Seasonal Variables Research Articles

Predicting global terrestrial biomes with the LeNet convolutional neural network

Abstract. A biome is a major regional ecological community characterized by distinctive life forms and principal plants. Many empirical schemes such as the Holdridge life zone (HLZ) system have been proposed and implemented to predict the global distribution of terrestrial biomes. Knowledge of physiological climatic limits has been employed to predict biomes, resulting in more precise simulation; however, this requires different sets of physiological limits for different vegetation classification schemes. Here, we demonstrate an accurate and practical method to construct empirical models for biome mapping: a convolutional neural network (CNN) was trained by an observation-based biome map, as well as images depicting air temperature and precipitation. Unlike previous approaches, which require assumption(s) of environmental constrain for each biome, this method automatically extracts non-linear seasonal patterns of climatic variables that are relevant in biome classification. The trained model accurately simulated a global map of current terrestrial biome distribution. Then, the trained model was applied to climate scenarios toward the end of the 21st century, predicting a significant shift in global biome distribution with rapid warming trends. Our results demonstrate that the proposed CNN approach can provide an efficient and objective method to generate preliminary estimations of the impact of climate change on biome distribution. Moreover, we anticipate that our approach could provide a basis for more general implementations to build empirical models of other climate-driven categorical phenomena.

Effects of Heat Stress across the Rural-Urban Interface on Phenotypic Trait Expressions of Dairy Cattle in a Tropical Savanna Region

Among all livestock systems in tropical regions, the dairy sector is facing huge challenges to sustain productivity under the rapidly changing climatic conditions. To date, there is a lack of knowledge on the combined effects of climate, season, and farm location on trait responses in different cattle breeds. Consequently, this study presents a novel approach to assess the impact of several climatic and geographical factors on production traits, energy efficiency indicators, and hygiene traits in dairy cattle reared across the rural–urban interface in the tropical savanna region of Bengaluru, a rising megacity in southern India. In total, 96 cattle were selected across Bengaluru’s rural–urban interface, reflecting a broad variety of social-ecological systems. The traits considered included test day milk yield (MY), body condition score (BCS), body weight (BW), hock assessment score (HAS), udder hygiene score (UHS), and upper leg hygiene score (ULHS). Apart from cow-related factors such as breed, lactation stage, lactation number, and milking frequency, the environmental classification variables of season, farm location (as expressed by survey stratification index, SSI), and temperature humidity index (THI) significantly affected most of the traits, with indication for breed-by-environment interactions. In particular, season significantly influenced production and hygiene traits. Furthermore, an evident breed variation was observed in the seasonal influence on BW, wherein exotic cows had a higher BW than crossbreds during the summer season. The distinct trend of SSI in its influence on most of the traits indicates that cows housed in urban areas had better trait expression than those in rural areas, thereby revealing a predominant role of management. The THI had a significant effect on MY, BCS, and HAS, and THI = 75 was identified as heat stress threshold. The results indicate the importance of considering ecological, social, and climatic factors simultaneously in order to improve primary and functional breed-specific traits of dairy cattle reared in challenging environments.

Gametogenic cycle of the marine gastropod Trophon geversianus in a harsh Patagonian intertidal environment

ABSTRACT The edible marine gastropod Trophon geversianus shows a large geographic distribution along the southern tip of South America, inhabiting both rocky intertidal and shallow subtidal zones, with distinct behavioural responses to environmental stressful conditions (desiccation, temperature amplitude and high irradiation). Nevertheless, its reproductive biology has been poorly reported to date. We studied the gametogenic cycle of an intertidal rocky shore population from northern Patagonia (Argentina), in order to investigate the relation between gonadal changes and seasonal environmental variables. Mature individuals were taken monthly during an annual reproductive season and processed by standard histology procedures (haematoxylin and eosin). Adults of T. geversianus registered different gonadal maturation stages along the year, showing a marked seasonal variation without a resting period. The highest gamete evacuation season in both sexes was registered when the environmental stress was minimum (winter), while a greater number of individuals with gonads in proliferation and growth were predominant during maximum environmental stress (summer). The reproduction pattern presented an extended period of evacuation and a short phase of non-evacuation, associated to drastic seasonal environmental changes in the intertidal zone of Atlantic Patagonia. This information is useful to protect the resource and assure the sustainability of intertidal populations.

Integrating historical observations alters projections of eastern North American spruce–fir habitat under climate change

AbstractSpruce–fir (Picea–Abies) forests of the North American Acadian Forest Region are at risk of disappearing from the northeastern United States and Canada due to climate change. Species distribution models (SDMs) have been used to predict changes in this critical transitional ecosystem in the past, but none have addressed how seasonal patterns of temperature and precipitation interact to influence tree species abundance. Inferences have also been limited by contemporary inventory data that could not fully characterize species ranges because they either, (1) only sampled species occurrence after large‐scale human disturbance and settlement, or (2) did not span critical geopolitical boundaries (e.g., the US–Canadian border) that intersect the focal species' range(s). Here, we built new SDM models to better assess the bioclimatic distribution of four spruce–fir species and to test the importance of seasonal climate interactions. We compiled an extensive database of tree occurrence and abundance from recent (~1955–2012) and historical time periods (1623–1869) to model current species distributions and to predict how these might change under future climate. We found that including historical tree data in our SDMs revealed previously unrecognized suitable habitat along the southern edge of species' contemporary ranges. Random forest models predicted occurrence with high accuracy (area under receiver operator curve >0.98), and the seasonal climate variables that emerged as most important for these cold‐adapted species all included interactions that reflected sensitivity to colder temperatures, and preferences for wet weather concentrated in the winter months. Under moderate climate warming (representative concentration pathway 6.0), the northeastern United States retained additional suitable habitat when historical data were included through 2060 for three of the four species: red spruce (Picea rubens), black spruce (Picea mariana), and balsam fir (Abies balsamea), while white spruce (Picea glauca) habitat contracted into Canada. In contrast, future predictions from models that used contemporary data alone forecast extirpation for all four species from the northeastern United States. Overall, these findings highlight that prediction of species ranges in transitional ecosystems that span geopolitical boundaries and gradients of intense land use are improved when historical data and seasonal climate interactions of both temperature and precipitation variables are incorporated.

Time-series cohort study to forecast emergency department visits in the city of Milan and predict high demand: a 2-day warning system

ObjectivesThe emergency department (ED) is one of the most critical areas in any hospital. Recently, many countries have seen a rise in the number of ED visits, with an increase...

Dengue fever occurrence in India, Brazil, Paraguay, Philippines and Singapore using Google trends

Background. Dengue is a mosquito-borne disease and its occurrence is a worldwide problem. The health seekers may turn to the Internet based data for possible complications and various therapeutic approaches for the treatment. Therefore, our objective is to identify and characterize the seasonal transition of dengue fever using Google Trends. Methods. This was a cross-sectional study of dengue fever internet searches between 2004 and 2017. The Google was indexed the aggregated web data and ranked based on the search volume with respect to time. The countries selected for the study include India, Brazil, Paraguay, Philippines and Singapore based on the maximum number of cases across world. The selected regions were further classified in to seasonal time points and the relative search volume ranks were categorized to two season variables. Mann–Whitney U test was applied to identify the statistical significance. Results. The summer to monsoon transitions were showed highest number cases beginning from 2004 in the selected regions. Furthermore, dengue searches were peaked in both the seasons of summer and monsoon in selected regions. The analyses of data further revealed that 16.2 % of variation in dengue trends is associated with the time between 2004 and 2017.

China's Economic Forecast Based on Machine Learning and Quantitative Easing.

In this paper, six variables, including export value, real exchange rate, Chinese GDP, and US IPI, and their seasonal variables, are used as determinants to model and forecast China's export value to the US using three methods: BP neural network, ARIMA, and AR-GARCH. Error indicators were chosen to compare the simulated and predicted results of the three models with the real values. It is found that the results of all three models are satisfactory, although there are some differences in their simulation and forecasting capabilities, but the ARIMA model has a clear advantage. This paper analyses the reasons for these results and proposes suggestions for improving China's exports in the context of the models.

Spatio-Temporal Differentiation Characteristic and Evolution Process of Meteorological Drought in Northwest China From 1960 to 2018

Against the backdrop of global climate change, the response characteristic of meteorological drought is of great concern, especially in the arid or semi-arid regions. By employing the Standardized Precipitation Index (SPI), TPFW-MK test, Run Theory, Moran’s I, and General G, the spatio-temporal evolution characteristic of drought was clarified and the spatial autocorrelation of local and global drought characteristic variables was explored based on the meteorological data from 122 stations in Northwest China (NWC) during 1960–2018. The results indicated that the drought situation of NWC was improving regardless of annual or seasonal scale. According to the Z-statistics by the TPFW-MK test, there existed an obvious wet trend in west NWC and a slight dry trend in east NWC. The center of gravity migration model revealed that the gravity center of SPI moved towards higher latitude over the last decades, there was a northwest (1960–1990) and northeast (1990–2018) variation in the covering shapes of the standard deviational ellipses of SPI, and the spatial distribution of SPI tended to be concentrated. Meanwhile, the distribution pattern of drought characteristics suggested that more droughts occurred in east of NWC, which were less harmful while fewer droughts happened in west NWC, which brought greater drought damage. The results of global Moran’s I (GMI) indicated that both annual and seasonal drought variables were characterized with significant spatial autocorrelation, the spatial distribution of winter drought variables was more disperse than other seasons, while the damage of summer and autumn drought was bigger than that in spring and winter. Besides, the results of local Moran’s I (LMI) showed that there was obvious agglomeration in the overall distribution of drought characteristic variables, which had a seesaw effect. The spatial distribution of hot spots and cold spots at different confidence levels indicated that Shaanxi Province experienced the most droughts but with shortest duration and lowest severity while northwest Xinjiang had the fewest droughts with longest duration and highest severity. The results of revealing the drought development process and identifying the location of drought aggregation will provide references for supporting climate adaptation strategies and preventing drought-related loss.

An inventory of human night-biting mosquitoes and their bionomics in Sumba, Indonesia.

Mosquitoes are important vectors that transmit pathogens to human and other vertebrates. Each mosquito species has specific ecological requirements and bionomic traits that impact human exposure to mosquito bites, and hence disease transmission and vector control. A study of human biting mosquitoes and their bionomic characteristics was conducted in West Sumba and Southwest Sumba Districts, Nusa Tenggara Timur Province, Indonesia from May 2015 to April 2018. Biweekly human landing catches (HLC) of night biting mosquitoes both indoors and outdoors caught a total of 73,507 mosquito specimens (59.7% non-Anopheles, 40.3% Anopheles). A minimum of 22 Culicinae species belonging to four genera (Aedes, Armigeres, Culex, Mansonia), and 13 Anophelinae species were identified. Culex quinquefasciatus was the dominant Culicinae species, Anopheles aconitus was the principal Anopheles species inland, while An. sundaicus was dominant closer to the coast. The overall human biting rate (HBR) was 10.548 bites per person per night (bpn) indoors and 10.551 bpn outdoors. Mosquitoes biting rates were slightly higher indoors for all genera with the exception of Anopheles, where biting rates were slightly higher outdoors. Diurnal and crepuscular Aedes and Armigeres demonstrated declining biting rates throughout the night while Culex and Anopheles biting rates peaked before midnight and then declined. Both anopheline and non-anopheline populations did not have a significant association with temperature (p = 0.3 and 0.88 respectively), or rainfall (p = 0.13 and 0.57 respectively). The point distribution of HBR and seasonal variables did not have a linear correlation. Data demonstrated similar mosquito-human interactions occurring outdoors and indoors and during early parts of the night implying both indoor and outdoor disease transmission potential in the area-pointing to the need for interventions in both spaces. Integrated vector analysis frameworks may enable better surveillance, monitoring and evaluation strategies for multiple diseases.

Assessment of temporal shifting of PM2.5, lockdown effect, and influences of seasonal meteorological factors over the fastest-growing megacity, Dhaka

Dhaka is subjected to high pollution levels throughout the year, holding some relatively high amounts of pollution readings, making its air unhealthy to breathe. The study examined hourly, shifting, seasonal fluctuations in particulate matter (PM2.5), the effects of seasonal meteorological variables, and the lockdown effect over the megacity of Dhaka from 2019 to 2021 using data from AirNow. The results indicate the daily average PM2.5 concentration between 2019 and 2021 was 112.49 µg/m3, about four times higher than the WHO limit and two times higher than the Bangladesh standard. Daily PM2.5 concentrations was high during morning and evening pick-up hours, reaching a maximum hourly concentration of 472.9 µg/m3 in February 2020. The maximum average PM2.5 concentration was 211.23 µg/m3 in March 2021 (winter season), and the lowest average was 27.58 µg/m3 in August 2020 (rainy season). The Pearson correlation coefficient (r) between the PM2.5 and meteorological variables were inverse with rainfall (− 0.62), temperature (− 0.73), humidity (− 0.82), but positive with wind (0.09). Daily average Air Quality Index (AQI) concentrations improved from 108.53 to 67.99 µg/m3 during the lockdown period. Finally, the study recommended many mitigation strategies that might assist accountable authorities in lowering the number of life-threatening components in the air.

Customer base analysis with recurrent neural networks

One of the primary goals that researchers look to achieve through customer base analysis is to leverage historical records of individual customer transactions and related context factors to forecast future behavior, and to link these forecasts with actionable characteristics of individuals, managerially significant customer sub-groups, and entire cohorts. This paper presents a new approach that helps firms leverage the automatic feature extraction capabilities of a specific type of deep learning models when applied to customer transaction histories in non-contractual business settings (i.e., when the time at which a customer becomes inactive is unobserved by the firm). We show how the proposed deep learning model improves on established models both in terms of individual-level accuracy and overall cohort-level bias. It also helps managers in capturing seasonal trends and other forms of purchase dynamics that are important to detect in a timely manner for the purpose of proactive customer-base management. We demonstrate the model performance in eight empirical real-life settings which vary broadly in transaction frequency, purchase (ir)regularity, customer attrition, availability of contextual information, seasonal variance, and cohort size. We showcase the flexibility of the approach and how the model further benefits from taking into account static (e.g., socio-economic variables, demographics) and dynamic context factors (e.g., weather, holiday seasons, marketing appeals). We make an open-source reference implementation of the newly developed method available at https://github.com/valendin/rfm2lstm.

Winter–Spring Prediction of Snow Avalanche Susceptibility Using Optimisation Multi-Source Heterogeneous Factors in the Western Tianshan Mountains, China

Data-driven methods are commonly applied in avalanche hazard evaluation. However, few studies have tapped into the relationship between the explanatory variables and avalanche hazard in arid–frigid areas, and the seasonal dynamics of avalanche hazard and its attribution has not been discussed. Therefore, to fill the gap in the hazard assessment of a dry–cold snow avalanche, quantify the dynamic driving process of seasonal nonlinear explanatory variables on avalanche hazard, and improve the reliability of the assessments, this study used Support Vector Machine (SVM), Random Forest (RF) and K-Nearest Neighbour (KNN) algorithms to construct three assessment models; these were used and verified in the western Tianshan Mountains, China. The following results were obtained: The causative factors of avalanches varied based on the season. In winter, terrain and snow depth played a major role, whereas spring was mainly influenced by snow depth and meteorological factors. The dynamic process of avalanche hazard was mainly governed by the seasonality of snow depth and temperature. The seasonal changes in avalanche hazard increased from low to high. The performance of all models was consistent for season and more reliable than the inter-annual evaluations. Among them, the RF model had the best prediction accuracy, with AUC values of 0.88, 0.91 and 0.78 in winter, spring and the control group, respectively. The overall accuracy of the model with multi-source heterogeneous factors was 0.212–0.444 higher than that of exclusive terrain factors. In general, the optimised model could accurately describe the complex nonlinear collaborative relationship between avalanche hazard and its explanatory variables, coupled with a more accurate evaluation. Moreover, free from inter-annual scale, the seasonal avalanche hazard assessment tweaked the model to the best performance.

Non-parametric seasonal unit root tests under periodic non-stationary volatility.

This paper presents a new non-parametric seasonal unit root testing framework that is robust to periodic non-stationary volatility in innovation variance by making an extension to the fractional seasonal variance ratio unit root tests of Eroğlu et al. (Econ Lett 167:75–80, 2018). The setup allows for both periodic heteroskedasticity structure of Burridge and Taylar (J Econ 104(1):91–117, 2001) and non-stationary volatility structure of Cavaliere and Taylor (Econ Theory 24(1):43-71, 2008). We show that the limiting null distributions of the variance ratio tests depend on nuisance parameters derived from the underlying volatility process. Monte Carlo simulations show that the standard variance ratio tests can be substantially oversized in the presence of such effects. Consequently, we propose wild bootstrap implementations of the variance ratio tests. Wild bootstrap resampling schemes are shown to deliver asymptotically pivotal inference. The simulation evidence depicts that the proposed bootstrap tests perform well in practice and essentially correct the size problems observed in the standard fractional seasonal variance ratio tests, even under extreme patterns of heteroskedasticity.Supplementary InformationThe online version contains supplementary material available at 10.1007/s00180-022-01211-w.

Influence of seasonal and environmental variables on the emission of methane from the mangrove sediments of Goa.

Mangrove sediments are known sources for methane emission that has a very high global warming potential. The spatio-temporal emission of methane in the mangrove sediments was quantified in the present study using the static closed chamber technique. Besides, the effects of environmental parameters on methane emission were estimated at Betim (mouth), Chorão (midstream), and Volvoi (upstream) stations along the tropical Mandovi estuary. On an average, the methane emission at the upstream estuarine station at Volvoi was maximum (1268.68 ± 176nMcm-2h-1) compared to the other two stations. Annually, the methane emission was significantly influenced by physicochemical parameters like salinity at Betim and Volvoi and, the redox potential at the midstream station at Chorão. The variation of methane emission between the 3 stations (P < 0.001) is attributed to the variation in methanotrophy (P < 0.05) and methanogenesis (P < 0.05) influenced by differences in the concentration of nutrients (P < 0.05) and organic carbon (P < 0.05). Seasonally, the highest methane emission at Chorão was during the post-monsoon, at Betim was during the monsoon season (1305.34 ± 108.58nMcm-2h-1), and at the upstream station at Volvoi, the emission of methane was highest during the pre-monsoon season (1514.68 ± 130.94nMcm-2h-1). The influence of environmental parameters was more prominent on methane emission at the 3 stations during the monsoon season. Spearman's correlation analysis indicated that seasonal changes in methane emissionare not only attributed to the influence of seasonal rainfall that leads to the fresh water input, but also to the variation in biogeochemical parameters.

HOW CHANGES IN CLIMATE AFFECT CROP YIELDS IN EASTERN INDIA

Climate change wreaks havoc on the planet’s ecosystem, affects weather patterns and undermines food security. This paper investigates the effect of changing climate on Indian agriculture, and the climate sensitivity of food grain yield in the eastern region. Using a panel set of physical crop yields and analyzing district-level data from 1990 to 2017, the study estimates agricultural production functions for specific food grains with extrinsic seasonal weather variables. These include average temperature, rainfall, evapotranspiration and wind speed. Spatially disaggregated district-level data provides information from low administrative units, creating a broader picture to be examined. The research indicates that climate change significantly affects east Indian agriculture, and its outcomes are consistent with other state and district-level studies conducted in India. Changing weather patterns over the year considerably influence crop production and disrupt food availability and quality. Hence, the study further expands the implications of results for inter-region inequality and recommends corrective measures.

Insect defoliation modulates influence of climate on the growth of tree species in the boreal mixed forests of eastern Canada.

Increasing air temperatures and changing precipitation patterns due to climate change can affect tree growth in boreal forests. Periodic insect outbreaks affect the growth trajectory of trees, making it difficult to quantify the climate signal in growth dynamics at scales longer than a year. We studied climate‐driven growth trends and the influence of spruce budworm (Choristoneura fumiferana Clem.) outbreaks on these trends by analyzing the basal area increment (BAI) of 2058 trees of Abies balsamea (L.) Mill., Picea glauca (Moench) Voss, Thuja occidentalis L., Populus tremuloides Michx., and Betula papyrifera Marsh, which co‐occurs in the boreal mixedwood forests of western Quebec. We used a generalized additive mixed model (GAMM) to analyze species‐specific trends in BAI dynamics from 1967 to 1991. The model relied on tree size, cambial age, degree of spruce budworm defoliation, and seasonal climatic variables. Overall, we observed a decreasing growth rate of the spruce budworm host species, A. balsamea and P. glauca between 1967 and 1991, and an increasing growth rate for the non‐host, P. tremuloides, B. papyrifera, and T. occidentalis. Our results suggest that insect outbreaks may offset growth increases resulting from a warmer climate. The observation warrants the inclusion of the spruce budworm defoliation into models predicting future forest productivity.

Influence of seasonal weather variables and habitat type on numbers of colonies of the giant honey bee in Nepal

Influence of seasonal weather variables and habitat type on numbers of colonies of the giant honey bee in Nepal

Seasonal variation and plankton physico-chemical characteristics of Omeremaduche River, Abia state, Niger delta, Nigeria

Omeremaduche river was surveyed for Physico-chemical and planktonic characteristics for a period of thirty-four (34) weeks. Three sections of the River were sampled and human activities were considered in the choice of sampling stations. Statistical analysis was done using the Analysis of Variance (ANOVA). The influence of Physico-chemical characteristics on plankton distribution were evaluated using Pearson’s Correlation. The mean values for pH, water-temperature, DO, BOD, phosphate and nitrate during the wet season were 6.91, 27.3, 5.86mg/l, 7.24mg/l, 0.38 and 4.18 respectively, while mean values of the same characteristics during the dry season were 6.68, 26.8, 7.22mg/l, 5.44mg/l, 0.25mg/l and 3.09mg/l. There was a significant difference in seasonal variation in DO, BOD, Nitrate and Phosphate, while there was no significant difference in pH and Water-temperature. The dry season variables were more favourable to plankton survival. Phytoplankton was characterized by four (4) phyla namely, Bacillariophyceae, Chlorophyceae, Euglenophyceae and Chrysophyceae while two (2) phyla Cladocera and Rhizopoda, represented the Zooplanktons. Bacillariophyceae recorded the highest number of individual species across the stations. The occurrence of plankton in Station 2, which is the point of dense anthropogenic activities, were less than what was obtained in the other sections. However, increased concentration of nutrients and dissolved oxygen were recorded in Station 3; this might have influenced the high plankton survival in the station. Changes in Physico-chemical characteristics of the River could be responsible for the varied plankton survival rate in the river. The seasonal variation may be attributed to varied rainfall patterns, while the partial variation may be due to the influence of various anthropogenic activities taking place around the river.

Sources and Seasonal Variance of Ambient Volatile Organic Compounds in the Typical Industrial City of Changzhi, Northern China

Volatile organic compounds (VOCs) emitted from industrial processes, which are major emission sources of air pollutants, could cause significant impacts on air quality. However, studies on the comprehensive analysis from sources contributing to the health risk perspective regarding ambient VOCs in industrial cities are limited. In this study, VOC samples were collected from 15 April 2018 to 19 October 2018 in Changzhi, a typical industrial city in northern China, and a total of 57 VOCs were measured for analysis. The average VOC concentrations were 54.4 µg·m−3, with the highest concentrations in autumn (58.4 µg·m−3). Ambient VOCs in spring, summer and autumn were all dominated by alkanes (66.8%), with contributions of 70.3%, 66.3% and 63.8%, respectively. The top five concentrations of total VOCs were isopentane (19.0%), ethane (9.5%), n-butane (8.1%), benzene (7.9%) and propane (5.2%), indicating that vehicle exhaust and coal combustion are the main sources of VOCs. Source apportionment by principal component analysis showed that vehicle exhaust (27.5%) and coal combustion (23.5%) were the main sources of VOCs in Changzhi, followed by industrial production (17.4%), solvent evaporation (13.5%), liquefied petroleum gas/natural gas leaking (9.5%), and biogenic emissions (8.7%). Sources of coal combustion and vehicle exhaust contributed more VOCs than industrial production. The carcinogenic risks of benzene (3.4 × 10−5) and ethylbenzene (2.2 × 10−6) were higher than the limit levels (1 × 10−6). Coal combustion contributed most (25.3%) to the carcinogenic risks because of its high VOC emissions. In an industrial city such as Changzhi, vehicle exhaust and coal combustion have become major sources of ambient air VOCs owing to the increasingly stringent industrial standards. Therefore, VOCs from vehicle exhaust and coal combustion also need to take into account mitigation measures for VOCs from the perspective of source contribution to health risk.

Nucleotide Variation in Drosophila cryptochrome Is Linked to Circadian Clock Function: An Association Analysis.

Cryptochrome (CRY) is a conserved protein associated with the circadian clock in a broad range of organisms, including plants, insects, and mammals. In Drosophila, cry is a pleiotropic gene that encodes a blue light-dedicated circadian photoreceptor, as well as an electromagnetic field sensor and a geotaxis behavior regulator. We have generated a panel of nearly-isogenic strains that originated from various wild populations and which carry different natural alleles of cry. Sequencing of these alleles revealed substantial polymorphism, the functional role of which was elusive. To link this natural molecular diversity to gene function, we relied on association mapping. Such analysis revealed two major haplogroups consisting of six linked nucleotides associated with circadian phase (haplotypes All1/All2). We also generated a maximum-likelihood gene-tree that uncovered an additional pair of haplogroups (B1/B2). Behavioral analysis of the different haplotypes indicated significant effect on circadian phase and period, as well on the amount of activity and sleep. The data also suggested substantial epistasis between the All and B haplogroups. Intriguingly, circadian photosensitivity, assessed by light-pulse experiments, did not differ between the genotypes. Using CRISPR-mediated transgenic flies, we verified the effect of B1/B2 polymorphism on circadian phase. The transgenic flies also exhibited substantially different levels of cry transcription. We, moreover, analyzed the geographical distribution of the B1/B2 haplotypes, focusing on a 12 bp insertion/deletion polymorphism that differentiates the two haplotypes. Analysis of cry sequences in wild populations across Europe revealed a geographical cline of B1/B2 indel frequency, which correlated with seasonal bioclimatic variables. This spatial distribution of cry polymorphism reinforces the functional importance of these haplotypes in the circadian system and local adaptation.