Bayesian analysis of semiparametric linear-circular models

In many environmental and agricultural studies, data are collected on both linear and circular random variables, with possible dependence between the variables. Classically, the analysis of such data has been carried out in a classical regression framework. We propose a Bayesian hierarchical framework to handle all forms of uncertainty arising in a linear-circular data set. One novelty of our multivariate linear-circular model is that, marginally, the circular component is assumed to be a mixture model with an unknown number of von Mises (or circular normal) distributions. We use the Dirichlet process to introduce variability in the model dimensionality, and develop a simple Gibbs sampling algorithm for simulating the mixture components. Although we illustrate our methodology on von Mises mixtures, it is widely applicable. We thus avoid complicated reversible-jump Markov chain Monte Carlo methods, which are considered ideal for analyzing mixtures of unknown number of distributions. We illustrate our methodologies with simulated and real data sets. Using pseudo-Bayes factors, we also compare different models associated with both fixed and variable numbers of von Mises distributions. Our findings suggest that models associated with varying numbers of mixture components perform at least as well as those with known numbers of mixture components. We tentatively argue that model averaging associated with variable number of mixture components improves the model’s predictive power, which compensates for the lack of knowledge of the actual number of mixture components.     

A new variance estimator for parameters of semiparametric generalized additive models

Generalized additive models (GAMs) have become popular in the air pollution epidemiology literature. Two problems, recently surfaced, concern implementation of these semiparametric models. The first problem, easily corrected, was laxity of the default convergence criteria. The other, noted independently by Klein, Flanders, and Tolbert, and Ramsay, Burnett, and Krewski concerned variance estimates produced by commercially available software. In simulations, they were as much as 50% too small. We derive an expression for a variance estimator for the parametric component of generalized additive models that can include up to three smoothing splines, and show how the standard error (SE) estimated by this method differs from the corresponding SE estimated with error in a study of air pollution and emergency room admissions for cardiorespiratory disease. The derivation is based on asymptotic linearity. Using Monte Carlo experiments, we evaluated performance of the estimator in finite samples. The estimator performed well in Monte Carlo experiments, in the situations considered. However, more work is needed to address performance in additional situations. Using data from our study of air pollution and cardiovascular disease, the standard error estimated using the new method was about 10% to 20% larger than the biased, commercially available standard error estimate.     

Phytoextraction of Cd and Zn with Thlaspi caerulescens in field trials

Abstract. Phytoextraction is the remediation of heavy metal contaminated soils using plants that take up metals. Hyperaccumulating plants such as Thlaspi caerulescens are often studied for their possible use for decontamination of Cd and Zn rich soils, but few field trials have been reported, although they are necessary to validate the results of hydroponic and pot studies. This article reports field data for T . caerulescens grown on a calcareous and an acidic soil, both contaminated 20 years ago by either atmospheric depositions or septic-tank wastes. Accelerated cropping using transplants grown three times in eight months was compared to Thlaspi sown twice during the same period. Both were followed by one crop of sown Thlaspi . High Cd and Zn concentrations in the plant shoots compensated for the low biomass production. Annual metal exports with transplanted Thlaspi were 130 g Cd ha⁻¹ and 3.7 kg Zn ha⁻¹ on the calcareous soil and 540 g Cd ha⁻¹ and 20 kg Zn ha⁻¹ on the acidic soil. We concluded that within the framework of the Swiss legislation, remediation of Cd-contaminated soils could be achieved within less than 10 years with one crop of Thlaspi per year, but differences in soil properties could affect the rate of phytoextraction significantly. Total Zn content in both soils was too high to be remediated by T . caerulescens in a realistic time span. Thlaspi did not decrease the NaNO₃-extractable fraction of Cd or Zn in either of the soils.     

Phytoextraction of Cd and Zn with Salix viminalis in field trials

Abstract. Use of high biomass crops such as the willow Salix viminalis to extract metals for soil remediation has been proposed as an alternative to the low biomass-producing hyperaccumulating plants. High yields compensate for the moderate heavy-metal concentrations in the shoots of such species. We report the first long-term trials using Salix viminalis to extract heavy metals from two contaminated soils, one calcareous (5 years) and one acidic (2 years). Total metals extracted by the plants were 170 g Cd ha⁻¹ and 13.4 kg Zn ha⁻¹ from the calcareous soil after 5 years, and 47 g Cd ha⁻¹ and 14.5 kg Zn ha⁻¹ from the acidic soil after 2 years; in the first year outputs were negligible. After 2 years, Salix had performed better on the acidic soil because of larger biomass production and higher metal concentrations in shoots. Addition of elemental sulphur to the soil did not yield any additional benefit in the long term, but application of an Fe chelate improved the biomass production. Cd and Zn concentrations were significantly higher in leaves than stems, highlighting the necessity to collect leaves as well as shoots. On both soils, concentration in shoots decreased with time, indicating a decrease in extraction efficiency.     

Enhanced diagnostics for the spatial analysis of field trials

We report an analysis of a series of uniformity field trials using the technique proposed by Gilmour, Cullis, and Verbyla. In particular, we clarify the role of the sample variogram and present a range of enhanced graphical diagnostics to aid the spatial modeling process. We highlight the implications of the presence of extraneous variation related to commonly used agronomic practices, such as serpentine harvesting.     

一种蔬菜保护地土壤调理剂的应用初报

土壤调理剂是以天然沸石矿粉为母体，配以中量、微量营养元素和营养协调物质复合而成，具有改善土壤理化性状，提高作物抗病能力，净化生产环境的功能，其突出特点是成本低、效果好、无毒副作用。蔬菜保护地施用土壤调理剂后可增加纯收入3150－11250元／hm^2，平均增产10％以上，延长保护地设施使用年限3年以上，土壤调理剂用量为1500kg/hm^2，成本为0.8元／kg，可替代基肥。     

生化抑制剂组合与施肥模式对黄泥田水稻养分累积及利用率的影响

添加生化抑制剂是提高水稻肥料利用率的有效途径之一,本研究结合不同施肥模式探讨其节肥增效的养分利用特征,以期寻找适合黄泥田地区水稻高产高效的施用方式。采用二因素随机区组设计,研究生化抑制剂[脲酶抑制剂N-丁基硫代磷酰三胺(NBPT)和N-丙基硫代磷酰三胺(NPPT)及硝化抑制剂2-氯-6-(三氯甲基)吡啶(CP)]组合与施肥模式(一次性和分次施肥)互作对黄泥田水稻养分吸收、利用和分配的影响,并探讨各养分间及其与产量的相互关系。结果表明:生化抑制剂组合和施肥模式对水稻主要生育期N、P、K累积、转运和分配的影响均存在一定的互作效应,互作条件下各生育期养分间吸收存在协同效应。尿素分次施用处理水稻成熟期N、P、K吸收量较一次性施用处理分别提高11.0%、0.9%、4.2%;氮肥吸收利用率、氮肥农学利用率分别显著提高27.5%、70.8%。不同施肥模式下,配施抑制剂组合(NBPT、NPPT/+CP)显著增加水稻N、P、K吸收量,促进抽穗后干物质生产和N素积累,提高籽粒中的养分分配及N素利用效率。新型脲酶抑制剂NPPT单独施用及与CP配施的水稻养分吸收和利用与NBPT相似。相关分析表明,不同施肥模式下水稻成熟期N、P、K吸收量与籽粒产量均呈极显著正相关。总之,通过施肥技术和抑制剂配施的集成与优化,有利于抽穗后N、P、K的吸收、转运,促进养分积累,大幅度同步提高黄泥田水稻的产量和养分利用效率。     

The use of surface resistance—soil moisture relationships in soil water budget models

An important feature of a soil water budget is the reduction of transpiration from a canopy below the rate of atmospheric demand with increasing soil dryness. Commonly, an empirical relationship between the ratio of actual evaporation (AE) to potential evaporation (PE) and soil water storage is adopted. Alternatively the Penman—Monteith equation can be used with a specified relationship between surface resistance and soil water storage.Using actual evaporation rates determined from instrumented soil water profiles, a relationship between surface resistance and soil water storage can be inferred, and results are presented for different crops and soil-types in the United Kingdom. These results are compared with the surface resistance values implicit in the performance of two layer soil moisture models adopting an empirical AE/PE relationship with soil moisture deficit. The performance of the two approaches with respect to soil moisture estimation is compared.     

The use of simulation models as resource management tools for restoring acidified waters

Several computer simulation models are available to resource managers for predicting the response of watersheds and adjacent surface waters to acidification and subsequent neutralization. Some of these models are useful in designing and implementing liming restoration programs. The simpler models emphasize in-lake biochemical processes involved in the dissolution of CaCO₃ and dilution hydrology; these are best suited for simulating straightforward direct applications to surface waters that have minimal hydrologic interaction with the terrestrial watershed. The more sophisticated models integrate various complex biogeochemical processes in land catchments in addition to in-lake and in-stream chemistry. These are best suited for rigorous, process-oriented research applications, for systems which have significant hydrologic interaction with the terrestrial catchment, and for simulating catchment liming applications. The approach and results from a recently published liming simulation study employing the modified Integrated Lake-Watershed Acidification (ILWAS) model are used to illustrate the methodology for using simulation models to complement field research and operational liming programs. Also, modifications to existing models are identified for increasing their suitability for more robust management applications under a wider variety of conditions in different regions.     

Organic fertilizers derived from plant materials Part II: Turnover in field trials

Our aim was to investigate two different organic fertilizers derived from plant materials (OFDP) with respect to their nitrogen (N) and carbon (C) turnover in field trials planted with small radish (Raphanus sativus L. var. sativus) and white cabbage (Brassica oleracea L. convar. capitata var. alba) or fallow. The two fertilizers investigated were coarse seed meal of yellow lupin (Lupinus luteus L.) and coarse meal of castor cake (Ricinus communis L.). Under cool spring conditions, the soil turnover of yellow lupin–seed meal was slightly enhanced compared to castor‐cake meal. During the vegetation period of the vegetables, N added with both fertilizers was metabolized more or less completely by soil microorganisms. Due to similar efficiencies of the fertilizers tested, no significant difference could be found in the N uptake of plants. From this point of view, yellow lupin–seed meal, which can be produced by farmers themselves, has the potential to replace the widely used castor‐cake meal. Considerable amounts of N may remain in the field after fertilization with OFDPs either as mineral N or as easily mineralizable organic N. This N should be utilized immediately by a succeeding crop to avoid leaching losses.     

Ploughless tillage in Sweden. Results and experiences from 15 years of field trials

Results and experience from 140 field trials carried out during the period 1974–1988 are presented. The whole of Sweden was represented in the investigation, but the majority of sites were in the central and southern areas of the country. In the unploughed experimental plots, conventional autumn ploughing to 20–25 cm was compared with two or three stubble cultivations to approximately 10–12 cm depth. on average, use of ploughless tillage reduced yields of winter wheat, spring barley, winter and spring oilseed rape and sugarbeet and somewhat improved yields of oats and potatoes. The results from ploughless tillage were generally improved by the use of dual rear tractor wheels instead of single wheels, by the placement of artificial fertilizer instead of broadcasting and by the removal of straw. The incidence of annual weeds increased on average by 25% and stoloniferous weeds by 100%. The most suitable soils were peats, sandy loamy tills, silty loams or silty clay loams and heavy clays.     

Sulphur deficiency in oilseed rape and cereals - Experience from Swedish field trials

Field trials in Sweden have shown responses to S-fertilizers by rapeseed, but not by barley. These responses varied from 0 to approximately 200 kg/ha seed. They are, however, hard to predict and the criteria of S availability have to be elucidated.     

Correlation of soil tests with pot and field trials in the evaluation of soil fertility

The practical importance of soil tests in bringing about efficient use of fertilizers is well established. Extensive areas in India are producing Only a fraction of their yield potential due to the lack of enough plant nutrients whose need could be easily assessed by soil tests and suitable recommendations of fertilizer use to increase crop production could be given.     

The use of reflectometric test kits for estimating nitrification and soil invertase activity in the field

The objective of this study was to test the applicability of different reflectometric test kits for the determination of potential nitrification and the activity of invertase in soil directly in the field. For the investigation of nitrification in the field, the incubation time was reduced because a linear correlation between incubation time and nitrification rates was found. Likewise, a linear correlation between invertase activity and the incubation temperature was observed which allowed for the calculation of temperature‐dependent factors. Activity measurements in the field at lower temperatures in comparison to the standard procedure were corrected by these factors. Ten different soils were investigated for potential nitrification and invertase activity with standard methods in comparison to field tests with reflectometric detection of the reaction products. The results showed that the reflectometric test kits are suitable for the estimation of potential nitrification and invertase activity directly in the field.     

不同产量水平的玉米田养分利用效应比较

采用大田试验方法,在不同土壤肥力水平和不同管理措施导致产量水平明显不同的玉米田中,对植株对氮、磷、钾养分的吸收、运转及利用效应进行了分析比较。结果表明,玉米对N、P吸收呈双峰趋势变化,吸收高峰均为拔节—大喇叭口期和灌浆—成熟期,K呈单峰吸收,但高产田和常规田出现在大喇叭口—吐丝期,低产田K吸收高峰提前至拔节—大喇叭口期。与常规田和低产田相比,高产田玉米对N、P、K养分的累积量、玉米营养体贮藏N、P、K养分的运转量和运转率,以及玉米营养体N、P养分向籽粒的转移效率均明显提高,但单位养分的生产效率以及氮、钾养分的生理效率降低。     

土地利用的多尺度空间自相关模型比较（英）

该文以广州市花都区为例,选用拉格朗日乘子检验和拟合度检验分析比较了经典线性回归模型、空间滞后模型、空间误差模型对于土地利用与驱动因子在不同尺度上的解释力和适用度。研究表明：1)经典线性回归模型的残差表现出正相关性,但比原始土地利用数据弱,表明经典线性回归模型能部分地解释土地利用空间布局但忽略了土地利用的空间依赖性;2)与经典线性回归模型相比,空间滞后模型、空间误差模型很好地消除了空间自相关性,并且有更好的拟合度;3)研究区不同地类、不同尺度的最优空间自回归模型的具体形式是不同的,表明最优空间自回归模型的具体形式也存在尺度依赖性。     

Soil nitrogen dynamics after slurry injection in field trials: Evaluation of a soil sampling strategy

Slurry injection below maize seeds is a rather new application technique developed to improve the nitrogen use efficiency of liquid organic manure. To enable the characterization of the spatial and temporal soil mineral nitrogen (SMN) dynamics after slurry injection, the present study aims to develop an appropriate soil sampling strategy. Three consecutive experiments were conducted. The first testing of the soil sampling approach was conducted in an existing field trial where the slurry was injected down to a depth of 12 cm (upper rim) below the soil surface. The soil profile (75 cm wide) centered below the maize row was sampled grid‐like to a depth of 90 cm. Around the injection zone, soil monoliths (SM) were sampled using a purpose‐built soil shovel. Below the SMs and in the interrow space (15 and 30 cm distance to the row) a standardized auger procedure was performed. The second experiment aimed at improving the sampling strategy with a focus on sample homogenization quality and necessary sample sizes per pooled sample. Furthermore, the risk of a carryover of slurry components along the soil core due to drilling an auger through a slurry band was analyzed. In the third experiment this improved sampling strategy was validated. Results from the first testing of the sampling procedure showed that the strategy is suitable, although some problems occurred (especially the high spread in values among the replications causing high coefficients of variation (CV) of mostly 40–60%). The improvement trial revealed that due to the high gradient of SMN concentration in the direct range of the injection zone an intensive homogenization of these samples is required. Suitable sample sizes (twelve auger samples and six soil monolith samples per pooled sample) have to be collected to obtain reliable SMN values. Drilling an auger through a slurry band to sample subjacent soil layers has to be avoided. Following this enhanced sampling strategy, in the final validation trial the spread in values were considerably reduced and resulted in CV values of mostly < 20%. The developed sampling strategy enables the characterization of the spatial and temporal SMN dynamics when slurry has been band‐injected below a maize row. The method can be transferred to other row crops and different slurry injection spacing.     

我国北方立体种养殖稻田氮素利用率研究

针对目前我国北方地区农业面源污染严重、氮肥利用率低的现象,选择北方典型稻区——天津市宝坻水稻种植区为研究区,以整个稻田生态系统为基本研究单元,建立氮素输入和输出模型,并以水稻普通种植模式(CK,水稻单作)为对照进行田间试验,研究水稻立体种养殖模式(RF,水稻-鱼-虾-蟹共作+田埂+沟渠)氮素的吸收利用率。结果表明,两种水稻种植模式氮素的输入主要来自灌溉、施肥和降雨,其中RF输入氮肥128.25 kg(N)·hm-2,与CK相比减少11.75 kg(N)·hm-2,与南方种植水稻地区相比,氮肥施用量减少14%~52%,RF从源头减少氮素输入,降低了营养元素流失风险。CK氮素的输出主要包括土壤固定、氨挥发、侧渗流失和水稻吸收,RF与CK相比,氮素的输出还包括鱼虾蟹的吸收,由于RF特殊的田埂-沟渠生态净化系统,通过侧渗损失的氮素(以NO3--N为主)较CK减少9.33 kg(N)·hm-2。试验期间,RF和CK氨累积挥发量分别为8.91kg(N)·hm-2和21.54 kg(N)·hm-2,RF氨挥发速率为6.9%,比CK低8.5%,比全国平均水平低10.3%;收获期,RF与CK相比,水稻产量增加6.65%,表明稻田养殖鱼虾蟹不会降低水稻产量。RF氮素利用率为64.3%,比CK高19.7%,既实现了水稻丰产,又减少了氮素流失。因此,在满足水稻灌溉需求的北方地区,可以开展水稻立体种养殖模式,以控制北方地区农业面源污染。     

Evaluation of calibrated passive sampling for quantifying ammonia emissions in multi-plot field trials with slurry application

Background

There is a great need for simple and inexpensive methods to quantify ammonia emissions in multi-plot field trials. However, methods that meet these criteria have to be thoroughly validated. In the calibrated passive sampling approach, acid traps placed in the center of quadratic plots absorb ammonia, enabling relative comparisons between plots. To quantify ammonia emissions, these acid trap samplings are scaled by means of a transfer coefficient (TC) obtained from simultaneous measurements with the dynamic tube method (DTM). However, dynamic tube measurements are also comparatively costly and time-consuming.

Aims

Our objective was to assess the best practice for using calibrated passive sampling in multi-plot field trials. One particular challenge in such experiments is to evaluate the influence of ammonia drift between plots.

Methods

In a series of eight multi-plot field trials, acid traps and DTM were used simultaneously on all plots to measure ammonia emissions caused by different slurry application techniques. Data obtained by both methods were correlated, and the influence of the ubiquitous ammonia background on both methods was evaluated by comparing net values, including the subtraction of the background with gross values (no background subtraction). Finally, we provide recommendations for calculating a TC for calibrating relative differences between plots, based on simultaneous acid trap and dynamic tube measurements on selected plots.

Results

Treatment mean values obtained by both methods correlated well. For most field trials, R² values between 0.6 and 0.8 were obtained. Ammonia background concentrations affected both methods. Drift between plots contributed to the background for the acid traps, whereas the contamination of the chamber system might have caused the background for the DTM. Treatments with low emissions were comparatively more affected by that background.

Conclusion

For a robust application of calibrated passive sampling, we recommend calculating the TC based on a treatment with high ammonia emissions, reducing the relative influence of the ubiquitous ammonia background.