagros region, and (ii) to estimate ecohydrological parameters of the individual Persian oak affected by climate change or climate variability. To our knowledge, no studies used meteorological variables to detect and monitor the oak trees decline in Iran.

MATERIALS AND METHODS

Study area

The research was conducted in the Zagros region of Iran. The region consists of the western and southern slopes of the Zagros Mountains range, which extends from the Turkish border to the Northwest to the Persian Gulf to the Southeast (Fig. 1).

In this region, the average annual precipitation (P) varies between 250 and 800 mm, while the mean annual air temperature (T) ranges from 9 ºC to 25 ºC (Fathizadeh et al. 2013).

Rainfall historically occurs during winter and spring. Summers are hot and dry (mean: 35 ºC) and winters are cool (mean: 7 ºC).

Meteorological data and evapotranspiration equation

Long-term meteorological data (1961-2010) from six synoptic meteorological stations (Fig. 2 and Table 1) were used to parameterize the De Martonne Aridity Index, I_DM (Baltas 2007).

We used the FAO Penman-Monteith (PM) combination equation to calculate daily ET₀. We also used PM method to ET₀ as the standard method proposed by the International Commission for Irrigation and Drainage (ICID) and Food and Agriculture Organization of the United Nation (FAO).

The PM equation for calculation of the daily ET₀ assumes the reference crop evapotranspiration. The hypothetical crop has an assumed height of 0.12 m, a surface resistance of 70 s m^-1 and an albedo of 0.23. It closely resembles the evaporation from a green grass of uniform height, actively growing and adequately watered (Allen et al. 1998):

(1)

whereET₀ (mm day⁻¹) is the reference ET; R_n (MJ m⁻² day⁻¹) is the net radiation at the crop surface; G (MJ m⁻² day⁻¹) is soil heat flux density; T(°C) is the mean daily air temperature at a height of 2 m; u₂ (m s⁻¹) is the wind speed at a height of 2 m; e_s (kPa) is saturation vapor pressure calculated as e_s= 0.6108 exp[(17.27 T ) / (T+237.3)]; e_a (kPa) is actual vapor pressure computed as e_a=(RH× e_s) ×100, where RH is daily relative humidity (%); e_s−e_a(kPa) is vapor pressure deficit (VPD); Δ (kPa °C⁻¹) is the slope of vapor pressure curve at the daily mean T; and γ (kPa °C⁻¹) is the psychometric constant calculated as 0.665× 10⁻³×AP; in which AP (kPa) is the atmospheric pressure. G at the daily scale beneath the grass reference surface was relatively small and thus could be ignored in PM combination equation. Therefore, G= 0 ((MJ m⁻² day⁻¹) (Goyal 2004; Maruyama et al. 2004). To calculate daily ET₀, we employed daily mean T, RH, WS at 2 m height and sunshine hours.

Fig.1. Location of the Zagros region, west of Iran

Fig. 2. Positions of the six synoptic meteorological stations in the Zagros region, west of Iran.

Trend test of the meteorological parameters and ET₀

The linear trends in meteorological parameters of Ta, P, WS, and ET₀ during the period 1961-

2010 in the Zagros region were evaluated by the Mann-Kendall (MK) test, which has been found to be an excellent tool for trend detection and extensively used to detect monotonic trends in series of meteorological data (e.g., Ludwig et al. 2004; Zhang et al. 2004; Patra et al. 2012). When the MK statistic (Z_MK) is greater than 1.96 (based on normal probability table), there is a significant increasing trend (a = 0.05), whereas negative Z_MK values lower than -1.96 show significant decreasing trends (a = 0.05).

Measurement of rainfall interception (I) and estimation of ecohydrological parameters Measurements of P and throughfall (TF) were conducted at Shurab Forest Park, Khoram-Abad City (Fig. 2). The field experiment was carried out over two months started at October, 26, 2013 and ended in December, 19, 2013. Five isolated, typical, and mature Q. brantii trees were selected with similar morphology features (i.e., tree height, h (7.0 m); diameter at breast height, DBH (30 cm); crown diameter, Cd (6 m²); crown-projected area, CPA (40 m²); and leaf area index, LAI, (2)). LAI was measured using a fish eye camera (Canon EOS 6D EF 8-15mm f / 4L) during the in-leaf season. The standard method of measuring the CPA is to project the edges of the crown to a horizontal surface (Levia 2004). CPA was measured with a clinometer and a tape.

Table 1. Characteristics of the synoptic meteorological stations located in the Zagros region, west of Iran. Climate classification was achieved according to the De Martonne Aridity Index (I_DM) after Baltas et al.(2007).

The crown radius was measured as the distance from the center of the tree bole to the edge of the crown. To obtain the best estimate of mean crown diameter, we used the average of 4 cardinal directions (N,E,S, and W) (Ahmadi et al. 2009). The canopies of these trees did not overlap with adjacent trees. P was collected using five rain-gauges (catch area, 64 cm² each), in a nearest neighboring open area. The average of the five rain-gauges was used to measure P. TF was measured on an event basis under the isolated crowns of five Q. brantii trees. For each tree, 16 TF-gauges (similar rain-gauges) were used, with four of these gauges placed along four transects (N, E, S, and W). We computed the total volume of TF captured underneath each tree for individual events using the total number of rain-gauges and the obtained average volume. SF was not measured, since it is usually a small fraction of each rain storm (e.g., Lloyd et al. 1988; Hutjes et al. 1990; Valente et al. 1997; Licata et al. 2011; Sun et al. 2013). Lastly, I was calculated as the difference between P and TF. A rain event was defined as a period of P bounded by 4 hours with no measurable rainfall as indicted by P rain-gauges (Fathizadeh et al. 2013). Four hours is long enough to allow the canopy to dry out completely in this semi-arid climate (I_DM = 18) (Carlyle-Moses et al. 2004). We measured the water collected in the gauges the day following each storm. P storms were grouped into three classes (Fathizadeh et al. 2013): P< 5 mm, 5 mm to P< 10 mm and equal or more than 10 mm, to allow for a better understanding of the relationship between P and I. We used indirect regression methods that relate TF and P to determine canopy ecohydrological parameters including canopy saturation point (P^'_G), canopy water storage capacity (S, mm), free throughfall coefficient (p), the ratio of mean evaporation rate from the wet canopy to the mean rainfall intensity (E/R). The amount of P that is necessary to saturate the canopy before the drip of TF occurs is defined as P^'_G, which is equal to S if p is zero, and can be estimated subjectively by finding the inflection point on a graph relating TF to P for multiple storms (Sadeghi et al. 2014, 2015a, b). In the present study, we used the mean method (Jackson 1975) for estimation of S. This method requires two regression lines relating P (x-axis) and TF (y-axis). The first regression line (R₁) fits to storms where P is ≥ P^'_G (R₁), while a second regression line fits to storms where P is less than P^'_G (R₂). The differences between P and TF at the intersection point of R₁ and R₂ offers the estimates of S. One minus slope of R₁ provides an estimate of E/R and the slope of R₂ arrange for an estimate of p.

RESULTS

Trends of meteorological parameters and ET₀

During 1961-2010, the mean annual T and P in the selected stations at the Zagros region were 16.7 ºC and 546 mm, respectively. The I_DM in the Zagros region ranged roughly from 8 (arid climate) in Ramhormoz to 33 (humid climate) in Yasuj. Mean I_DM across the meteorological stations was roughly 17 (semi-dry climate). Mean annual WS, and ET₀ were 2.0 m s^-1, and 4.2 mm day, respectively. Table 2 shows the ET₀ and WS trends in terms of Z_MK statistic within the long-term period (1961-2010). Three out of six stations exhibited statistically significant positive (2 sites increasing) and negative (1 site decreasing) trends (P < 0.05) for the annual T, whereas P and ET₀ indicated statistically negative trends only in one station (Z_MK = -0.4 for P; Z_MK = -0.2 for ET₀).

Table 2. MK statistic (Z_MK) obtained through the Mann-Kendall method for the meteorological parameters as well as reference evapotranspiration (ET₀) in the Zagros region of western Iran within the period 1961-2010.

Meteorological parameters and ET₀ changed in recent years (2000-2010) relative to the previous period (1961-1999) (Table 3). Average T at the five synoptic stations increased by an average of approximately 4%. P decreased an average of 10% at five of the six stations in the period 2000-2010 compared 1961-1999 whereas WS increased by an average of 22% at five stations and decreased by 5% in one station. As a result of changing meteorological parameters, mean ET₀ increased by an average of 6.5% at the five stations and decreased by 18% in one station.

Table 3. Changes in meteorological parameters, reference evapotranspiration in the Zagros region, west of Iran, in the period 2000 - 2010 compared to previous years (before 1999).

Interception and ecohydrological parameters of oak trees

During the study period, 15 rainfall events were recorded, and cumulative depth of these P storms was 137.6 mm with individual events averaging 9.2 mm and ranging from 1.1 to 25.3 mm. The average (± standard error) cumulative I for the five trees was 36.8 ± 0.2 mm (26.7% of P).

The mean annual event based ratio of I to P (relative I or I: P) value was 32.7% (SE: 3.8%). I: P ranged from 15.9%

of P for the larger P events (25.3 mm) to 77.1% of P (1.1 mm) for smallest P. I: P was strongly correlated with P. A strong correlation was also observed between relative interception (I:P)%) and P[(I:P)%= -0.16Ln(P)+0.65; R²= 0.64] (Fig. 3). The mean I:P values in classes of the P< 5 mm, 5 mm to P< 10 mm and equal or more than 10 mm were 47.5%, 30.3%, and 23.8%, respectively (Table 4). I for the 15 events was 32.7% (Table 5). During the measurement period, P^'_G, S, P, and E/R were 4.5 mm, 1.4 mm, 0.67, and 0.11, respectively.

Fig. 3. Regression analyses between relative interception (I:P) and precipitation (P) for individual oak trees (Quercus brantii var. persica). Error bars denote the standard error.

Table 4. Cumulative precipitation (P) depth, the percent of average relative interception (I:P)%, standard error (SE), and coefficient of variation (CV) divided into three P classes.

Table 5. A summary of the ecohydrological parameters [(I:P)%, the percent of relative I; P'_G, canopy saturation; S, canopy water storage capacity; p, free throughfall coefficient; and E/R, the ratio of mean evaporation rate from the wet canopy to the mean rainfall intensity] of individual Persian oak (Quercus brantii var. persica) trees.