Some Comments on Gerechte Designs

In a companion paper we derived the correct analysis for gerechte designs with uncorrected errors. Here we show that this correct analysis cannot be justified by the usual randomization argument. However, when the regions are rectangular there is a randomization procedure which validates an analysis with three separate error terms. We also outline other developments in design and analysis that may be more satisfactory.     

The Relationship Between the Chemical Properties of Native Woody Species and the Soil,in a Secondary Forest of the Eastern Chaco (Argentina)

Abstract

The horizontal distribution of trees within forests is due in part to the composition of the substrate, but there are also effects of the trees on the properties of the soil. This work focused on the relationship between the chemical composition of the foliar tissues of four species of the most abundant deciduous trees found in the “quebrachal” (xerophitic forest), and the chemical properties of the soil under their crowns. The species were two trees: Schinopsis balansae Engl. and Acacia praecox Gris. and two bushes, Maytenus vitis-idaea Gris. and Achatocarpus praecox Gris. The chemical variables that were analysed were: C/N, N, Ca, Mg, K and P. The most important association between the foliar tissues and the soil was found for P, under the crown of S. balansae. These results would reassure the idea that the most important woody species within the “quebrachal” would be modifying the conditions of the substrate.     

Foraging mode: a factor affecting the susceptibility of spiders (Araneae) to insecticide applications

Field experiments have revealed that some species of spiders are more sensitive to insecticides than others. Among many factors influencing their susceptibility, foraging mode seems to play an important role. Aspects of foraging mode that appear to be relevant are whether the spider is diurnal or nocturnal, a hunter or a web-maker. Six spider species, Araniella opisthographa, Clubiona neglecta, Dictyna uncinata, Pardosa agrestis, Philodromus cespitum and Theridion impressum were used in the study. P agrestis and P cespitum are diurnal hunters that may come into direct contact with insecticide. C neglecta is nocturnal and so is exposed to residues only. The remaining three species are web-makers building webs that vary in the extent to which they can protect the spider from direct spray. The effect of sprays was tested under laboratory conditions (Potter tower) with three commercial insecticides, an insect growth regulator (hexaflumuron), a selective organophosphorus (phosalone) and a non-selective pyrethroid insecticide (permethrin) using a four-day exposure period. Data were analysed using bootstrap method and randomization tests. The results obtained showed that hunting spiders were more susceptible to the insecticides tested than web-makers (in their webs). Diurnal hunting spiders (Philodromus and Pardosa) were severely affected only by permethrin. A high mortality was observed for the nocturnal hunter, Clubiona, after application of phosalone and permethrin. This species appears to be very sensitive to residues of both insecticides. Comparing the effect on web-making spiders, with and without webs, it was observed that the sparse orb-web of Araniella did not protect its owner at all, but the dense cribellate and frame-webs of Dictyna and Theridion, respectively, reduced the mortality caused by permethrin significantly in comparison with specimens without webs. Of other factors studied, posture (normal and upside-down position) did not influence the susceptibility. Mortality increased slightly with body size after permethrin application. © 1999 Society of Chemical Industry     

不同免疫接种方法评价:多元方差分析方法及其应用

应用多元方差分析方法评价不同免疫接种方法(交巢与胸肌注射)接种乙肝疫苗与对照的抗-HBsAg的水平,结果表明:(1)在不同时间测量得到的抗-HBsAg水平,在不同处理的鸡只间存在差异,同时在同一处理的不同鸡只间也存在程度不同的差异,对照鸡只抗体含量明显低于其他两种处理,而交巢与胸肌注射之间比较接近。(2)多元方差分析的几种统计量均说明不同免疫接种方法处理的鸡只抗-HbsAg水平存在极显著的差异(F检验,P<0.001);胸肌注射与交巢注射明显优于对照,而这两种方法之间不存在显著差异。(3)不同处理在不同测量时间所获的抗-HBsAg平均水平的最小二乘估计值及其相应的95%置信区间说明:除试验初次测量值外,其它各反应量均存在显著差异,交巢与胸肌注射处理明显优于对照,而这2个处理之间不存在显著差异。(4)应用Bonferonni95%置信区间同时比较不同处理鸡只抗-HBsAg水平测量值之间差异显示:除初次与第2次测量值之外,其余抗-HBsAg水平测量结果之间均存在显著差异,交巢与胸肌注射明显优于对照,而这两种方法之间不存在显著差异。由此可见,不同处理间单一反应量的比较与所有反应量同时进行统计比较,其结果是不一样的。     

A multi-scale spatial analysis method for point data

This paper presents a nearest neighbor method for the spatial analysis of data collected from discrete field sampling sites. The method was applied to point counts of birds at permanent survey sites in the Nicolet National Forest of northeastern Wisconsin. The spatial analysis method we developed uses a Monte Carlo randomization approach to test for non-randomness not only of the mean nearest neighbor distance between n points but also the mean second nearest, third nearest,..., to (n–1)th nearest distances to reveal spatial information at multiple scales. Because the bird survey sites are not randomly distributed throughout the forest, the survey sites at which a given species was recorded were compared with random samples drawn from the total survey sites rather than from all possible points within the forest. More refined analyses restricted the randomization by (a) habitat type, in order to separate the effects of non-randomly distributed habitat types on species' distributions; and (b) north-south regions of the forest, in order to account for regional gradients in distribution which were evident for some species. Spatial patterns among the sites at which the birds were detected reveal information about the scale at which the birds are distributed in their environment and provide a more complete picture of multi-scale bird population dynamics.     

Some Comments on Gerechte Designs

Gerechte designs are row and column designs which have an additional blocking structure formed by spatially compact regions. In this paper, and a companion paper (B ailey et al. 1990) we consider the correct analysis of these designs. In this paper we discuss the purpose of gerechte designs, and we note that the regions are not orthogonal to the rows and columns. This means that the usual sums of squares for rows, columns and regions are not additive. We show how the correct analysis can be performed, either through the use of appropriate formulae, or by the use of a statistical package. In the companion paper we show that the correct analysis cannot be justified by a randomization argument, and we outline other developments in analysis and design that may be more satisfactory.     

Why Randomize Agricultural Experiments?

This study illustrates the importance of randomization using two hypothetical field trials, one with a marked systematic trend and the other with a more erratic spatial pattern. The insights from these two examples are reinforced by analysis of a uniformity trial and a small simulation study. Results illustrate that both model‐based spatial analysis and randomization‐based analysis assuming independent errors are valid with full randomization but may be invalidated when randomization is lacking. It is concluded that randomization provides protection against different forms of spatial trend. The examples given in the study serve as a general reminder that agricultural experiments should be randomized whenever possible.     

A Hitchhiker's Guide to Mixed Models for Randomized Experiments

Designed experiments conducted by crop scientists often give rise to several random sources of variation. Pertinent examples are split‐plot designs, series of experiments and repeated measurements taken on the same field plot. Data arising from such experiments may be conveniently analysed by mixed models. While the mixed model framework is by now very well developed theoretically, and good software is readily available, the technology is still under‐utilized. The purpose of the present paper is, therefore, to encourage more widespread use of mixed models. We outline basic principles, which help in setting up mixed models appropriate in a given situation, the main task required from users of mixed model software. Several examples are considered to demonstrate key issues. The theoretical underpinnings are briefly sketched in so far as they are practically relevant for making informed use of mixed‐model computer packages. Finally, a brief review is given of some recent methodological developments, which are of interest to the plant sciences. A German version of this paper is available from the corresponding author upon request.