Historical and future winter chill for temperate fruit and nut trees in Afghanistan

doi:10.21203/rs.3.rs-4687846/v1

Download PDF

Research Article

Historical and future winter chill for temperate fruit and nut trees in Afghanistan

https://doi.org/10.21203/rs.3.rs-4687846/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Temperate fruit trees require exposure to chill and heat conditions to overcome the dormant period in winter and resume growth in spring. Rising temperatures due to climate change have already impacted winter chill accumulation in horticulturally vital growing regions worldwide. Afghanistan’s intra-regional differences make the country favorable for producing a wide range of fruits and nuts, which constitute approximately 30% of the country's total export earnings. However, expected future temperature increases may be detrimental to the trees’ ability to fulfill their chilling requirements. To quantify this risk, we computed and mapped historical and future chill accumulation in Afghanistan. For this purpose, we used long-term daily temperature data from 1980 to 2020 for 51 hydrometeorological stations across the country. Based on the temperature data, we produced four future Shared Socioeconomic Pathway (SSP) scenarios (SSP126, SSP245, SSP370, and SSP585) for five General Circulation Models (GCMs) from the Coupled Model Intercomparison Project Phase 6 (CMIP6). We used the RMAWGEN weather generator to produce 100 synthetic realizations of weather records for historical (1980, 1990, 2000, 2010, and 2020) and future conditions (2035–2065 and 2070–2100). We quantified winter chill using the Dynamic Model. Additionally, we computed Safe Winter Chill (SWC), which is defined as the level of chill that is expected to be exceeded in 90% of years. We found that SWC has decreased in the low-lying regions mainly in the eastern, southern, and parts of the northern regions, and is projected to further decrease in the future. This reveals a possible future chill risk for high chill-requiring cultivars of different fruit and nut species, including apricot, peach, plum, pistachio and almond, in these regions. In contrast, we recorded an increase in SWC in the northern and central regions, marking these regions as potentially favorable for high-chill species in the future. Our findings support fruit and nut growers in Afghanistan in adapting their orchards to match species’ and cultivars' chill requirements with the expected future winter chill.

Fruit trees

Dormancy

Chill accumulation

Dynamic model

Spatial interpolation

Chill risk

Temperate fruit trees undergo a dormancy stage during their seasonal cycle to withstand cold winter temperatures (Rohde and Bhalerao 2007; Luedeling 2012; Fadón et al. 2020). This dormancy stage consists of two distinct phases known as endo-dormancy and eco-dormancy. Endo-dormancy is characterized by suspension of plant growth even under conducive environmental conditions. It is controlled by low temperatures that trigger the accumulation of winter chill, which eventually leads to the release of endo-dormancy. During the subsequent eco-dormancy, buds are receptive to warm temperature, accumulating heat (Lang et al. 1987; Campoy et al. 2011; Luedeling 2012; Fadón et al. 2020) which ultimately leads to resumption of growth in spring. The combined effect of cold and warm temperatures during the two phases influences the timing of spring phenological events in temperate trees (Luedeling 2012; Luedeling and Fernandez 2023). Consequently, changes in temperature seasonality have direct implications for fruit and net tree phenology.

Climate change has already affected the timing of phenological events in temperate fruit trees due to its direct influence on chill accumulation (Campoy et al. 2011; Luedeling et al. 2015). Besides shifts in flowering time, inadequate chill accumulation can result in incomplete dormancy release, manifesting in delayed bud burst, bud drop, and leafing and flowering asynchrony (Erez 2000; Campoy et al. 2011; Guo et al. 2023). Climate change may thus result in a loss of viability of commercial production of perennial fruits and nuts in some areas (Darbyshire et al. 2011; Benmoussa et al. 2018).

We explore the future viability of orchards in Afghanistan, a country that is expected to be strongly affected by climate change (Kreft et al. 2016; Eckstein et al. 2021). Afghanistan, being part of the Hindu Kush-Himalayan region, faced a significant decrease in cold events between 1961 and 2015 (Sun et al. 2017) and an increase of 1.8°C in mean temperature from 1951 to 2010 (Aich et al. 2017). Future projections using the General Circulation Models (GCMs) of the Coupled Model Intercomparison Project phase 6 (CMIP6) indicate substantial increases in minimum and maximum temperature of up to 2.5°C for the SSP585 scenario by 2020–2059 and 5°C by 2060–2099 (Sediqi et al. 2022).

Horticultural fruit production is a leading agricultural sub-sector in Afghanistan. Official estimates suggest that fresh and dried fruits contributed approximately 30% to the total exports earnings in 2023 (NSIA 2023). The country’s diverse geography and continental climate, with four distinct seasons, provides fertile ground for growing a wide range of deciduous and subtropical fruits. However, in a warming future, increasingly mild winters could be a serious concern in Afghanistan, since available chill might drop to levels that are insufficient for many temperate fruit and nut tree species. A robust estimation of recent and future chill in important growing regions is therefore vital for the management of orchards.

A number of models have been developed for quantifying winter chill for specific regions and estimating the chill requirements of different species and cultivars. The available literature shows that there is no single universally accepted model for computing winter chill for all fruit species. However, the Dynamic Model (Fishman et al. 1987b; Fishman et al. 1987a; Erez et al. 1990) has been found to be superior in accuracy, while delivering more moderate chill change estimations compared with other chill models (Luedeling et al. 2009c; Luedeling and Brown 2011; Luedeling 2012; Fadón et al. 2020). Despite the availability of modeling techniques that allow computing fairly accurate estimates of chill and heat requirements of different species and cultivars (Luedeling and Brown 2011; Luedeling et al. 2013), and despite the urgent need for robust projections, a region-based chill analysis for horticultural perennials in Afghanistan is noticeably absent.

Here, we aim to provide scientifically rigorous information on chill mapping in Afghanistan using refined methods to estimate state-of-the-art historical and future Safe Winter Chill (SWC) projections. This will guide fruit growers in relating problems that occur in fruit trees to the possibility of chill shortfalls, and it will guide policymakers and growers in planning large-scale commercial orchards.

Study area

This study aims at quantifying Safe Winter Chill (SWC) for the entire country of Afghanistan. The country experiences an arid and semi-arid continental climate characterized by hot summers and cold winters. Its climatic and geographic diversity has enabled the cultivation of a wide variety of subtropical and temperate fruit crops. For many of these crops, sufficient winter chill is critical. We used long-term daily temperature data from 51 weather stations that span two climatic zones (subtropical and temperate) and cover an elevation range from 372 to 2,939 meters above mean sea level (m a.s.l.) (Fig. 1). The criteria for selecting these stations among the 144 existing ones are 1) availability of credible and long (2008 to 2020) temperature records, and 2) availability of more than 90% of records for the given period.

Temperature data

Historical data

Records of daily minimum and maximum temperature for 51 stations covering the period from 2008 to 2020 were acquired from the Directorate of Hydrometeorology of the Ministry of Energy and Water (MEW) of Afghanistan. The automatic hydrometeorological stations record the current temperature every 15 minutes. In three stations, data gaps were observed. We used the ‘patch_daily_temperatures’ function from the chillR package (Luedeling et al. 2023) to fill a gap of 11 days in Eshkashim with information from the nearby station Ishkashim in Tajikistan (1.7 km distance), a gap of 1023 days in Dakah with data from the station Peshawar in Pakistan (56.2 km away), and a gap of 501 days in Nazdik-i-Kandahar with data from the stations FOB Pasab/Wilson (13.3 km) and Kandahar (28.8 km). The choice of stations was based on the distance to the target station, as well as similarity in elevation and temperature variation. Auxiliary records were adjusted for between-station temperature biases before being used to fill gaps.

The timespan of the recorded temperature data was not sufficient for analyzing long-term trends in Safe Winter Chill in Afghanistan. Therefore, we used simulated temperature data from 1980 to 2007 to compile historical temperature data for 40 years. The simulated temperature records were downloaded for each location from ClimateEngine (Huntington et al. 2017).

The acquisition of simulated data from ClimateEngine, which is based on several GCMs, introduces inherent biases. Therefore, bias correction was applied to enhance the quality of simulated data. We tested different methods of bias correction including Quantile Mapping Linear (QML; Boé et al. 2007; Gudmundsson et al. 2012), Quantile Mapping Tricub (QMT; Gudmundsson et al. 2012), and Quantile Delta Mapping (QDM; Cannon et al. 2015). Overall, all three methods demonstrated similar performance, however, visualization of mapping results indicated that results of QDM were slightly better than those of the other two methods (Fig. S1 and S2 in Supplementary Material). Earlier studies have thoroughly investigated the performance of various bias correction methods over time (Pierce et al. 2015; Maraun 2016; Kim et al. 2021; Dinh and Aires 2023). The Quantile Delta Method (QDM) for bias correction, which is known to explicitly preserve relative changes in the quantiles of distributions (Cannon et al. 2015), performed best in our study. We therefore applied this approach to the temperature data using the ‘QDM’ function from the MBC package (Cannon et al. 2015) in the R programming language. The results were analyzed using the Root Mean Square Error of Prediction (RMSEP) and the Ratio of Performance to Inter-Quartile distance (RPIQ) before and after bias correction (Fig. S3 and S4 Supplementary Material). Details about the bias correction of the calibration period for daily minimum and maximum temperature for all stations are available online in a public GitHub repository.

(https://github.com/Atifshinwari/Chill_quantification_in_Afghanistan)

Winter chill projection

Historical scenarios

For the computation of past chill accumulation, historical weather scenarios were produced based on the extended weather records (1980 to 2020). For this purpose, we used the RMAWGEN weather generator (Cordano and Eccel 2016) to produce 100 random realizations of synthetic weather records for 1980, 1990, 2000, 2010 and 2020 for all stations, based on the available temperature records. This process adhered to the procedures explained and implemented in previous studies (Benmoussa et al. 2020; Buerkert et al. 2020; Fernandez et al. 2021). Weather generators are statistical models that use random processes to simulate meteorological variables such as temperature, precipitation, humidity, and wind speed. The generation of synthetic weather is based on a characterization of typical historical weather distributions, which are crucial for capturing local agro-climatic conditions. A weather generator considers the local climate as the virtual collection of all plausible weathers, with annual weather being a random draw from this collection. A sample consisting of multiple random draws that represent possible manifestations of the weather of a given year enables the assessment of climatic risks. Such a sample can be instrumental in supporting agronomic decisions, e.g. the selection of tree cultivars in orchard planning. Historical scenarios were generated for 1980, 1990, 2000, 2010, and 2020 based on the historical temperature records from 1980 to 2020 for all the stations. Each scenario consisted of 100 replicates of plausible weather for the given scenario year for our location of interest.

Future scenarios

For future chill projections, we used scenarios of the Coupled Model Intercomparison Project Phase 6 (CMIP6), namely the Shared Socioeconomic Pathway (SSP) scenarios SSP126, SSP245, SSP375 and SSP585 (downloaded from the Earth System Grid Federation, ESGF). The SSP scenarios were downloaded for 5 GCMs (‘BCC-CSM2-MR’, ‘GFDL-ESM4’, ‘INM-CM4-8’, ‘INM-CM5-0’, and ‘MRI-ESM2-0’). We obtained data for daily minimum and maximum temperatures with a resolution of 100 km. To facilitate the interpretation of the future scenarios, we considered the original intent of the SSP scenarios, interpreting the SSP126 scenario as ‘optimistic’, SSP245 as ‘intermediate’, SSP375 as ‘moderately pessimistic’, and SSP585 as ‘most pessimistic’.

Based on these data, we generated future climate scenarios with an approach similar to the generation of historical scenarios. The weather generator was calibrated with the site-specific observed weather data from 1980 to 2020. Future scenarios were then imposed as specified by the CMIP scenarios, and 100 random realizations of future weather records were simulated. To encompass both near-term and long-term projections, we focused on two distinct periods: 2035–2065, represented by its central year 2050, and 2070–2100, represented by its central year 2085.

The Dynamic Model for winter chill quantification

For all the weather records and historical and future scenarios, winter chill was estimated in R using the chillR package (Luedeling et al. 2023). For chill quantification, we used the Dynamic Model with the original parameters that were used in the original version of the model (Fishman et al. 1987b; Fishman et al. 1987a; Erez et al. 1990). The Dynamic Model is widely used around the world with the same parameter set and it has shown consistent and reliable performance across different locations, in contrast to other models (Balandier et al. 1993; Luedeling and Brown 2011; Darbyshire et al. 2011; Campoy et al. 2011; Guo et al. 2015). Erez et al. (1990) discussed the effectiveness of this model in providing accurate assessments of winter chill for both cold and warm regions. Luedeling & Brown (2011) also found this model to be superior to other chill models for fruits and nuts, especially in studies covering wide temperature gradients. Therefore, considering the intra-regional differences in Afghanistan, we relied on the Dynamic Model for our analysis.

Quantification of Safe Winter Chill

For compatibility with the Dynamic Model, we converted the daily temperature data to an hourly format following the procedure described by Luedeling (2018), which is based on daily temperature extremes and the latitude of the stations. Thereafter, we applied the concept of Safe Winter Chill (SWC), which represents the minimum chill level that is likely to be met in 90% of years and is calculated as the 10th percentile of the chill distribution (Luedeling et al. 2009b).

To enhance the interpretability of the SWC analysis, we subdivided the country into the sub-regions ‘North’, ‘Northeast’, ‘East’, ‘South’, ‘West’, and ‘Central’ and assigned the weather stations according to their locations.

To capture the emerging climate signals, we evaluated the degree of overlap between the 100 synthetic years of historical and future scenarios. Additionally, we computed the absolute differences in chill accumulation between different time points of historical and future scenarios. To better examine the subsequent changes in chill accumulation, we compared the simulated historical scenarios for 1980 and 2020. Similarly, each simulated future scenario was compared with the median of all historical scenarios.

Countrywide chill mapping using spatial interpolation

We used Kriging as a spatial interpolation technique to estimate chill accumulation across the country. Similar approaches have been applied for chill mapping in California (Luedeling et al. 2009b), Tunisia (Benmoussa et al. 2020), South America (Fernandez et al. 2021) and the Mediterranean region (Fernandez et al. 2023).

To enhance the quality of the interpolated surface, we adopted the mean daily temperature in December as an auxiliary variable. Gridded data representing mean daily temperature in December between 1970 to 2000 were downloaded from the WorldClim database (Fick and Hijmans 2017) with a spatial resolution of 30 seconds (~ 1 km²). Based on a temperature variogram, which was fitted using R’s automap package (Hiemstra et al. 2009), we interpolated temperature data across the grid. Similarly, a variogram was fitted for each chill scenarios and Kriging was performed to produce an interpolated chill layer. Kriging was implemented using the gstat package in R (Gräler et al. 2016).

Following these two steps, we fitted a linear regression model to capture the relationship between temperature and chill accumulation for each scenario. We estimated typical chill values using the interpolated temperature and the linear model. We then computed the difference between typical chill and interpolated chill layer to capture the temperature effect and include the chill contribution associated with the temperature that remained unaccounted for in the initial Kriging of station-specific chill values. The computed difference in chill was added to the interpolated chill layer to derive the final chill layer. We applied this process to all scenarios and plotted the final maps using the tmap package in R (Tennekes 2018).

Data analysis and visualization

The R statistical software (R Development Core Team, 2022) and in particular the chillR package (Luedeling et al. 2023) were used for implementing all steps involved in this study.

We discovered a considerable change in chill accumulation across Afghanistan over the historical and future periods. The general pattern showed a development towards more extreme environments over time, with the observed change in chill accumulation ranging from a notable decrease in the lowlands to an increase in the highlands. Change trends in accumulated chill showed strong variation across regions and weather stations, varying with temperature and elevation. This variation led to the emergence of considerable regional differences.

Historical Safe Winter Chill

In the eastern lowlands and southern regions, we recorded a general decrease in chill accumulation (Fig. 2). We illustrated this trend by evaluating the degree to which the 100 simulated chill accumulation totals for the 1980 historical scenario overlap with those for the 2020 scenario. The strongest decrease in chill accumulation (-14 CP) was recorded for Nawabad (NBD), a horticulturally important station in the lowlands of the eastern region (Fig. 3). Between 1980 and 2020, SWC at this station decreased from 53 CP to 39 CP. The degree of overlap between the two scenario distributions indicated that 42% of the simulated chill accumulation totals for 2020 were outside the range of simulated values for 1980 (Fig. 3).

We observed an overall increase in SWC accumulation across the historical scenarios for the northern, northeastern, western, and central regions (Fig. 2). The strongest increase was observed at Cheghcheran (CCN), a station in the central highland region (+ 10 CP), with absolute values ranging from 51 CP in 1980 to 61 CP in 2020 (Fig. 3). Sixteen percent of the 2020 values surpassed the maximum of the chill accumulation range observed for 1980 (Fig. 3). The northern region, apart from exhibiting a general increase in accumulation, showed minor chill losses, especially in the lowland areas bordering Uzbekistan.

Considering the extremes across all stations in 2020, we recorded the lowest SWC, 27 CP, at Gardandiwal (GDL), a high-altitude station located in the central highlands at an elevation of 2,939 m a.s.l.. The highest SWC registered was 79 CP in the high-mountain station of Faizabad (FZD), which is located at an elevation of 1180 m a.s.l. in the northeastern region. In the lowlands, the station Dakah (DKH), located at an elevation of 430 m a.s.l. in the eastern plains, recorded the lowest SWC of 35 CP (Fig. 3).

Future Safe Winter Chill

The lowland regions, particularly in the eastern and southern parts of the country, exhibited high chill losses in future SSP scenarios for both time points (Fig. 4). Nawabad (NBD), a valley station in the eastern region, recorded the highest chill losses across all future scenarios compared to historical SWC (Fig. 5). The values of SWC in Nawabad (NBD) decreased by 6 CP to an absolute SWC of 43 CP under the ‘optimistic’ scenario by 2050. SWC accumulation was further decreased by 10 CP to 33 CP for the ‘moderately pessimistic’ scenario (SSP375) by 2050 and by another 16 CP to 17 CP under the ‘most pessimistic’ scenario (SSP585) by 2085. Except for the central highlands, many stations in the north, northeast, and western regions also experienced chill losses, especially in the low-lying regions (Fig. 4).

The emerging climate signals, represented by the degree of overlap between the simulated chill estimates for historical and future scenarios, were mainly apparent in low to midland regions. The degree of overlap for Adraskan (ADN), a station in the western region, showed that for the ‘optimistic’ scenario (SSP126), 19% of the simulated chill accumulation range for 2050 fell below the minimum chill accumulation estimated for the historical scenarios. Projected changes were even more dramatic for Asmar (ASR), a station in the eastern lowlands, where 70% of the simulated chill accumulation under the ‘intermediate’ scenario by 2050 and 100% of the values under the ‘moderately’ and ‘most pessimistic’ scenarios by 2085 were below the minimum historical chill baseline.

In the highland regions, many stations retained their SWC levels, with the central highlands even exhibiting a notable increase in chill accumulation across all future scenarios (Fig. 4). For the high-altitude station of Gardandiwal (GDL), relative to historical scenarios, SWC increased by 9 CP to an absolute value of 34 CP for the ‘optimistic’ scenario by 2050. SWC was further increased by 13 CP, reaching the highest absolute value of 47 CP for the ‘most pessimistic’ scenario in 2085 (Fig. 5). Considering the degree of overlap between historical and future simulated chill distributions in the highlands, for Keraman (KRN) in the central region, a station with a relatively cool climate, under the ‘moderately’ pessimistic future scenario, 30% of chill accumulation in 2050 surpassed the maximum of the historical chill accumulation range. Similarly, for Anjuman (AJN) in the high-mountain area of the northeastern region, under the ‘moderately’ and ‘most pessimistic’ scenarios, 74% and 59% of values in 2085 were above the maximum historical chill accumulation range.

We observed considerable differences across the 5 GCMs in estimating winter chill for stations in the warm and cold regions. The climate model BCC-CSM2-MR projected the overall lowest chill levels in warm regions for different SSP scenarios and time points. However, the same model produced high chill estimations for cold regions (Fig. 5). Contrary to this, the GFDL-ESM4 GCM generally projected the highest chill in warm regions but indicated low chill accumulation in cold regions. The GCMs INM-CM5-0 and INM-CM4-8 exhibited a high degree of similarity in chill estimation in both warm and cold regions.

Across all simulations, the lowest chill accumulation was projected for Dakah (DKH) in the eastern plains, where an SWC of 30 CP was estimated for the optimistic scenario (SSP126) for 2050, with a further decrease to 10 CP by 2085 under pessimistic assumptions. On the other end of the spectrum, the highland station of Estalef (ETF) in the central region recorded the highest SWC of 74 CP under the optimistic scenario (SSP126) in 2050 and 78 CP under the intermediate scenario (SSP245) in 2085. For the moderately and most pessimistic scenarios (SSP370, SSP585), Keraman (KRN) in the same central region recorded the highest SWC at 77 CP and 75 CP, respectively, in 2085.

Our study delivers the first complete chill profile of Afghanistan. By combining a historical assessment with future projections, the study offers useful insights into the impact of climate change on chill accumulation across the whole country and a robust estimation of its suitability for horticultural fruit and nut species.

We found considerable winter chill losses for both historical and future scenarios. The overall highest SWC losses amounted to up to 20 CP between 1980 and 2020 (Fig. 2) and continued to increase, reaching 70 CP under the most pessimistic scenario in 2085 (Fig. 5). Such losses were mainly observed in regions with considerable horticultural importance for perennial fruits. Growers in the eastern region cultivate apricots, peaches, citrus, and persimmon, while the southern region produces pomegranates, grapes, almonds, pistachios and figs. The chill losses we projected and the reduction in the degree of chill accumulation overlap between historical and future scenarios illustrate the emerging signals of climate change in the country. This was mainly apparent for Asmar (ASR), a station in the eastern lowlands, where 40% of chill accumulation in 2050 and 100% in 2085 are projected to fall outside the chill accumulation range recorded for 2020, even under the moderately pessimistic scenario. These findings demonstrate alarming prospects for temperate fruits and nuts in Afghanistan. It is important to note that similar chill losses have been observed in different regions across the world, such as Oman (Luedeling et al. 2009a), California (Luedeling et al. 2009b), Australia (Darbyshire et al. 2011), Tunisia (Benmoussa et al. 2020), and the Mediterranean region (Fernandez et al. 2023). The findings further imply that chill requirements play a vital role in selecting and introducing fruit cultivars for a particular region. This is mainly important for fruit trees that are native to cold regions but grown in warm regions (Erez 2000). Hence the fulfillment of chill and heat needs is a crucial consideration for fruit growers in Afghanistan in a warming future.

At high altitudes, the emerging signals of changes in future climate are evident in the form of chill gains, which range from a low or moderate SWC range (25 to 40 CP) to a considerably higher range (45 to 55 CP) for future climate scenarios, for example in stations in the central (Gardandiwal (GDL), 2,939 m a.s.l.; Waras (WRS), 2,498 m a.s.l.) and northeastern regions (Eshkashem (EKM), 2,502 m a.s.l.). While this might sound counter-intuitive at first sight, a closer look at the underlying mechanisms driving chill accumulation explains this pattern. The Dynamic Model does not attribute chilling effectiveness to freezing temperatures, with its effective temperature range limited to the interval between − 2°C and 14°C and maximum efficiency attained at between 6°C and 8°C (Erez et al. 1990). Hence the historical chill levels in these locations were low due to harsh cold but future warming will drive temperature closer to the effective range, facilitating gains in chill accumulation. Studies conducted in other regions have also projected future chill gains or maintenance of historical chill levels at high altitudes (Luedeling et al. 2011; Delgado et al. 2021; Fernandez et al. 2021). These findings offer potential direction for investigating the adaptation and selection of a different set of fruit cultivars in the long run. Previous studies have suggested that different fruit cultivars within the same species have varying chill and heat requirements based on their climatic origin (Luedeling et al. 2013; Delgado et al. 2021). Therefore, it might be possible to introduce and cultivate high-chill species such as apples and cherries in Afghanistan’s highlands, assuming that growers can also meet other needs, including water availability, soil suitability, and appropriate agronomic practices. Nevertheless, this needs a cautious approach and is only advisable once the species and cultivar-specific chill requirements have been determined.

In the near term, the most effective adaptation strategy is probably to quantify cultivar-specific chill requirements to enable selection of appropriate varieties. Numerous studies have been conducted worldwide for the quantification of cultivar-specific chill needs using experimental (Ruiz et al. 2007; Ruiz et al. 2018) and statistical (Luedeling et al. 2013; Guo et al. 2015; Benmoussa et al. 2018) methods. For the statistical approach, Partial Least Squares (PLS) regression was developed and applied by Luedeling & Gassner (2012) as a statistical opportunity to determine the climatic requirements of temperate fruits. A more recent development in this context is the PhenoFlex model, which combines the Dynamic Model and a heat model to simulate tree dormancy and predict spring phenology (Luedeling et al. 2021). However, this method requires availability of long-term phenology records and temperature data. Therefore, we strongly propose that the Agricultural Research Institute of Afghanistan (ARIA) within the Ministry of Agriculture, Irrigation and Livestock (MAIL) pay particular attention to consistent recording of flowering dates, thereby fostering long-term datasets essential for climate-informed decisions in the future.

The performance of chill proxies for spatial interpolation has varied between regions (Benmoussa et al. 2020; Buerkert et al. 2020; Fernandez et al. 2021) indicating that the choice of proxy data should be informed by considering the climatic variability and intra-regional variation in the study region. In our study, elevation was not an ideal proxy because of substantial differences in the climatic conditions between the sub-regions. Unlike the eastern and southern lowlands, stations in the northern lowlands (elevation ranges from 372 m to 556 m a.s.l.) experience colder climates and high chill accumulation. In addition, co-variables of minimum and maximum temperature produced unreliable results due to a difference of more than 2°C between the observed temperature data and the WorldClim database (Fig. S5 in Supplementary Material). Nevertheless, by leveraging the existing limited observed temperature data records and applying Kriging interpolation, using mean temperature in December as an auxiliary variable, we obtained valuable insights into the chill profile of Afghanistan, a region confronted with the challenge of data scarcity.

For most tree cultivars grown in Afghanistan, information on specific chill requirements is unavailable, making the selection of appropriate germplasm for future-resilient orchards difficult. Considering the chill losses we projected, especially for the southern and eastern regions, such information would greatly facilitate adaptation efforts. At present, however, such science-based adaption is held back by a lack of sufficiently long time series of phenology data, as well as by limited accessibility of the few data series that do exist. Our projections of future changes in chill accumulation can support planning of orchard designs or adaptation policies based on the specific chill profiles of Afghanistan’s regions. While we did not project complete loss of available chill for any growing regions, the projected changes indicate a need to prepare for future warming by considering location-specific chill profiles when selecting tree cultivars. For instance, regions that have experienced chill losses, such as the eastern and southern plains, are likely to require low-chill cultivars of the same fruit species or even the introduction of new species to these regions. On the other hand, regions where chill has increased, mainly the central and northern regions, may present opportunities for cultivating valuable high-chill cultivars. In the near term, strategic relocation of cultivars within regions may help avert immediate chill-related problems. However, such relocation should not be narrowly focused on ensuring that cultivars receive sufficient chill; it would also need to account for further site characteristics, such as additional agroclimatic considerations (e.g. excessive heat, hail risk, rainfall patterns), water availability and access to input and output markets. Through a thoughtful and holistic approach to managing cultivar selection, the future chill changes we project can likely be transformed from liabilities to assets, with new opportunities for designing productive and resilient orchards.

Acknowledgments We acknowledge the support of the Surface Water Resources Department of the Ministry of Energy and Water (MEW) of Afghanistan for providing the long-term daily temperature data.

Funding The research project is funded by the German Academic Exchange Service (DAAD) under the funding programme of ‘Hilde Domin Programme Afghanistan’ as part of the PhD scholarship (2022/2023) for the first author.

Competing Interests The authors declare that they have no conflict of interest, neither financial nor non-financial, related to this research work.

Author Contribution AS developed the research proposal under the supervision and guidance of EL. AS conducted all the analysis, R code implementation, results visualization and manuscript writing. LC provided support on R code implementation and validation and reviewed the manuscript. KS contributed to the analysis in various ways, supporting visualizations and providing guidance on the manuscript. EL supervised the overall research work, provided regular guidance on major aspects of the research and thoroughly reviewed and edited the manuscript.

Data Availability All the scripts related to the data processing, analysis and figure generation of this study are available online in a public GitHub repository.

(https://github.com/Atifshinwari/Chill_quantification_in_Afghanistan)

Aich V, Akhundzadah N, Knuerr A, Khoshbeen A, Hattermann F, Paeth H, Scanlon A, Paton E (2017) Climate Change in Afghanistan Deduced from Reanalysis and Coordinated Regional Climate Downscaling Experiment (CORDEX)—South Asia Simulations. Climate 5:38. https://doi.org/10.3390/cli5020038
Balandier P, Bonhomme M, Rageau R, Capitan F, Parisot E (1993) Leaf bud endodormancy release in peach trees: evaluation of temperature models in temperate and tropical climates. Agric For Meteorol 67:95–113. https://doi.org/10.1016/0168-1923(93)90052-J
Benmoussa H, Ben Mimoun M, Ghrab M, Luedeling E (2018) Climate change threatens central Tunisian nut orchards. Int J Biometeorol 62:2245–2255. https://doi.org/10.1007/s00484-018-1628-x
Benmoussa H, Luedeling E, Ghrab M, Ben Mimoun M (2020) Severe winter chill decline impacts Tunisian fruit and nut orchards. Clim Change 162:1249–1267. https://doi.org/10.1007/s10584-020-02774-7
Boé J, Terray L, Habets F, Martin E (2007) Statistical and dynamical downscaling of the Seine basin climate for hydro‐meteorological studies. Int J Climatol 27:1643–1655. https://doi.org/10.1002/joc.1602
Buerkert A, Fernandez E, Tietjen B, Luedeling E (2020) Revisiting climate change effects on winter chill in mountain oases of northern Oman. Clim Change 162:1399–1417. https://doi.org/10.1007/s10584-020-02862-8
Campoy JA, Ruiz D, Egea J (2011) Dormancy in temperate fruit trees in a global warming context: A review. Sci Hortic 130:357–372. https://doi.org/10.1016/j.scienta.2011.07.011
Cannon AJ, Sobie SR, Murdock TQ (2015) Bias Correction of GCM Precipitation by Quantile Mapping: How Well Do Methods Preserve Changes in Quantiles and Extremes? J Clim 28:6938–6959. https://doi.org/10.1175/JCLI-D-14-00754.1
Cordano E, Eccel E (2016) Tools for stochastic weather series generation in R environment. Ital J Agrometeorol 21. https://doi.org/10.19199/2016.3.2038-5625.031
Darbyshire R, Webb L, Goodwin I, Barlow S (2011) Winter chilling trends for deciduous fruit trees in Australia. Agric For Meteorol 151:1074–1085. https://doi.org/10.1016/j.agrformet.2011.03.010
Delgado A, Dapena E, Fernandez E, Luedeling E (2021) Climatic requirements during dormancy in apple trees from northwestern Spain – Global warming may threaten the cultivation of high-chill cultivars. Eur J Agron 130:126374. https://doi.org/10.1016/j.eja.2021.126374
Dinh TLA, Aires F (2023) Revisiting the bias correction of climate models for impact studies. Clim Change 176:140. https://doi.org/10.1007/s10584-023-03597-y
Eckstein D, Künzel V, Schäfer L (2021) Global Climate Risk Index 2021 | Germanwatch e.V. https://www.germanwatch.org/en/19777. Accessed 31 Mar 2024
Erez A (2000) Bud dormancy; phenomenon, problems and solutions in the tropics and subtropics. In: Erez A (ed) Temperate Fruit Crops in Warm Climates. Springer Netherlands, Dordrecht, pp 17–48
Erez A, Fishman S, Linsley-Noakes GC, Allan P (1990) The dynamic model for rest completion in peach buds. Acta Hortic 165–174. https://doi.org/10.17660/ActaHortic.1990.276.18
Fadón E, Fernandez E, Behn H, Luedeling E (2020) A Conceptual Framework for Winter Dormancy in Deciduous Trees. Agronomy 10:241. https://doi.org/10.3390/agronomy10020241
Fernandez E, Caspersen L, Illert I, Luedeling E (2021) Warm winters challenge the cultivation of temperate species in South America—a spatial analysis of chill accumulation. Clim Change 169:28. https://doi.org/10.1007/s10584-021-03276-w
Fernandez E, Mojahid H, Fadón E, Rodrigo J, Ruiz D, Egea JA, Ben Mimoun M, Kodad O, El Yaacoubi A, Ghrab M, Egea J, Benmoussa H, Borgini N, Elloumi O, Luedeling E (2023) Climate change impacts on winter chill in Mediterranean temperate fruit orchards. Reg Environ Change 23:7. https://doi.org/10.1007/s10113-022-02006-x
Fick SE, Hijmans RJ (2017) WorldClim 2: new 1-km spatial resolution climate surfaces for global land areas. Int J Climatol 37:4302–4315. https://doi.org/10.1002/joc.5086
Fishman S, Erez A, Couvillon GA (1987a) The temperature dependence of dormancy breaking in plants: Computer simulation of processes studied under controlled temperatures. J Theor Biol 126:309–321. https://doi.org/10.1016/S0022-5193(87)80237-0
Fishman S, Erez A, Couvillon GA (1987b) The temperature dependence of dormancy breaking in plants: Mathematical analysis of a two-step model involving a cooperative transition. J Theor Biol 124:473–483. https://doi.org/10.1016/S0022-5193(87)80221-7
Gräler B, Pebesma E, Heuvelink G (2016) Spatio-Temporal Interpolation using gstat. R J 8:204–218
Gudmundsson L, Bremnes JB, Haugen JE, Engen-Skaugen T (2012) Technical Note: Downscaling RCM precipitation to the station scale using statistical transformations – a comparison of methods. Hydrol Earth Syst Sci 16:3383–3390. https://doi.org/10.5194/hess-16-3383-2012
Guo L, Liu X, Alatalo JM, Wang C, Xu J, Yu H, Chen J, Yu Q, Peng C, Dai J, Luedeling E (2023) Climatic drivers and ecological implications of variation in the time interval between leaf-out and flowering. Curr Biol 33:3338-3349.e3. https://doi.org/10.1016/j.cub.2023.06.064
Guo L, Xu J, Dai J, Cheng J, Luedeling E (2015) Statistical identification of chilling and heat requirements for apricot flower buds in Beijing, China. Sci Hortic 195:138–144. https://doi.org/10.1016/j.scienta.2015.09.006
Hiemstra PH, Pebesma EJ, Twenhöfel CJW, Heuvelink GBM (2009) Real-time automatic interpolation of ambient gamma dose rates from the Dutch radioactivity monitoring network. Comput Geosci 35:1711–1721. https://doi.org/10.1016/j.cageo.2008.10.011
Huntington JL, Hegewisch KC, Daudert B, Morton CG, Abatzoglou JT, McEvoy DJ, Erickson T (2017) Climate Engine: Cloud Computing and Visualization of Climate and Remote Sensing Data for Advanced Natural Resource Monitoring and Process Understanding. Bull Am Meteorol Soc 98:2397–2410. https://doi.org/10.1175/BAMS-D-15-00324.1
Kim S, Joo K, Kim H, Shin J-Y, Heo J-H (2021) Regional quantile delta mapping method using regional frequency analysis for regional climate model precipitation. J Hydrol 596:125685. https://doi.org/10.1016/j.jhydrol.2020.125685
Kreft S, Eckstein D, Melchior I (2016) Global Climate Risk Index 2017 Who Suffers Most From Extreme Weather Events? Weather-related Loss Events in 2015 and 1996 to 2015. Germanwatch Nord-Süd Initiative e.V, Bonn
Lang GA, Early JD, Martin GC, Darnell RL (1987) Endo-, Para-, and Ecodormancy: Physiological Term inology and Classification for Dorm ancy Research. 22
Luedeling E (2012) Climate change impacts on winter chill for temperate fruit and nut production: A review. Sci Hortic 144:218–229. https://doi.org/10.1016/j.scienta.2012.07.011
Luedeling E (2018) Interpolating hourly temperatures for computing agroclimatic metrics. Int J Biometeorol 62:1799–1807. https://doi.org/10.1007/s00484-018-1582-7
Luedeling E, Blanke M, Gebauer J (2015) Chilling challenges in a warming world. Acta Hortic 901–907. https://doi.org/10.17660/ActaHortic.2015.1099.115
Luedeling E, Brown PH (2011) A global analysis of the comparability of winter chill models for fruit and nut trees. Int J Biometeorol 55:411–421. https://doi.org/10.1007/s00484-010-0352-y
Luedeling E, Caspersen L, Fernandez E (2023) chillR: Statistical Methods for Phenology Analysis in Temperate Fruit Trees
Luedeling E, Fernandez E (2023) Forecasting tree phenology in a climate change context. Acta Hortic 69–78. https://doi.org/10.17660/ActaHortic.2023.1372.10
Luedeling E, Gassner A (2012) Partial Least Squares Regression for analyzing walnut phenology in California. Agric For Meteorol 158–159:43–52. https://doi.org/10.1016/j.agrformet.2011.10.020
Luedeling E, Gebauer J, Buerkert A (2009a) Climate change effects on winter chill for tree crops with chilling requirements on the Arabian Peninsula. Clim Change 96:219–237. https://doi.org/10.1007/s10584-009-9581-7
Luedeling E, Girvetz EH, Semenov MA, Brown PH (2011) Climate Change Affects Winter Chill for Temperate Fruit and Nut Trees. PLoS ONE 6:e20155. https://doi.org/10.1371/journal.pone.0020155
Luedeling E, Kunz A, Blanke MM (2013) Identification of chilling and heat requirements of cherry trees—a statistical approach. Int J Biometeorol 57:679–689. https://doi.org/10.1007/s00484-012-0594-y
Luedeling E, Schiffers K, Fohrmann T, Urbach C (2021) PhenoFlex - an integrated model to predict spring phenology in temperate fruit trees. Agric For Meteorol 307:108491. https://doi.org/10.1016/j.agrformet.2021.108491
Luedeling E, Zhang M, Girvetz EH (2009b) Climatic changes lead to declining winter chill for fruit and nut trees in California during 1950-2099. PLoS ONE 4:e6166. https://doi.org/10.1371/journal.pone.0006166
Luedeling E, Zhang M, Luedeling V, Girvetz EH (2009c) Sensitivity of winter chill models for fruit and nut trees to climatic changes expected in California’s Central Valley. Agric Ecosyst Environ 133:23–31. https://doi.org/10.1016/j.agee.2009.04.016
Maraun D (2016) Bias Correcting Climate Change Simulations - a Critical Review. Curr Clim Change Rep 2:211–220. https://doi.org/10.1007/s40641-016-0050-x
NSIA (2023) Statistical Yearbook 2022-23. In: Google Docs. https://drive.google.com/file/d/1qMpj8_b56_in43QUbzx94jstwAeswdfj/view?usp=sharing&usp=embed_facebook. Accessed 31 Mar 2024
Pierce DW, Cayan DR, Maurer EP, Abatzoglou JT, Hegewisch KC (2015) Improved Bias Correction Techniques for Hydrological Simulations of Climate Change*. J Hydrometeorol 16:2421–2442. https://doi.org/10.1175/JHM-D-14-0236.1
Rohde A, Bhalerao RP (2007) Plant dormancy in the perennial context. Trends Plant Sci 12:217–223. https://doi.org/10.1016/j.tplants.2007.03.012
Ruiz D, Campoy J, Egea J (2007) Chilling and heat requirements of apricot cultivars for flowering. Environ Exp Bot 61:254–263. https://doi.org/10.1016/j.envexpbot.2007.06.008
Ruiz D, Egea J, Salazar JA, Campoy JA (2018) Chilling and heat requirements of Japanese plum cultivars for flowering. Sci Hortic 242:164–169. https://doi.org/10.1016/j.scienta.2018.07.014
Sediqi MN, Hendrawan VSA, Komori D (2022) Climate projections over different climatic regions of Afghanistan under shared socioeconomic scenarios. Theor Appl Climatol 149:511–524. https://doi.org/10.1007/s00704-022-04063-y
Sun X-B, Ren G-Y, Shrestha AB, Ren Y-Y, You Q-L, Zhan Y-J, Xu Y, Rajbhandari R (2017) Changes in extreme temperature events over the Hindu Kush Himalaya during 1961–2015. Adv Clim Change Res 8:157–165. https://doi.org/10.1016/j.accre.2017.07.001
Tennekes M (2018) tmap: Thematic Maps in R. J Stat Softw 84:1–39. https://doi.org/10.18637/jss.v084.i06

240704Supplementrymaterialfinal.docx

Download PDF

Reviewers agreed at journal
06 Aug, 2024
Reviewers invited by journal
26 Jul, 2024
Editor assigned by journal
08 Jul, 2024
First submitted to journal
04 Jul, 2024

You are reading this latest preprint version

Historical and future winter chill for temperate fruit and nut trees in Afghanistan

Status:

Version 1

Abstract

Figures

Introduction

Materials and methods

Study area

Temperature data

Historical data

Winter chill projection

Historical scenarios

Future scenarios

The Dynamic Model for winter chill quantification

Quantification of Safe Winter Chill

Countrywide chill mapping using spatial interpolation

Data analysis and visualization

Results

Historical Safe Winter Chill

Future Safe Winter Chill

Discussion

Conclusion

Declarations

References

Supplementary Files

Status:

Version 1