Statistical inference for calibration points in nonlinear mixed effects models

Many disciplines conduct studies in which the primary objectives depend on inference based on a nonlinear relationship between the treatment and response. In particular, interest often focuses on calibration—that is, estimating the best treatment level to achieve a particular result. Often, data for such calibration come from experiments with split-plots or other features that result in multiple error terms or other nontrivial error structures. One such example is the time-of-weed-removal study in weed science, designed to estimate the critical period of weed control. Calibration, or inverse prediction, is not a trivial problem with simple linear regression, and the complexities of experiments such as the time-of-weed-removal study further complicate the procedure. In this article, we extend existing calibration techniques to nonlinear mixed effects models, and illustrate the procedure using data from a time-of-weed-removal study.     

Dynamic manganese-enhanced MRI signal intensity processing based on nonlinear mixed modeling to study changes in neuronal activity

We analyze data on the impact of testosterone on the dynamics of Mn²⁺ accumulation measured by magnetic resonance imaging in three songbird brain areas: the nucleus robustus arcopallii (RA), area X, and the high vocal center (HVC). Birds with and without testosterone were included in the experiment, and repeated measurements were available in both a preand post-drug administration period. We formulate a nonlinear modeling strategy, allowing for the incorporation of (1) within-bird correlation, (2) the nonlinearity of the profiles, and (3) the effect of treatment. For two of the outcomes (RA and area X), biological theory suggests a parametric form, but for HVC this is not the case. Because the HVC outcome bears some resemblance with the two-compartment model known from pharmacokinetics, this model was considered a sensible choice. We use a different model, based on fractional polynomials, as a sensitivity analysis for the latter. All methods used provide good fits to the data, confirm results from previous, simple analyses undertaken in the literature, but were able to detect additional effects of treatment that had so far gone undetected. The fractional polynomial and two-compartment models provide similar substantive conclusions; the two together can be seen as a form of sensitivity analysis.     

The use of a water budget model and yield maps to characterize water availability in a landscape

Crop yield maps may contain substantial corollary information regarding the distribution of yield related soil properties across a landscape. One of these properties is water holding capacity (WHC). Since WHC is an important parameter for crop models and is also critical for crop yield, our objective was to determine if WHC could be estimated by matching simulated yield with yield map data. We collected soil cores for water retention measurements and recorded plant phenological stages from 60 plots on four transects over two growing seasons (1997 and 1998). Soil cores were also sampled on 40 other locations set out on a grid pattern. We utilized a simple water budget model that uses the relative transpiration ratio to calculate relative yield from available water in the soil profile. Rainfall, potential evapotranspiration and soil water holding capacity are input. An optimization program varies the WHC to produce a grain yield similar to the one from the yield map at a particular location. This analysis was carried out over several scales by averaging yields over 55 m×71 m, 27 m×35 m, and 11 m×14 m areas. Yield data from 2 years were used. Yields from the transects in both years were significantly related to measured WHC in the surface 0–10 cm of soil. The calculated stress indices from the water budget model and estimated available WHC calculated for the 1997 data were similar to those calculated for the 1998 data where data were aggregated in 27 by 35 m or larger blocks. The contour map of estimated WHC was similar to the map of measured WHC for some features though there were also some differences. Use of multiple years of yield data are required to give stable results for estimated water holding capacities.. This information could be used in a farm management plan by allowing a producer to classify a field into areas that are buffered against drought and areas more susceptible to drought.     

Incorporating LASSO effects into a mixed model for quantitative trait loci detection

The identification of quantitative trait loci (QTL) can be viewed as a subset selection problem. In a simulation study the least absolute selection and shrinkage operator (LASSO) is shown to be a useful and powerful tool for QTL identification. LASSO effects are embedded into a mixed model allowing simultaneous modeling of genetic and experimental effects. This provides the flexibility to model the experiment in conjunction with the power of LASSO QTL identification. Estimation is performed using an approximation to the restricted likelihood and modified Gaussian elimination. The extended mixed model is used to analyze a cattle gene mapping dataset.     

Combined effects of technical grade fenitrothion, humic acids and particulate matter on cholinesterase activity in freshwater invertebrates

Purpose

The relative sensitivity of two freshwater invertebrate organisms to the organophosphorus insecticide fenitrothion was assessed by measuring cholinesterase (ChE) activity, a well-known biomarker of both exposure and effect to organophosphorus pesticides. The influence of different concentrations of humic acids (HAs) and particulate matter on fenitrothion bioavailability was assessed in the more sensitive species.

Materials and methods

The selected invertebrates were the dwelling feeding oligochaete Lumbriculus variegatus and the pulmonate gastropod Biomphalaria glabrata. Acute 48-h bioassays were performed exposing organisms to different fenitrothion concentrations. The concentrations that induced 50 % inhibition of enzyme activity (EC₅₀) were calculated. Fenitrothion bioavailability was investigated using different concentrations of commercial HA or particulate matter. Sand and a diverse selection of chromatographic resins that have been proposed as analogues of natural sediments were selected. For these experiments, animals were exposed to a fenitrothion value similar to the EC₅₀.

Results and discussion

The 48-h EC₅₀ values were 12?±?2 and 23?±?3 μg?l^?1 for L. variegatus and B. glabrata, respectively. Depending on HA concentration and the characteristics of particles, ChE activity was similar or higher than the value recorded for animals exposed only to the pesticide in aqueous solution.

Conclusion

The results indicated that L. variegatus was the more sensitive species of the two. In this species, fenitrothion bioavailability did not increase due to the presence of either different HA concentrations or particulate matter. The experimental approach may constitute a useful tool to predict the influence of dissolved organic matter and sediment particles on fenitrothion bioavailability and toxicity to non-target aquatic invertebrates.     

Using computed tomography images to characterize the effects of soil compaction resulting from large machinery on three-dimensional pore characteristics in an opencast coal mine dump

Journal of Soils and Sediments - The use of heavy machinery in mining areas increases during mining and dumping of waste soils, causing severe soil compaction and potentially affecting soil...     

森林养分循环模型应用

The nutrient cycling model NuCM is one of the most detailed models for simulating processes that influence nutrient cycling in forest ecosystems. A field study was conducted at Tieshanping, a Masson pine （Pinus massoniana Lamb.） forest site, in Chongqing, China, to monitor the impacts of acidic precipitation on nutrient cycling. NuCM simulations were compared with observed data from the study site. The model produced an approximate fit with the observed data. It simulated the mean annual soil solution concentrations in the two simulation years, whereas it sometimes failed to reproduce seasonal variation. Even though some of the parameters required by model running were measured in the field, some others were still highly uncertain and the uncertainties were analyzed. Some of the uncertain parameters necessary for model running should be measured and calibrated to produce a better fit between modeled results and field data.     

Tissue iron distribution and adaptation of iron absorption in rats exposed to a high dietary level of NaFeEDTA

Although it has been shown that iron absorption from NaFeEDTA, a promising iron fortificant, is effectively down-regulated in iron-loaded rats, effects of prolonged exposure to high dietary levels of NaFeEDTA are not well understood. The objectives of this study were to determine whether rats can adapt to a high dietary level of NaFeEDTA by down-regulating iron absorption, and to determine effects on tissue iron distribution, with or without an iron absorption inhibitor. Male Sprague-Dawley rats were exposed to diets supplemented with FeSO4 or NaFeEDTA at 1200 mg of Fe/kg of diet, with or without tea, for 27 days. Iron absorption measured by whole-body counting before and after exposure showed that rats adapted to the high dietary level of FeSO4 or NaFeEDTA by down-regulating iron absorption to a similar extent. However, nonheme iron concentrations in liver and spleen were about 35-50% lower, whereas the concentration in kidney was about 300% higher in rats fed NaFeEDTA, compared to rats fed FeSO4. Tea had no major impact on iron absorption or iron status, regardless of iron source. Our results showed that although iron absorption was down-regulated similarly, body iron distribution was markedly different between rats exposed to FeSO4 and those exposed to NaFeEDTA. Further studies are warranted to determine the effects of prolonged exposure to dietary NaFeEDTA on kidney iron accumulation and kidney function.     

The effects of cations and pH on the phosphatase activity of a mixed enzyme system extracted from soil microorganisms

Enzymes extracted from soil microorganisms exhibited activity with a number of inorganic phosphate substrates. Maximum hydrolysis of sodium and ammonium polyphosphates took place at about pH 8.5 in the presence of magnesium or manganese. Optimum cation concentrations were 2 mM for magnesium and 0.8 mM for manganese. Calcium inhibited both magnesium and manganese activation and was itself a non-activator, but an increase in the amounts of magnesium and manganese beyond their optimum concentrations reduced the inhibitory effects of calcium. There was a decrease in the rate of enzymatic hydrolysis as chain length of the polyphosphates increased, but the substrate cation had no effect.     

Using a simplified field progestagen method to assess ovarian activity in female African elephants

Non-invasive endocrine monitoring is an important conservation tool frequently used to support the management of wildlife species. Investigating ovarian activity and pregnancy rates of free-ranging African elephants can enhance our knowledge about the physiology of this highly social and long-lived species. The utility of a new, field-friendly progestagen enzymeimmunoassay (EIA) method was tested by monitoring ovarian activity of 40 females in Addo Elephant National Park (AENP), South Africa. Fecal samples were collected opportunistically from known individuals (June 2007–September 2009) and fecal progestagen metabolites (FPM) were extracted in the field and analyzed on-site using a visual, qualitative EIA. Those results were compared to the laboratory quantitative EIA assessment of hormone concentrations in the U.S. Qualitative FPM concentrations were strongly correlated with quantitative concentrations (r_s = 0.72). Fecal samples collected from pregnant elephants (n = 22) had significantly higher FPM concentrations than non-pregnant individuals, and month of gestation was the strongest predictor of FPM concentrations. Endocrine profiles suggested that five older females were no longer cycling normally. Simplified field endocrine techniques could transform endangered species research under field conditions and provide reliable and economical new tools for making sensible conservation management decisions based on real-time hormone results.     

An ecological dose-response model approach to short- and long-term effects of heavy metals on arylsulphatase activity in soil

Summary The aim of this study was to provide data to evaluate the short- and long-term effects of heavy metals on arylsulphatase activity in five soils. The effects are fitted on a logistic dose-response model and are presented graphically as the ecological dose (heavy metal concentration corresponding to 50% inhibition; ED₅₀) and ecological dose range (heavy metal concentration range corresponding to 10–90% inhibition; EDR). In 7 out of 22 comparable soil-metal combinations the ED₅₀ decreased significantly over 6 weeks to 18 months of incubation and in two cases the ED₅₀ increased. Toxicity (defined as ED₅₀) was highest in sand and sandy loam and lowest in sandy peat. Cd toxicity in sand, silty loam, and clay varied from 1.08 to 9.04 mmol kg^-1. Both Cr and Ni toxicity varied strongly and decreased with time in some soils while increasing in others. The Cu toxicity ranged from 4.51 to 2 mmol kg^-1 in sand and silty loam, respectively, but remained fairly constant over time. Pb was the least toxic element (14.5 to 59.9 mmol kg^-1). The toxicity of Zn ranged from 5.73 to 148 mmol kg^-1 in sand and sandy peat, respectively. At critical concentrations set by the Dutch Soil Protection Act, Cr, Cu, Ni, and Zn inhibited arylsulphatase by 53, 35, 48 and 97%, respectively.     

Assaying the effects of chemical ameliorants with earthworms and plants exposed to a heavily polluted metalliferous soil

The study examined the effects of chemical ameliorant additions (1% montmorillonite, 1% hydroxylapatite, or 1% ferrous oxide) on the availability of cadmium (Cd), copper (Cu), lead (Pb) and zinc (Zn) to the earthworm, Lumbricus rubellus, exposed for 4 weeks to a circumneutral heavily polluted soil (Cd = 220 μg g^–1; Cu = 35 μg g^–1; Pb = 6070 μg g^–1; Zn = 124500 μg g^–1) in 1:0–1:3 dilutions with a clean soil, under laboratory conditions. Soil type (i.e. the dilution series) had a strong influence on the 1 M ammonium acetate extractable metal fractions in soil and on worm-tissue concentrations of Cd, Pb and Zn. Soil treatments (i.e. amelioration) significantly reduced only the soil Zn extractable fraction; Zn concentrations in worms tended to be lower in amended soils. A second experiment, involving curly cress (Lepidium sativum), grown either directly in the serial soil dilutions with 5% ameliorant additions or in the water-extractable fractions of the soils, indicated that root growth is a more sensitive endpoint of metal availability than chlorophyll assays. It was concluded that: (i) chemical immobilization of metals is probably most effective in soils with low to moderate degrees of metal pollution; (ii) an integrated suite of bioassays incorporating different, ecologically relevant, taxa is to be recommended for monitoring metal bioavailabilities and biological effects.     

Dating chicks: Calibration and discrimination in a nonlinear multivariate hierarchical growth model

The motivation for this work was to investigate the possibility of accurately determining the age of a tern chick using easily obtained body measurements. We describe the construction of a nonlinear multivariate hierarchical model for chick growth and show how it can be estimated using Markov chain Monte Carlo techniques. A simple extension of the analysis allows for estimation of the ages of unknown chicks. Posterior distributions of the unknown ages are derived, so that the accuracy of age determination can be examined. We further extend our model and analysis to include the possibility that chicks fall into distinct groups with different growth characteristics. The technique is illustrated using data on the weight and wing length of black-fronted terns from the Ohau River, New Zealand. It is found that dating to within one day is possible, but only in some areas of the data space. The concept of “braiding” of multivariate growth curves is introduced to explain the varying accuracy of age determination.     

Using a modified DNDC model to estimate N2O fluxes from semi-arid grassland in China

Field measurement results showed that approximately 40% of the annual N₂O losses occurred during the non-growing season and confirmed the importance of spring and autumn periods for assessment of total N₂O losses from semi-arid temperate grassland in China. In the previous study, we found that the 7.2 version of Denitrification-Decomposition (DNDC) model had significantly lower estimates of N₂O losses in spring and autumn time. In this study, three modifications, which mainly focus on the nitrification sub-model, the impact of soil frost and snow cover on gas production and emission, had been made to the model code. Based on field measurement, we concluded that modified version of DNDC model is more suitable for estimating the magnitude and seasonal trends of N₂O losses from this region on a plot scale. By extrapolating the field data with our modified model, we estimated that the annual N₂O emission rates for natural temperate grasslands of northern China is ca. 0.056T_g N₂O-N y⁻¹, i.e. ca. 40% lower than the estimate based only on field measurements (ca. 0.092T_g N₂O-N y⁻¹).     

Porcine skeletal muscle troponin is a good source of peptides with Angiotensin-I converting enzyme inhibitory activity and antihypertensive effects in spontaneously hypertensive rats

In the search for novel peptides that inhibit the angiotensin I-converting enzyme (ACE), porcine skeletal troponin was hydrolyzed with pepsin, and the products were subjected to various types of chromatography to isolate active peptides. Glu-Lys-Glu-Arg-Glu-Arg-Gln (EKERERQ) and Lys-Arg-Gln-Lys-Tyr-Asp-Ile (KRQKYDI) were identified as active peptides, and their 50% inhibitory concentrations were found to be 552.5 and 26.2 microM, respectively. These are novel ACE inhibitory peptides, and the activity of KRQKYDI was the strongest among previously reported troponin-originated peptides. KRQKYDI was slowly hydrolyzed by treatment with ACE, and kinetic studies indicated that this peptide was a competitive inhibitor of the enzyme. When KRQKYDI was administered orally to spontaneously hypertensive rats (SHR) at a dose of 10 mg/kg, a temporary antihypertensive activity was observed at 3 and 6 h after administration.     

黑土有机碳变化的DNDC模拟预测

为探讨黑土有机碳的长期变化规律及DNDC模型在土壤有机碳预测方面的适用性,本文利用吉林省公主岭地区黑土不同施肥措施下的长期定位试验数据,选取不施肥(CK)、单施化肥(NPK)、配施有机肥(NPKM)和增施有机肥(M2+NPK)4个处理进行土壤有机碳分析,并将数据用作DNDC模型验证。验证结果表明:各处理DNDC验证中RMSE值均小于10%(分别为5.09%、6.11%、9.38%、8.36%),说明模拟值与观测值一致性良好,模型可用于该地区土壤有机碳模拟。选取了化肥施用、有机肥施用、秸秆还田比率、温度及降水5个因子进行模型的敏感性分析,结果表明:有机肥的施用对土壤有机碳含量的影响最显著,且这种影响具有持久性。最后模拟了4种施肥情境下未来(至2100年)的土壤有机碳变化情况。结果表明:对照不施肥处理(CK)土壤有机碳含量略有下降,至2100年土壤有机碳含量为11.55 g·kg-1,较试验前土壤初始有机碳(13.2 g·kg-1)下降约12.5%。单施化肥处理(NPK)土壤有机碳含量较为稳定,并未出现土壤有机碳含量下降。配施有机肥(NPKM)和增施有机肥(M2+NPK)处理土壤有机碳含量增加明显,至2100年土壤有机碳含量为24.4 g·kg-1和27.6 g·kg-1,分别较初始有机碳含量上升84.8%和109.1%。     

利用数字高程模型改进高山灰岩坑土壤pH值预测

Among spatial interpolation techniques,geostatistics is generally preferred because it takes into account the spatial correlation between neighbouring observations in order to predict attribute values at unsampled locations.A doline of approximately 15 000 m 2 at 1 900 m above sea level (North Italy) was selected as the study area to estimate a digital elevation model (DEM) using geostatistics,to provide a realistic distribution of the errors and to demonstrate whether using widely available secondary data provided more accurate estimates of soil pH than those obtained by univariate kriging.Elevation was measured at 467 randomly distributed points that were converted into a regular DEM using ordinary kriging.Further,110 pits were located using spatial simulated annealing (SSA) method.The interpolation techniques were multi-linear regression analysis (MLR),ordinary kriging (OK),regression kriging (RK),kriging with external drift (KED) and multi-collocated ordinary cokriging (CKmc).A cross-validation test was used to assess the prediction performances of the different algorithms and then evaluate which methods performed best.RK and KED yielded better results than the more complex CKmc and OK.The choice of the most appropriate interpolation method accounting for redundant auxiliary information was strongly conditioned by site specific situations.     

CH4 production potential in a paddy soil exposed to atmospheric CO2 enrichment

Abstract

An anaerobic incubation experiment was conducted to investigate methane (CH₄) production potential in soil samples collected from a paddy field after exposure to free-air CO₂ enrichment (FACE). The FACE experiment with two CO₂ levels, ambient and ambient + 200 p.p.m.v CO₂ during the rice growing season, was conducted at Shizukuishi, Iwate Prefecture, Japan. The soil was a wet Andosol. Soil samples were taken from the surface (0–1 cm) and the sub-surface (1–10 cm) soil layers 2 months after rice harvest. Sub-samples of the fresh soils were put into glass bottles and submerged under N₂ gas headspace during the incubation. The results showed that, prior to incubation, the contents of total C and dissolved organic C (DOC) were significantly greater in FACE soil than ambient soil. During the incubation, CH₄ production potential was approximately 2–4-fold higher in FACE soil than ambient soil and approximately 500–1,000-fold greater in surface soil than sub-surface soil. In general, the FACE soil contained more DOC than ambient soil, particularly in the surface soil layer. These findings suggest that FACE treatment exerted long-term positive effects on CH₄ production and increased organic C content in this paddy soil, particularly in the surface soil layer.     

Microbial activities in forest soils exposed to chronic depositions from a lignite power plant

Atmospheric emissions of fly ash and SO₂ from lignite-fired power plants strongly affect large forest areas in Germany. The impact of different deposition loads on the microbial biomass and enzyme activities was studied at three forest sites (Picea abies (L.) Karst.) along an emission gradient of 3, 6, and 15 km downwind of a coal-fired power plant (sites Ia, II, and III, respectively), representing high, moderate and low emission rates. An additional site (site Ib) at a distance of 3 km from the power plant was chosen to study the influence of forest type on microbial parameters in coniferous forest soils under fly ash and SO₂ emissions. Soil microbial biomass C and N, CO₂ evolved and activities of l-asparaginase, l-glutaminase, β -glucosidase, acid phosphatase and arylsulfatase (expressed on dry soil and organic C basis) were determined in the forest floor (L, Of and Oh horizon) and mineral top soil (0-10 cm). The emission-induced increases in ferromagnetic susceptibility, soil pH, concentrations of mobile (NH₄NO₃ extractable) Cd, Cr, and Ni, effective cation exchange capacity and base saturation in the humus layer along the 15 km long transect significantly (P<0.05) reflected the effect of past depositions of alkaline fly ash. Soil microbial and biochemical parameters were significantly (P<0.05) affected by chronic fly ash depositions. The effect of forest type (i.e. comparison of sites Ia and Ib) on the studied parameters was generally dominated by the deposition effect. Alkaline depositions significantly (P<0.05) decreased the microbial biomass C and N, microbial biomass C-to-N ratios and microbial biomass C-to-organic C ratios. Microbial respiration, metabolic quotient (qCO₂) and the activities of l-asparaginase, l-glutaminase, β-glucosidase, acid phosphatase and arylsulfatase were increased by long-term depositions from the power plants. Acid phosphatase had the highest specific (enzyme activities expressed per unit organic C) activity values among the enzymes studied and arylsulfatase the lowest. The responses of the microbial biomass and soil respiration data to different atmospheric deposition loads were mainly controlled by the content of organic C and cation exchange capacity, while those of enzyme activities were governed by the soil pH and concentrations of mobile heavy metals. We concluded that chronic fly ash depositions decrease litter decomposition by influencing specific microbial and enzymatic processes in forest soils.