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.     

Estimating the proportions of closely related species: Performance of the two-phase ratio estimator

In many ecological field contexts, accurately classifying closely related species based on phenotypic characteristics may be difficult. In such cases, definitive classification of species may require expensive genetic analysis of tissues or extensive quantitative measurements. If inexpensive phenotype-based species classifications are highly correlated with expensive but definitive classifications, however, then estimating the proportion of a target species in a mixed species complex using a two-phase ratio estimator may prove cost-efficient. In two-phase ratio estimation, a first-phase sample is randomly selected, and phenotype is used to classify all individuals in the sample. A smaller, second-phase sample is then randomly selected from the first-phase sample, and a definitive method is used to classify the individuals in this subsample. Net relative efficiency (i.e., cost-effectiveness) of the optimally allocated two-phase ratio estimator of a proportion depends on the relative costs of classification at the first and second phases, on the phenotypic classification accuracy for the target species (sensitivity) and nontarget species (specificity), and on the magnitude of the target species proportion. Results are presented that allow assessment of the circumstances in which this two-phase estimation approach can be recommended over an equal-cost single-phase approach based only on expensive but definitive classification. An illustrative application of these methods is provided using an example of two closely related, sympatric, anadromous salmonid species: steelhead (Oncorhynchus mykiss) and coastal cutthroat trout (O. clarki clarki).     

Can the sample variance estimator be improved by using a covariate?

Variance componentsare quantities of central interest in many applications, e.g., in cultivar yield stability analysis and in the analysis of measurement errors. In some applications, the feasible sample size is rather limited, leading to estimates of variance components that are subject to considerable sampling variation. For example, new crop cultivars are tested in only a few environments before release to the market, so the sample size for the variance across environments is small. Similarly, testing a new measurement instrument for some chemical compound may be costly, allowing only a limited number of replications. This article investigates the potential for improving the usual sample variance estimator by exploiting covariate information. In a cultivartrial, yield data may be available for only a few environments while meteorological data or data on a standard cultivar has been recorded for a very large number of environments. Likewise, in the analysis of measurement errors, there may be long-term data on a standard measurement procedure that can be used as a covariate to improve the variance estimate for a new instrument. It is shown in this article that the gain in accuracy achieved by using a covariate can be considerable, provided there is sufficient correlation between the covariate and the variable of interest.     

Randomization tests for time effects and heterogeneity in capture probabilities for closed populations

We propose the use of randomization tests to detect time effects and heterogeneity in capture probabilities of individuals in a closed-population mark-recapture study. The results of a simulation study show the randomization procedures can be implemented so that tests are robust to extraneous sources of variation, with good power to detect the desired effects. We apply the tests to two examples of real data, finding very strong evidence of heterogeneous capture probabilities in both data sets and highly significant time effects for one of the sets.     

Evaluating the physical capture method of terminal restriction fragment length polymorphism for comparison of soil microbial communities

Terminal restriction fragment length polymorphism (T-RFLP) is a popular method of comparative microbial community analysis which is normally accomplished by tagging terminal restriction fragments (T-RFs) with a fluorescent primer. Here, we evaluate an alternative method of T-RFLP where T-RFs are physically captured using a biotinylated primer and streptavidin-coated beads. This eliminates one of the primary criticisms of T-RFLP, namely that T-RFs cannot be identified by sequence analysis, and also represents an alternative method for collecting T-RFLP profiles. Microbial communities from forest, agricultural, and turf soils were investigated using several sets of primers specific for different microbial groups. The physical capture method of T-RFLP resulted in similar profiles to those generated by fluorescent T-RFLP. The relationships among ecosystem types captured by both methods and revealed by ordination were virtually identical. The total variance in the profiles that was attributed to ecosystem type was approximately equal, or greater, when generated by the physical capture method, depending on the primers used. However, physical capture T-RFLP resolved fewer T-RFs than fluorescent T-RFLP, and this may reduce the sensitivity to changes in non-dominant populations within the community. Direct cloning and sequencing of physical capture T-RFs revealed that most bands were not comprised of sequences related to those in the database that would generate T-RFs of similar size. T-RFs should therefore be identified by sequencing, rather than by comparing the sizes of T-RFs to computer digests of database sequences. Physical capture T-RFLP should be a useful tool to identify T-RFs by sequencing, and for laboratories without economical access to equipment required to perform fluorescent T-RFLP.     

Identification of mtDNA lineages of Sus scrofa by multiplex single base extension for the authentication of processed food products

A genetic method to identify the breed of origin could serve as a useful tool for inspecting the authenticity of the increasing number of monobreed foodstuffs, such as those derived from small local European pig breeds. Mitochondrial DNA (mtDNA) is practically the only reliable genomic target for PCR in processed products, and its haploid nature and strict maternal inheritance greatly facilitate genetic analysis. As a result of strategies that sought to improve the production traits of European pigs, most industrial breeds presently show a high frequency of Asian alleles, while the absence or low frequency of such Asian alleles has been observed in small rustic breeds from which highly prized dry-cured and other traditional products are derived. Therefore, the detection of Asian ancestry would indicate nonconformity in Protected Denomination of Origin products. This study presents a single base extension assay based on 15 diagnostic mtDNA single nucleotide polymorphisms to discriminate between Asian and European Sus scrofa lineages. The test was robust, sensitive and accurate in a wide range of processed foodstuffs and allowed accurate detection of pig genetic material and identification of maternal ancestry. A market survey suggested that nonconformity of products derived from Portuguese breeds is an unusual event at present, but regular surveys both in the local populations and in commercial products would be advisible. Taking into consideration the limitations presented by other methodologies, this mtDNA-based test probably attains the highest resolution for the direct genetic test for population of origin in Sus scrofa food products.     

Collaborative study of an extension of the Mills et al. method for the determination of pesticide residues in foods.

Fifteen laboratories collaboratively determined parathion, o,p'-DDT, p,p'-DDT, and p,p'-DDE in kale, and lindane, dieldrin, heptachlor epoxide, and p,p'-DDE in eggs, using electron capture gas-liquid chromatography (potassium chloride thermionic detector for parathion). Pesticide levels ranged from 0.033 ppm heptachlor epoxide to 3.32 ppm p,p'-DDT. Mean recoveries (+/- standard deviations) were: parathion 91+/-15%, o,p'-DDT 99+/-8%, P,P'-DDT 97+/-8%, lindane 94+/-4%, dieldrin 106+/-8%, heptachloro epoxide 106+/-5%, p,p'-DDE 94+/-8% in eggs and 101+/-9% in kale. The methods have been adopted as official first action.     

An appreciation of the contribution of Frank Stevenson to the advancement of studies of soil organic matter and humic substances

Purpose

The aims of this paper are to outline the state of knowledge with regard to the chemistry of soil organic matter (SOM) prior to 1950; then, to review and evaluate the contributions made by Frank Stevenson during Stevenson’s research period (1950–1994); and subsequently to outline advances that are being made in the modern era.

Progress in the Stevenson period

Frank Stevenson’s research career began in the middle of the twentieth century when a number of techniques of colloid chemistry were available to him, but relatively few of the recently developed instrumental techniques and other procedures of analytical organic chemistry that have enabled significant advances to be made in the chemistry and properties of SOM components.The contributions that Frank Stevenson has made to the chemistry of SOM and of humic substances (HSs) are an integral part of his book (first and second editions) Humus Chemistry: Genesis, Composition, Reactions. The validity of the terms humus and HSs are being questioned as legitimate terms that describe definable components, and SOM is being viewed as a ‘continuum of progressively decomposing organic compounds’. The legitimacy of isolating the organic matter components from their native soil environment is questioned. Those who pose such questions would do well to consider how progress could have been made in the vital life sciences areas of, for example, proteomics and genomics, without the isolation of the relevant cellular components. We recognise the importance of clear and rigorous definitions of HS components and stress the need to isolate these components from the SOM matrix as a prerequisite to the study of the composition, structure and reactivity of these components. We disagree with proposals or suggestions that do not recognise HSs as a scientific entity, and we feel sure that Frank Stevenson would have supported this stance. Various studies of SOM and of HSs have taken place over the centuries, but progress was slow because the tools required to study such complex systems were not available. Frank Stevenson’s research involved many areas of humic chemistry, and his major advances were in aspects of functionality and in the interactions of humic functional groups with metals and to a lesser extent with anthropogenic organic chemicals. His studies of nitrogen and of ammonia in relation to organic matter also had a very great impact.

Progress in the modern era

Frank Stevenson can be said to have provided the stimulation that enabled beginners and established scientists to obtain a good grasp of the fundamentals of SOM and the humic sciences. His scientific contributions have catalysed many of the significant advances that have been made in the field since he retired. In the final section, some of the advances that have been made using modern analytical techniques are addressed and some of the controversial topics that have recently arisen are discussed.

     

浅析农技推广信息综合服务系统建设中的风险管理

本文以农技推广信息综合服务系统项目的建设为例,通过一定的专家评分方法,得出项目建设过程中存在的六个主要风险,并提出相应的风险控制对策,以降低对项日的影响。     

Incidental capture of marine turtles by the Italian trawl fishery in the north Adriatic Sea

In the north Adriatic Sea on-board observations on midwater and bottom trawlers were carried out during years 1999 and 2000. Results indicate that the north Adriatic Sea, and especially the north-east part, is a very important foraging and overwintering habitat for loggerhead turtles (Caretta caretta). Since the distribution of fishing effort in the area is not known, the total turtle catch by the Italian bottom trawl fleet was conservatively estimated from the lower catch rate observed in the south-west sub-area: 4273 turtle captures/year (95% C.I.=2186-8546). The actual total catch in the whole north Adriatic is likely to be much higher, due both to an unknown Italian trawling effort in the south-west part with multi-gear vessels and to the Italian and Croatian trawling efforts in the north-east part, where catch rate was 15 times greater than in the south-west part. In the south-west sub-area 9.4% of captured turtles were dead and potential mortality (assuming that all comatose turtles would die too) was 43.8%. Hence, trawling in the north Adriatic is likely to represent a serious threat to the populations, and possible conservation measures are discussed.     

Ecological studies in the ratanica catchment (Carpathian foothills,S Poland) — An overview

This paper includes an overview of ecological studies conducted since 1986 in the Ratanica pine-beech forested catchment located in the polluted, high populated southern part of Poland. General characteristics of the catchment (including soil and vegetation, air pollution, input/output of nutrients and pollutants, element budget data and forest health assessment) are presented. Based on biogeochemical and bioindication results, the Ratanica catchment has been classified as a moderately to heavily deteriorated area. Predictions for this forested catchment for various deposition of anthropogenic pollutants, are also disscused.     

An iterative method for establishing priorities for the selection of nature reserves: An example from Tasmania

Past work ascribing priorities for the selection of nature reserves has weighted attributes and applied formulae in a single stage process. This single application of a formula means that areas of differing priority may contain similar mixes of species, communities or habitats, and may thus lead to imbalance in representation when preservation takes place, with some previously unpreserved or poorly preserved species, communities or habitats being found in several of the new reserves, and others being absent. An iterative method that has been applied to assess priorities for the preservation of threatened species in the central east coast of Tasmania overcomes this difficulty. The area with the highest score in the initial stages of analysis is assumed to be preserved and the weightings of attributes are altered accordingly. These new weightings are applied to locate the area of next highest priority, and the process continues until predetermined preservation goals are met.     

Solid phase extraction gas chromatography/electron capture detector method for the determination of organochlorine pesticides in wildlife whole blood.

A gas chromatographic method for the analysis of 10 organochlorine pesticides in 0.5 mL of whole blood is described. Sample preparation involved an ethyl ether and hexane extraction, followed by a silica solid phase extraction cleanup. The pesticides are quantified by gas chromatography/electron capture detection. Method limits of detection ranged from 1.1 to 5.2 microg/L. The mean and standard deviation for the recovery of 10 pesticides was 97.9 +/- 5.5%. Recoveries from whole blood were comparable to recoveries from plasma. This indicates that the preparation of plasma is unnecessary for the quantification of organochlorine pesticides in blood. This approach is particularly useful as a nonlethal approach for monitoring pesticide contamination in small animals for which the volume of blood is limiting.     

Phytoestrogen-low diet for endocrine disruptor studies

Hormonally active chemicals (HACs) that are capable of inducing adverse effects on wildlife as well as human beings are featured as "endocrine disruptors". Various animal studies conducted to clarify the characteristics of HACs, including the uterotrophic assay, are sufficiently sensitive to detect the effect of 17-beta-estradiol in micrograms per kilogram of body weight or lower. In such systems, a trace amount of HACs in the dietary pellets may interfere with the test results and thus can be a serious problem for the low-dose issue, which is now a major topic in the field of endocrine disruptor research. Here, the significance of the hormonal effects of phytoestrogen components in the NIH-07 diet is confirmed and a NIH-07-based open formula "phytoestrogen-low diet" (PLD) is proposed, which effectively reduces uterine weight as well as the uterine luminal epithelial labeling index in ovariectomized rats.     

Methodologies for interrill soil erosion studies

Due to wide range of experimental techniques reported in the literature for determining interrill erodibility and soil loss values, meaningful comparisons between experiments often cannot be made. Furthermore, inaccurate concepts are developed because erosion processes are dependent upon methodologies. The purpose of this paper is to discuss problems related to both laboratory and field rainfall simulator experiments. Rainfall simulators cannot duplicate a wide range of rainfall intensities and, at the same time, have similar energies as natural rainstorms, unless several different nozzles are used. Rainfall intensity in most simulators is created by varying the frequency of spray oscillation. This intermittent spray characteristic of most simulators, and the constant drop size characteristic of other simulators, greatly affects results. Erosion pan design for laboratory studies and preparation of soil samples placed in the pans also can influence erosion results. We conclude that standardization of rainfall simulator design and test procedures will allow better comparison of erosion results to be made among researchers.     

Species and ecosystem level bioindicators of airborne pollution: An analysis of two major studies

Bioindication of air pollution effects has received considerable attention in recent years. It has been almost entirely focused on individual species and relatively little notice has been given to ecosystem level process and function monitors. Longterm research projects in the Niepolomice Forest in southern Poland and the Colstrip area in southeast Montana, U.S.A., were analyzed for both organismic and system level indicators and monitors for SO, trace elements, and fluoride pollution originating in nearby coal-fired industrial processes. Species of lichens exhibited changes in morphology and survival and pine species exhibited pollutant accumulation in needles at both sites. Declines in Scotch pine growth in Poland of up to 20% were compared with declines in western wheatgrass rhizone biomass in Montana to illustrate system wide effects on primary productivity. Directly observable declines in decomposition rate were noted for both sites at higher pollution levels and tied to system wide occurrences of nutrient deficiency and toxicant buildup in soil pools. Pollutant increases in deer antler composition, changes in grasshopper dietary patterns, and lichen density and health were postulated to have system level implications as well.     

Efficacy of perforating the soil to capture and store rain during fallow in dry regions

Perforating topsoil of poor conductivity with artificial macropores can often improve the capture and storage of rain in dryland cropping. However, several features can render it of little merit. These include the inability of artificial macropores effectively to harvest coalescing water on the soil surface, and the increased evaporation caused by the greater surface area of subsoil exposed to the atmosphere in such pores. To ascertain the extent of these effects I used a sprinkling infiltrometer to determine the time-to-ponding of small areas of perforated and unperforated soil subjected to different intensities of ‘rain’, and determined the rate of evaporation on perforated and unperforated microlysimeters. On average the perforated plots tended to take longer to pond than the unperforated ones, but the water-harvesting efficacy of individual artificial macropores was highly variable, depending on their position in the microlandscape, the degree of connectivity with subsoil pores, the rate at which the water was applied, and the antecedent water content of the topsoil. The microlysimeter experiments revealed consistently greater evaporation from the perforated samples, with average daily differences between the treatments ranging from 2 to 26%. Antecedent soil water content and air temperature and windspeed affected the amounts of evaporation from both treatments. The ultimate efficacy of perforation for capturing water will vary spatially and temporally according to the extent to which the properties of the soil and environment impact on the behaviour of the penetrating water.