Controlling nitrogen release from farm ponds with a subsurface outflow device: Implications for improved water quality in receiving streams

The retention of nutrients in farm ponds has many potential benefits, including reduction of nitrogen and phosphorus (promoters of eutrophication) in receiving streams. The aim of this study was to evaluate the efficacy of a commercial subsurface pond outflow control device (Pond Management System™) on nutrient retention in farm ponds. Four ponds of similar size and water chemistry in the upper Tar River basin of North Carolina, USA were studied; three were equipped with the pond outflow control device and one was retained without a device (normal surface outflow) that served as a reference site. Water samples were collected monthly from each pond at 0.3 m intervals from the surface to 2.1 m at a fixed station adjacent to the pond standpipe and from the pond outflow pipe from March to October 2005. The water samples were analyzed for total Kjeldahl nitrogen (N), total phosphorus (P), chlorophyll a, and a suite of other physicochemical variables. In ponds with the subsurface outflow device, the mean N concentrations in the outflow were substantially less (6.2–20.7%) than concentrations at the pond surface. Concentrations of N in the outflow were similar to N concentrations at intermediate pond depths (0.9–1.5 m), the depth of the outflow devices, indicating water was drawn from these depths and that N was being retained in the surface layers of the pond. Also, mean water temperatures were 1.1–1.9 °C cooler at intermediate depths compared to the surface, suggesting potential application of the outflow device for minimizing warm water outflows to receiving streams. These results provide evidence that under these conditions a subsurface pond outflow device can reduce nutrient release to receiving streams, thereby increasing overall stream water quality.     

Water production functions of wheat (Triticum aestivum L.) irrigated with saline and alkali waters using double-line source sprinkler system

Expected yield losses as a function of quality and quantity of water applied for irrigation are required to formulate guidelines for the effective utilisation of marginal quality waters. In an experiment conducted during 2004-2006, double-line source sprinklers were used to determine the separate and interactive effects of saline and alkali irrigation waters on wheat (Triticum aestivum L.). The study included three water qualities: groundwater (GW; electrical conductivity of water, ECw 3.5 dS m⁻¹; sodium adsorption ratio, SAR 9.8 mmol L⁻¹; residual sodium carbonate, RSC, nil) available at the site, and two synthesized waters, saline (SW; ECw 9.4 dS m⁻¹, SAR 10.3 mmol L⁻¹; RSC nil) and alkali (AW; ECw 3.7 dS m⁻¹, SAR 15.1 mmol L⁻¹; RSC 9.6 meq. L⁻¹). The depths of applied SW, AW, and GW per irrigation ranged from 0.7 to 3.5 cm; the depths of applied mixtures of GW with either SW (MSW) or AW (MAW) ranged from 3.2 to 5 cm. Thereby, the water applied for post-plant irrigations using either of GW, SW or AW ranged between 15.2 and 34.6 cm and 17.1 and 48.1 cm during 2004-2005 and 2005-2006, respectively and the range was 32.1-37.0 and 53.1-60.0 cm for MSW or MAW. Grain yields, when averaged for two years, ranged between 3.08 and 4.36 Mg ha⁻¹, 2.57 and 3.70 Mg ha⁻¹ and 2.73 and 3.74 Mg ha⁻¹ with various quantities of water applied using GW, SW and AW, respectively, and between 3.47 and 3.75 Mg ha⁻¹ and 3.63 and 3.77 Mg ha⁻¹ for MSW and MAW, respectively. The water production functions developed for the two sets of water quality treatments could be represented as: RY = 0.528 + 0.843(WA/OPE) − 0.359(WA/OPE)² − 0.027ECw + 0.44 × 10⁻²(WA/OPE) × ECw for SW (R² = 0.63); RY = 0.446 + 0.816(OPE/WA) − 0.326(WA/OPE)² − 0.0124RSC − 0.55 × 10⁻⁴(WA/OPE) × RSC for AW (R² = 0.56). Here, RY, WA and OPE are the relative yields in reference to the maximum yield obtained with GW, water applied for pre- and post-plant irrigations (cm), and open pan evaporation, respectively. Crop yield increased with increasing amount of applied water for all of the irrigation waters but the maximum yields as obtained with GW, could not be attained even with increased quantities of SW and AW. Increased frequency of irrigation with sprinklers reduced the rate of yield decline with increasing salinity in irrigation water. The sodium contents of plants increased with salinity/alkalinity of sprinkled waters as also with their quantities. Simultaneous decrease in potassium contents resulted in remarkable increase in Na:K ratio.     

An assessment of collective action for pond management in Zhanghe Irrigation System (ZIS), China

Ponds are common property resources that allow users to obtain water on-demand because of their capacity to store rainwater and other surplus water close to users. Effective management of these ponds is crucial, especially with increasing water scarcity and decreasing water deliveries for agriculture that are taking place in many irrigation systems, including the Zhanghe Irrigation System (ZIS) in China. The purpose of this study was to determine the nature and performance of collective action in pond management, and to quantify the factors affecting that collective action. We used two different approaches to measure the performance of collective action: a performance indicator based on soil conditions, and farmer’s perceptions based on their satisfaction with the current level of management activities. The results using the two approaches were consistent, and indicated that, among other variables, dependency on pond water, pond size, and household size were important variables affecting the performance of collective action. There was also limited evidence that clarification of property rights through the use of contractors for pond management improved outcomes for farmers.     

Response of hot pepper (Capsicum annuum L.) to mulching practices under planted greenhouse condition

Mulch is considered a desirable management technology for conserving soil moisture, improving soil temperature and soil quality. This study aimed to investigate soil conditions and hot pepper (Capsicum annuum L.) performance in terms of leaf photosynthetic capacity, fruit yield and quality, and irrigation water use efficiency (IWUE) under such practices in greenhouse condition. A field experiment across 3 years was carried out with four types of mulch (without mulch [CK], wheat straw mulch [SM], plastic film mulch [FM], and combined mulch with plastic film and wheat straw [CM]). Mulch could improve soil physical properties regardless of mulch materials. FM and CM treatments improved soil moistures status and soil temperature in comparison to CK control, while SM increased soil water content and decreased soil temperature. Mulch increased leaf net photosynthesis rate (P_N), stomatal conductance to water vapor (gs), intercellular CO₂ concentration (Ci), and transpiration rate (E), but declined instant water use efficiency (WUEi). No significant effect of mulch application on chlorophyll fluorescence was existent for the entire growth season. Fruit yield and irrigation water use efficiency (IWUE) showed some increment under all the mulch conditions. Compared to CK, the yield was enhanced by 82.3%, 65.0%, and 111.5% in 2008; 38.1%, 17.4%, and 46.5% in 2009; and 14.3%, 6.5%, and 19.6% in 2010 under SM, FM, and CM conditions, respectively. Although FM produced better fruit quality than other treatments, CM is the recommended practice for hot pepper cultivation in greenhouse condition due to working well to facilitate soil condition (moisture and temperature), plant growth, and marketable yield.     

Unrestricted cattle access to streams and water quality in till landscape of the Midwest

Unrestricted cattle access to streams in traditionally pastoral regions has been linked to increased concentrations of bacteria, suspended sediments and associated contaminants in streams. However, there is a dearth of data available regarding the impact of cattle access to streams in poorly drained landscapes of the Midwest. In this study, we investigate changes in water quality on a 1005 m long stream section impacted by cattle grazing on the upper 130 m. We monitor discharge, water quality [nitrate, ammonium, total Kjeldahl nitrogen (TKN), total phosphorus (TP), total suspended sediments (TSS), turbidity, Escherichia coli] and chloride, atrazine, silica and major cation concentrations over a 12-month period. Cattle access to the stream does not significantly affect nitrate concentration but leads to large increases in TKN (fourfold increase), TP (fivefold increase), ammonium (fourfold increase), TSS (11-fold increase), turbidity (13-fold increase) and E. coli (36-fold increase) in the summer/fall period. In particular, E. coli concentration in the stream in the summer/fall period exceeds 235 colony forming unit (CFU)/100 ml 64% of the time upstream from the section impacted by cattle, but exceeds this same threshold 89% of the time immediately downstream. Despite the negative impact of cattle access to the stream on water quality, data indicate that dilution, in-stream processes, and natural stream geometry downstream from the impacted section help mitigate this pollution. We expect that this study will be an incentive for policy makers to promote stream rehabilitation and develop more stringent guidelines limiting cattle access to streams in many Midwestern states and other regions with poorly drained soils where the impact of cattle access to streams on water quality is often ignored.     

The hydroponic production of lettuce (Lactuca sativa L) by using hybrid catfish (Clarias macrocephalus × C. gariepinus) pond water: Potentials and constraints

An experiment was conducted at the Asian Institute of Technology, Thailand, for 54 days to investigate on the effect of pond water filtration and aggregates (hydroponic substrates) on lettuce (Lactuca sativa L.) production, nutrient uptake and growth. A 2 × 3 factorial experiment consisting of 2 filtration regimes (unfiltered versus partially filtered pond water) and 3 aggregates (styrofoam as control, sand, and gravel). The hydroponic units were placed on a bamboo platform across the pond surface. An earthen pond (226 m²) was stocked with 8000 hybrid catfish with an average initial weight of 6.58 ± 1.72 g. The fish were fed with commercial pelleted feeds containing 30% crude protein. Twenty-one days old lettuce seedlings were transplanted on to hydroponic units after stocking catfish for 4 months in the pond. Pond water was partially filtered by using a settling tank and filtration tank, which contained netting materials to trap solids. Lettuce plots were irrigated twice daily at 0800 and 1600 h using either filtered or unfiltered pond water for 45 min.The highest head weight and yield of lettuce were observed with plants grown on the sand media followed by gravel and the control treatments. Partially filtered pond water treatments had significantly higher lettuce yield (P < 0.05) and filtration increased lettuce yield of the control, gravel and sand treatments by 87, 63 and 52%, respectively, over unfiltered water treatments. Lettuce grown on gravel and the filtered water control treatments had adequate tissue N content. Nutrient rich hybrid catfish (Clarias macrocephalus × Clarias gariepinus) pond water provides an opportunity to use it in the hydroponic vegetable production.Although pond water is mainly constrained by the high suspended solids, and low dissolved oxygen, the study showed that filtration of catfish pond water enhanced the potential to use it for the hydroponic lettuce production.     

Combining hydrological modeling and GIS approaches to determine the spatial distribution of groundwater recharge in an arid irrigation scheme

Accurate quantification of the rate of groundwater (GW) recharge, a pre-requisite for the sustainable management of GW resources, needs to capture complex processes, such as the upward flow of water under shallow GW conditions, which are often disregarded when estimating recharge at a larger scale. This paper provides (1) a method to determine GW recharge at the field level, (2) a consequent procedure for up-scaling these findings from field to irrigation scheme level and (3) an assessment of the impacts of improved irrigation efficiency on the rate of GW recharge. The study is based on field data from the 2007 growing season in a Water Users Association (WUA Shomakhulum) in Khorezm district of Uzbekistan, Central Asia, an arid region that is characterized by a predominance of cotton, wheat and rice under irrigation. Previous qualitative studies in the region reported irrigation water supplies far above the crop water requirements, which cause GW recharge. A field water balance model was adapted to the local irrigation scheme; recharge was considered to be a fraction of the irrigation water losses, determined as the difference between net and gross irrigation requirements. Capillary rise contribution from shallow GW levels was determined with the HYDRUS-1D model. Six hydrological response units (HRUs) were created based on GW levels and soil texture using GIS and remote sensing techniques. Recharge calculated at the field level was up-scaled first to these HRUs and then to the whole WUA. To quantify the impact of improved irrigation efficiency on recharge rates, four improved irrigation efficiency scenarios were developed. The area under cotton had the second highest recharge (895 mm) in the peak irrigation period, after rice with 2,514 mm. But with a low area share of rice in the WUA of <1 %, rice impacted the total recharge only marginally. Due to the higher recharge rates of cotton, which is grown on about 40 % of the cropped area, HRUs with a higher share of cotton showed higher recharge (9.6 mm day^?1 during August) than those with a lower share of cotton (4.4 mm day^?1). The high recharge rates in the cotton fields were caused by its water requirements and the special treatment given to this crop by water management planners due to its strategic importance in the country. The scenario simulations showed that seasonal recharge under improved irrigation efficiency could potentially be reduced from 4 mm day^?1 (business-as-usual scenario) to 1.4 mm day^?1 (scenario with maximum achievable efficiency). The combination of field-level modeling/monitoring and GIS approaches improved recharge estimates because spatial variability was accounted for, which can assist water managers to assess the impact of improved irrigation efficiencies on groundwater recharge. This impact assessment enables managers to identify options for a recharge policy, which is an important component of integrated management of surface and groundwater resources.     

Influence of deficit irrigation in phase III of fruit growth on fruit quality in ‘lane late’ sweet orange

The aim of this work was to apply one strategy of deficit irrigation (DI) to improve the final fruit quality in 10-year-old ‘Lane late’ sweet orange grafted on Carrizo citrange (Citrus sinensis L. Osb. × Poncirus trifoliata L.). The experiment was carried out over 2 years in an experimental orchard located in Torre Pacheco (Murcia, south-east Spain). The deficit irrigation treatment consisted of the stopping of irrigation in phase III of fruit growth (1st October-28th February). The irrigation cut-off in phase III reduced the midday stem water potential (Ψ_md), the plant water status being heavily influenced by rainfall. In both years, the DI treatment did not alter fruit yield although mean fruit weight was slightly reduced. The main effects of DI on the final fruit quality were increases of total soluble solids (TSS) and titratable acidity (TA) and a decrease of juice percentage without altering the final maturity index. Plant water-stress integral (S_Ψ) was correlated positively with TSS and TA and negatively with juice percentage. In conclusion, a DI strategy could be useful for improving the final content of TSS and the TA, therefore allowing the harvest to be delayed.     

Water quality assessment of different reservoir types in relation to nutrient solution use in hydroponics

Hydroponics requires good quality water. For this purpose, water quality is based on concentrations of specific ions and phytotoxic substances as well as the presence of organisms and substances that can clog irrigation systems. Here, four irrigation reservoirs, i.e. two rainwater ponds, a peat ditch, and a natural lake, were analyzed to determine whether or not they conform to water quality guidelines. Based on our data, the four reservoirs could be divided into two categories in respect to their water quality. The two rainwater ponds belong to the category characterized by low input of ionic strength (480 μmol m⁻¹), low concentration of unwanted ions, such as SO₄²⁻ (63 μmol l⁻¹) and Zn²⁺ (3.9 μmol l⁻¹), a moderate bacterial population (lg 4.9 CFU m⁻¹), and moderate algae density (lg 6.0 cells ml⁻¹). The rainwater ponds were found to contain a good diversity in bacteria (45 species from 25 genera), and a poor diversity of algae (15 species from 4 groups). The other category, to which the peat ditch and natural lake belong, is characterized by a high ionic strength (12,200 μmol l⁻¹), high concentrations of alkali ions (Mg²⁺: 890 μmol l⁻¹; Ca²⁺: 3.260 μmol l⁻¹; K⁺: 470 μmol l⁻¹), a moderate bacterial (lg 4.7 CFU ml⁻¹), but low algae density (lg 5.0 cells ml⁻¹). In comparison to the first category, the diversity of the bacteria was poor (seven species from three genera). However, in sharp contrast was the rich algal community detected in the peat ditch, for which 32 species from six groups were found, whereas in the natural lake, only one group with seven species was identified. In all reservoirs, species of the genera Paenibacillus and Bacillus were detected, and small green algae, e.g. Scenedesmus spp., also dominated in each case. Overall, the bacterial and algal densities showed wide fluctuations between water sources, and neither caused filter clogging as observed in investigations of others. The quality of the rainwater investigated was assessed to be well suited for use in hydroponics due to appropriate nutrient concentration (except Zn²⁺ in one pond), and lack of potential bacterial and algal development. However, we recommend water from the natural lake and the peat ditch to be used with care because of the high nutrient concentration.     

Water quality implications of raising crop water productivity

Because of a growing and more affluent population, demand for agricultural products will increase rapidly over the coming decades, with serious implications for agricultural water demand. Symptoms of water scarcity are increasingly apparent, threatening ecosystem services and the sustainability of food production. Improved water productivity will reduce the additional water requirements in agriculture. However, there is a tradeoff between the quantity of water used in agriculture and the quality of return flow. Where yields are low due to limited nitrogen (N) and water supply, water productivity can be enhanced through higher fertilizer applications and improved water management. This limits the amount of additional water needed for increased food demand, thus leaving more water for environmental requirements. But it also increases the amount of nitrate (NO₃–N) leaching, thus adversely affecting the water quality of return flows.This paper quantifies the tradeoff between enhanced water productivity and NO₃–N leaching and shows the importance of the right management of water and N applications. Using the Decision Support System for Agro-technology Transfer (DSSAT) crop model, several scenarios combining different water and N application regimes are examined for maize (Zea mays L.) in Gainesville, FL, USA. Without adequate water, nitrogen use efficiency (NUE) remains low, resulting in substantial NO₃–N leaching. Too much water leads to excessive NO₃–N leaching and lower water productivity. The lack of N is a cause of low water productivity but too much of it leads to lower NUE and higher losses. The paper concludes that increased NO₃–N leaching is an inevitable by-product of increased water productivity, but its adverse impacts can greatly be reduced by better management of water and N application. The paper briefly shows that leaching can be reduced and water productivity increased by split application of N-fertilizer. This implies that improved water and nutrient management at field- and scheme-level is a prerequisite to limit adverse impacts of agriculture on ecosystems, now and especially in the future.     

Water savings, nutrient leaching, and fruit yield in a young avocado orchard as affected by irrigation and nutrient management

This project was designed to determine the effect of fertilizer rate and irrigation scheduling on water use, nutrient leaching, and fruit yield of young avocado trees (Persea americana Mill. cv. Simmonds). Seven nutrient and irrigation management practices were evaluated: (1) irrigation based on crop evapotranspiration (ET) with 50% fertilizer at a standard rate (FSR); (2) ET irrigation with FSR (typical for avocado production in the area); (3) ET irrigation with 200% FSR; (4) irrigation based on exceedance of 15-kPa (SW) soil water suction with 50% FSR; (5) SW with FSR; (6) SW with 200% FSR; and (7) irrigation at a set schedule (based on timing and frequency typically used in local avocado production) with FSR. The SW with FSR treatment saved 87% of the water volume applied and reduced total phosphorus leached by 74% compared to the set schedule irrigation with FSR. The SW with FSR treatment had higher avocado fruit production, tree water-use efficiency, and fertilizer-use efficiency than the other six treatments. Thus, the use of soil water monitoring for irrigation management can substantially increase sustainability of young avocado orchards in southern Florida.     

Containment basin water quality fluctuation and implications for crop health management

Containment basins (CB) are an integral part of recycling irrigation systems that foster agricultural sustainability through water resource conservation. However, little is known regarding this aquatic ecosystem and the lack of water quality data has become an increasingly serious liability in crop health management. Here we report on four distinct seasonal and two diurnal patterns of change in water quality in the CBs. The four seasonal patterns are (a) periodic fluctuation in chlorophyll a, pH, and dissolved oxygen (DO), (b) oxidation–reduction potential (ORP) rises with decreasing DO, (c) tendency for increase in electrical conductivity, salinity, and total dissolved solids, and (d) weather-dependent changes in turbidity and temperature. The two diurnal patterns are (1) chlorophyll a, pH, DO, and temperature consistently peak between 16:00 and 17:00 hours and bottom out around 08:00 hours, and (2) ORP peaks in the morning and bottoms in the evening. Eight of the nine parameters excluding temperature were correlated; and algal blooms appear to be a major driving force for changes in the other seven parameters. These results underscore the importance of water quality monitoring in irrigation management and provide a framework for better understanding of pathogen aquatic ecology and how changes in water quality might be employed in a manner that suppresses plant pathogens and improves crop quality and productivity.     

Assessing grain crop water productivity of China using a hydro-model-coupled-statistics approach: Part I: Method development and validation

Identifying and quantifying the links between water resources and food production is crucial in addressing the intensified conflicts between water scarcity and food security. We proposed an integrated framework for quantifying relationships between food and water based on the concept of green water (GW), blue water (BW) and crop water productivity (CWP). An estimation method coupling hydrologic model and crop and water statistics was developed and validated to quantify basin-scale GW, BW and CWP in breadbasket basins of China. A basin-scale GW and BW assessment method was developed by using the Soil and Water Assessment Tool (SWAT). Monthly-step calibration and validation were performed at 15 discharge flow stations in seven first-order river basins of the country. The coefficient of determination (r²) and Nash-Sutcliffe Efficiency (NSE) in calibration stage ranged from 0.18 to 0.95, and −4.22 to 0.93, respectively; while in validation period, r² ranged from 0.02 to 0.97 and NSE ranged from −266.7 to 0.96. The simulated available soil water was validated against the observed soil moisture data, and the results showed that the model can reflect the yearly average values of soil water storage. Overall, the modeling performance for river basins with 4.94 million km² of drainage areas in total was acceptable. The simulated hydrologic components were then coupled with crop-and-water-statistics-based estimation method for assessing basin-scale CWP on four staple grain crops, i.e. rice, wheat, maize, and soybean. The results were validated by comparing with the similar investigations in China and around the globe. It was concluded that the overall performance of the estimation method was acceptable, and the method can be applied in assessing basin-scale GW, BW and CWP in China.     

Phosphorus availability in sediments from a tidal river receiving runoff water from agricultural fields

Eutrophication of surface water is a worldwide concern. Sediments may play an important role in buffering phosphorus (P) concentration in the overlying water column. However, information on the spatial variation of sediment P availability as affected by agricultural water discharge and hydrological conditions is limited. In this study river sediments were sampled in spring, summer, fall and winter, respectively from seven locations along a main tributary (Ten Mile Creek, TMC), which receives surface runoff water from agricultural lands and discharges into the Indian River Lagoon, south Florida, USA, and analyzed for P availability. Simultaneously, hydrological variables were measured on the spot and river water samples were collected for analyses of water quality. The results demonstrated that available P in the sediments of TMC as measured by several commonly used extraction procedures had a large spatial variation. The downstream locations had a greater amount of available P in the sediments than the upstream locations, which is attributable to the settlement of finer particles due to slower water flow and increased influence from salt water in the downstream locations where the fresh river water gradually mixed with salt water from the Indian River Lagoon. Phosphorus availability in the sediments appeared to be related to P sorption by iron and aluminum oxides and the competition for adsorbing sites between SO₄²⁻ and PO₄³⁻. This spatial variation of P availability agrees with the elevation of chlorophyll a (Chla) in overlying water body in the downstream locations of TMC, indicating that the internal P source plays an important role in triggering an algal boom in surface water systems.     

Interception and storage of surface run-off in ponds in small agricultural watersheds,Andhra Pradesh,India

Summary A 8.93 ha graded agricultural watershed was developed with a 0.4% slope of cultivation line farming. Within this agricultural watershed, 6 ponds, each with a volume of about 180 m³, were dug for an average catchment size of 0.91 ha which was referred to as a small watershed. This attempt to store water is the first of its kind in this region at this scale under arable conditions. It was found that about 20% of the annual run-off could be retained by these ponds. With effective sealants, such as plastic lining overlaid with brick work or cement plastering on brick work, water could be retained in the ponds for longer periods to provide for the needs of crops at stress periods. Natural silting was not effective in controlling seepage in these small ponds. Despite wide variations in rainfall during the period 1976–1985, it was observed that the ponds were filled annually from run-off. In this study area increasing pond size to 300 m³ for a 1.0 ha catchment would be desirable to provide enough stored water for supplemental irrigation to the cropping system.     

基于GS的南方水库灌区塘堰蓄水

探讨了基于GIS技术建立塘堰蓄水能力与地形因子关系的方法。提出用稻田塘堰容积率指标来描述塘堰的相对供水能力。用GIS 的空间分析功能，从 DEM获得了研究区域的坡度、高程及河网密度等地形因子。基于实地调查数据，在分析稻田塘堰容积率与单个地形因子关系的基础上，建立了稻田塘堰容积率与地形因子的综合方程。为精确快速的估算区域塘堰供水能力提供了简易的途径，它也可为指导灌区水资源优化管理和灌区改造中新建塘堰容量及位置的确定提供决策依据。     

基于DeepAR-RELM的池塘溶解氧时空预测方法研究

水体溶解氧(Dissolved oxygen，DO)是养殖水产品健康生长的重要生态因子。池塘溶解氧易受多种因素的影响，会产生时间和空间上分布的差异，现有的溶解氧预测方法大多是针对单监测点的时间序列预测，无法描述池塘溶解氧的空间分布，因此，对池塘溶解氧进行时间和空间预测非常重要。本文提出一种基于自回归循环神经网络(Autoregressive recurrent neural network，DeepAR)和正则化极限学习机(Regularized extreme learning machine，RELM)的池塘溶解氧时空预测方法。首先采用样本熵(Sample entropy，SE)衡量各个监测点溶解氧序列的波动程度，采用最大互信息系数(Maximum mutual information coefficient，MIC)衡量监测点溶解氧序列之间的相关性，综合选取出溶解氧序列波动程度较小且与各个监测点相关性较大的监测点作为中心监测点，并以中心监测点为原点，建立池塘空间坐标系；其次采用DeepAR算法构建中心监测点的溶解氧时间序列预测模型，实现中心监测点溶解氧时间序列预测；最后采用RELM算法构建中心监测点与池塘各位置溶解氧之间的空间映射关系模型，结合中心监测点溶解氧时间序列预测值和池塘空间坐标，实现对未来时刻池塘溶解氧的空间预测。该方法在提高时间序列预测精度的同时，实现了对未来时刻池塘溶解氧空间状态的预测。在真实的数据集上进行测试，预测未来24h的池塘空间溶解氧值，均方根误差(RMSE)为1.2633mg/L、平均绝对误差(MAE)为0.9755mg/L、平均绝对百分比误差(MAPE)为14.8732%。并与标准极限学习机（Extreme learning machine，ELM）、径向基神经网络（Radial basis function neural network，RBFNN）、梯度提升回归树（Gradient boosting regression tree ，GBRT）和随机森林（Random forest，RF）4种预测方法进行对比，各评价指标的性能均有较大幅度提升，表明该方法有较好的预测精度和泛化能力，能够较准确地实现池塘溶解氧时空预测。     

A method for spatial prediction of daily soil water status for precise irrigation scheduling

Available water holding capacity (AWC) and field capacity (FC) maps have been produced using regression models of high resolution apparent electrical conductivity (EC_a) data against AWC (adj. R² = 0.76) and FC (adj. R² = 0.77). A daily time step has been added to field capacity maps to spatially predict soil water status on any day using data obtained from a wireless soil moisture sensing network which transmitted hourly logged data from embedded time domain transmission (TDT) sensors in EC_a-defined management zones. In addition, regular time domain reflectometry (TDR) monitoring of 50 positions in the study area was used to assess spatial variability within each zone and overall temporal stability of soil moisture patterns. Spatial variability of soil moisture within each zone at any one time was significant (coefficient of variation [% CV] of volumetric soil moisture content (θ) = 3-16%), while temporal stability of this pattern was moderate to strong (bivariate correlation, R = 0.52-0.95), suggesting an intrinsic soil and topographic control. Therefore, predictive ability of this method for spatial characterisation of soil water status, at this site, was limited by the ability of the sensor network to account for the spatial variability of the soil moisture pattern within each zone. Significant variability of soil moisture within each EC_a-defined zone is thought to be due to the variable nature of the young alluvial soils at this site, as well as micro-topographic effects on water movement, such as low-lying ponding areas. In summary, this paper develops a method for predicting daily soil water status in EC_a-defined zones; digital information available for uploading to a software-controlled automated variable rate irrigation system with the aim of improved water use efficiency. Accuracy of prediction is determined by the extent to which spatial variability is predicted within as well as between EC_a-defined zones.     

Using midday stem water potential for scheduling deficit irrigation in mid–late maturing peach trees under Mediterranean conditions

Irrigation techniques that reduce water applications are increasingly applied in areas with scarce water resources. In this study, the effect of two regulated deficit irrigation (RDI) strategies on peach [Prunus persica (L.) Batsch cv. “Catherine”] performance was studied over three growing seasons. The experimental site was located in Murcia (SE Spain), a Mediterranean region. Two RDI strategies (restricting water applications at stage II of fruit development and postharvest) based on stem water potential (Ψ_s) thresholds (?1.5 and ?1.8 MPa during fruit growth and ?1.5 and ?2.0 MPa during postharvest) were compared to a fully irrigated control. Soil water content (θ_v), Ψ_s, gas exchange parameters, vegetative growth, crop load, yield and fruit quality were determined. RDI treatments showed significantly lower values of θ_v and Ψ_s than control trees when irrigation water was restricted, causing reductions in stomatal conductance and photosynthesis rates. Vegetative growth was reduced by RDI, as lower shoot lengths and pruning weights were observed under those treatments when compared to control. However, fruit size and yield were unaffected, and fruit quality was slightly improved by RDI. Water savings from 43 to 65 % were achieved depending on the year and the RDI strategy, and no negative carryover effect was detected during the study period. In conclusion, RDI strategies using Ψ_s thresholds for scheduling irrigation in mid–late maturing peach trees under Mediterranean conditions are viable options to save water without compromising yield and even improving fruit quality.     

General irrigation efficiency for field water management

On-farm water management systems are traditionally evaluated using a set of performance indices which are inconvenient for evaluation and comparison. We propose a general efficiency (E_g) which is defined as the ratio of crop transpiration to the sum of the volume of applied water and the volume of deficit. E_g combines the characteristics of traditionally used irrigation efficiencies: application efficiency (E_a), storage efficiency (E_s) and the Christiansen's coefficient of uniformity (U_c). Thus, it is possible to compare the performance of different water application systems and/or design and management scenarios using a single index. The relationships of E_g vs. E_a, E_s and U_c are also presented using a transpiration fraction (α) which is defined as the ratio of transpiration to evapotranspiration.