I am sorry to announce that John Nelder died on Saturday 7th August in Luton & Dunstable Hospital, where he was recovering from a fall. John was very active even at the age of 85, and retained the strong interest in our work – and statistics generally – that we will all remember with deep affection. On 13 May 2010 I took him to the Numerical Algorithms Group’s 40^th Anniversary Celebration near Oxford, where he was pleased to catch up with many old friends from his time on their Technical Policy Committee. Then, on 4 July 2010, he entertained Youngjo Lee, Yudi Pawitan, Mike Kenward, James Roger and myself to lunch – and to some challenging statistical discussions. However, he was becoming increasingly frail and it was a shock but perhaps, in retrospect, not a surprise to hear that he had died peacefully in his sleep.

If you would like to leave messages of condolence below, I will pass them on to his family.

Obituary

John Ashworth Nelder was one of the most influential statisticians of his generation, whose work will continue to have an important and widespread effect on statistical analysis.

John was born on 8 October 1924 in Dulverton, Somerset, UK. He was educated at Blundell’s School and at Sidney Sussex College, Cambridge where he read Mathematics (interrupted by war service in the RAF) from 1942-8, and then took the Diploma in Mathematical Statistics.

Most of John’s formal career was spent as a statistician in the UK Agricultural Research Service, later renamed Agricultural and Food Research Service (AFRC), and now Biotechnology and Biological Sciences Research Council (BBSRC). His first job, from October 1949, was at the newly set-up Vegetable Research Station, later renamed National Vegetable Research Station (NVRS), and now Horticultural Research International, Wellesbourne. Then, in 1968, he became Head of the Statistics Department at Rothamsted, and continued there until his first retirement in 1984. The role of statistician in AFRC was very conducive for John, not only because of his strong interests in biology (and especially ornithology), but also because it allowed him to display his outstanding skill of developing new statistical theory to solve real biological problems. So, for example at NVRS, John developed the theory of general balance to provide a unifying framework for the wide range of designs that are needed in agricultural research (see Nelder 1965). Then, at Rothamsted, he developed the theory of generalized linear models with the late Robert Wedderburn, to overcome the problems of analysing response variables like counts and proportions that do not come from Normal distributions; see the citation classic Nelder & Wedderburn (1972) or the book by McCullagh & Nelder (1989).

This idea of directing statistical research at real biological problems began with the two earlier Heads of Statistics at Rothamsted, R.A Fisher and F. Yates, to whom John became such a worthy successor. However, John emphasized an important additional aspect, namely that the new theory should be implemented in widely-distributed statistical software to enable it to become widely used in practice.

The initial aim for John’s first statistical program, GenStat, was to provide analysis of variance for generally balanced designs. The underlying ideas took shape in 1965–1966 when John visited the Waite Institute of the University of Adelaide to work with Graham Wilkinson, who was then on secondment there from CSIRO (Commonwealth Scientific and Industrial Research Organization). More intensive development began in 1968 when John joined Rothamsted, and the wider statistical and computing expertise available at Rothamsted allowed him to develop GenStat as a truly general-purpose statistical system. GenStat continues in widespread use today, and is distributed by VSN International to users in more than 120 countries. I was honoured to take over leadership of the GenStat development in 1985, after John’s retirement from Rothamsted, and glad that John continued as an enthusiastic (although sometimes critical!) user. GenStat embodied John’s originally-very-novel view that statistical programs should provide a programming environment for the development of new methodology. The procedure structure introduced in 1987 allowed the resulting new programs to be added to the system as new commands. A good example is the suite of procedures for Lee & Nelder’s hierarchical generalized linear models (see below).

John’s other two major contributions to statistical computing came about while he was Chairman of the Royal Statistical Society’s Working Party on Statistical Computing (1967-1984). The first, in 1968, was the Applied Statistics Algorithms, which aimed to support good computing practice by providing implementations of the basic building blocks of a statistical program. Later much more complicated techniques were added, and the publication of an algorithm for a new piece of methodology became an equally valid (and perhaps more effective) way of registering a new idea. The second contribution was the program GLIM which first appeared in 1974, with 4 further releases up to the final GLIM4 in 1993. This implemented Nelder & Wedderburn’s generalized linear models, and led to a dramatic improvement in the quality of statistical analysis allowing unsatisfactory approximate analyses, such as those involving the angular transformation of percentage data, to be discarded. It had an immense influence on the new generation of practical statisticians. For many it provided their first experience of analysing data interactively. It encouraged them to think about each data set, instead of directing it at a black box with a request for “statistics all”. It provided opportunities to investigate a rich set of models, and good diagnostics to assess which one would be most appropriate.

John retired from Rothamsted in 1984 at the age of 60, but continued his research at Imperial College (of Science, Technology, & Medicine, London) where, since 1972, he had been a Visiting Professor. He retired from Imperial College in October 2009. His first task there was to lead the GLIMPSE project (Nelder 1991), which was funded by the UK Government’s Alvey programme to produce a knowledge-based front-end for GLIM. The GLIMPSE system provided advice on data validation, data exploration and model selection. However, it seems to have been intended more as a guide for experts, than as a system to provide expert help to novices, and it never achieved the widespread use that it deserved. However, it contained many very interesting and far-sighted ideas and, when it was released in 1989, it was one of the first statistical expert systems to be made available commercially – and perhaps one of the few to deliver what the originators had promised.

John’s other major activity at Imperial College was his collaboration with Youngjo Lee to develop the theory of hierarchical generalized linear models (HGLMs); see the papers by Lee & Nelder (1996, 2001, 2006) and the book by Lee, Nelder & Pawitan (2006). The 1996 and 2006 papers were presented as “read papers” at meetings of the Royal Statistical Society; it is impressive to note that John was 81 years old when he and Youngjo presented the 2006 paper. HGLMs aimed to provide satisfactory methods of analysis for non-Normal data when there is more than one source of random variation. John viewed generalized linear models as a way of liberating statisticians from the “tyranny” of the Normal distribution, and was a little bemused to see this same tyranny reestablished in methods that were devised initially to extend generalized linear models. These generalized linear mixed models (GLMMs) catered for additional random variation by adding additional Normally-distributed random effects into the linear model of the generalized linear model. John and Youngjo’s new HGLMs extended the methodology to include the beta-binomial, gamma and inverse-gamma distributions, and showed that the conjugate HGLMs (namely binomial GLM with additional beta-binomial random effects, or Poisson with gamma, or gamma with inverse gamma) had attractive advantages in their mathematical theory, computing algorithms and philosophical interpretation. HGLMs can be fitted very efficiently by two interlinked generalized linear models. So we have access to a familiar repertoire of model checking techniques, and can base our choice of error distributions on the data rather than on prejudice or software limitations. Furthermore the analysis can still be carried out interactively – always a very important consideration for John.

With John’s many achievements in statistics, it is important not to forget his other interests. He shared a keen interest in gardening with his wife Mary (nee Hawkes), whom he met and married in 1955 while he was at NVRS; they have a son Jan and a daughter Rosalind. John and Mary were also keen birdwatchers, and were two of the three finders of Britain’s first Siberian Thrush (Zoothera sibirica); see Andrew, Nelder & Hawkes (1955). John was very proud of his other paper in British Birds (Nelder 1962), which gave a rigorous statistical assessment of the implausibility of the “Hastings Rarities” and provided convincing evidence for their subsequent removal from the British List. Finally he was a very keen musician and a virtuoso piano player, and his musical soirees at his house in Redbourn will be remembered by the attendees with lasting pleasure.

John received many honours during his career. He had a DSc from University of Birmingham, and received an honorary DSc. from Universite Paul Sabatier, Toulouse, in 1981. He was also elected a Fellow of the Royal Society in 1981. He was President of the International Biometric Society from 1978-1979, and was made an Honorary Life Member in 2006. He was President of the Royal Statistical Society from 1985-1986, and was awarded Guy Medals of the Society in Silver in 1977, and in Gold in 2005. He wrote three books and over 120 papers in statistical and biological journals, including two citation classics: the Nelder & Wedderburn (1972) paper on generalized linear models already mentioned, and his paper written with Roger Mead while at NVRS describing their now very widely-used adaptive simplex optimization algorithm (see Nelder & Mead 1965).

More important perhaps is his statistical legacy of general balance, generalized linear models, hierarchical general linear models – and GenStat – which will keep him always in our thoughts.

Roger Payne

References and Bibliography

Andrew, D.G., Nelder, J.A. & Hawkes, M. 1955. Siberian Thrush on the Isle of May: a new British bird. British Birds, 48, 21-25.

Lee, Y., & Nelder, J.A. (1996). Hierarchical generalized linear models (with discussion). Journal of the Royal Statistical Society, Series B, 58, 619-678.

Lee, Y., & Nelder, J.A. (2001). Hierarchical generalized linear models: a synthesis of generalised linear models, random-effect models and structured dispersions. Biometrika, 88, 987-1006.

Lee, Y. & Nelder, J.A. (2006). Double hierarchical generalized linear models (with discussion). Appl. Statist., 55, 139-185.

Lee, Y., Nelder, J.A. & Pawitan, Y., (2006). Generalized Linear Models with Random Effects: Unified Analysis via H-likelihood. CRC Press, London.

McCullagh, P. & Nelder, J.A. (1989). Generalized Linear Models (second edition). Chapman & Hall, London.

Nelder, J.A. (1962). A statistical examination of the Hastings Rarities. British Birds, 55, 283-298.

Nelder, J.A. (1965a). The analysis of randomized experiments with orthogonal block structure. I Block structure and the null analysis of variance. Proceedings of the Royal Society, Series A, 283, 147–162.

Nelder, J.A. (1965b). II Treatment structure and the general analysis of variance. Proceedings of the Royal Society, Series A, 283, 163–178.

Nelder, J.A. & Mead, R. (1965). A simplex method for function minimization. Computer Journal, 7, 303-333.

Nelder, J.A. (1991). GLIMPSE, a knowledge-based front end for GLIM. In: IMA Volume in Mathematics and its Applications 36: Computing and Graphics in Statistics (Ed. A. Buja & P.A. Tukey), pp. 125–131. New York: Springer Verlag.

Nelder, J.A. & Wedderburn, R.W.M. (1972). Generalized linear models. Journal of the Royal Statistical Society, Series A, 135, 370–384.

Payne, R. (2004). Algorithms, data structures and languages — the computational ingredients for innovative analysis. In: Methods and Models in Statistics — In Honour of Professor John Nelder, FRS. (Ed. N. Adams, M. Crowder, D.J. H and & D. Stephens), 95-118. London: Imperial College Press.

Senn, S. (2003). A Conversation with John Nelder. Statistical Science, 18, 118–131.

With the release of GenStat 13, we have taken further steps to develop and extend its functionality. The existing suite of menus for QTL analysis, of particular interest to plant researchers and breeders, has been extended to allow for the detection of QTLs in outcross populations and by association mapping. This suite allows researchers to easily access techniques for QTL analysis, without complex programming.

The continued development on our core areas has led to trellis plots of predicted means from regression and REML analysis, which will allow for prediction from any model and extends the ability for uses to visually display their data.

Other enhancements include quantile nonlinear regression, stepwise discriminant analysis to augment GenStat’s existing multivariate analysis techniques, new data mining commands for neural networks and random classification forests, and sampling for stratified surveys.

You can see the full list of new developments in GenStat 13 here.

Existing supported users can upgrade here.

Whilst any and all of the various different types of weed can cause significant damage, and most will require careful management techniques to control, one specific type of weed – an invasive weed, can have devastating effects. An invasive species of weed is one which can be non-indigenous or natural to the local environment, but has a tendency to crowd out other plants or crops. This disruption can affect the local ecosystem significantly and cause economic implications and loss to farmers who find themselves over run by invasive weeds.

And yet there is only a limited amount of information on the population dynamics and ecology for invasive plants, and the same is true for the associated soil seed bank ecology or seed persistence: hence it has been difficult to formulate well planned strategies for the management or elimination of invasive weeds. The soil seed bank ecology and seed persistence of one particular species of invasive plant however has recently been studied by Australian researchers at the Alan Fletcher Research Station in Queensland. Lantana camara is now considered one of the most invasive species of weed in tropical and subtropical areas, and has been labelled as a Weed of National Significance in Australia, hence the researchers chose to try and understand the factors affecting seed survival and seedling emergence. Historically it has been a difficult weed to manage through biological measures or through introduction of specific pests to control its spread. Additionally as its seeds are bird-dispersed it is difficult to manage through the removal of seed sources, given the sources can be wide ranging. So one possible method of controlling or eliminating Lantana camara is through depleting its soil seed bank, and in order to do this a deeper understanding of the Lantana’s seed germination and survival is needed.

A detailed study was carried out over a number of years in Queensland using controlled sites with artificial irrigation and more natural sites using rainfall for watering. Information was taken on seed size, germination responses including monitoring the depth the seed was planted and the climatic conditions, seedling emergence and seed survival as well as the different environmental conditions.

The results from the study certainly back up the concept of Lantana being a difficult weed to control. As well as the seeds being bird-dispersed and thus requiring control measures for a wide area, the study showed that seedlings germinate in a variety of conditions, although are most successful in warmer and lighter conditions, suggesting that the closer the seeds are to the surface the more likely they will germinate and are most likely to germinate during the Summer months. The Lantana seedling emergence was affected by many factors, although it was lower for the more deeply buried seeds, suggesting that forms of control through seed suppression, such as mulching could be beneficial in controlling Lantana. Alternatively the seeds could be encouraged to the surface to stimulate emergence and then treat the weeds as they emerge. Through regression techniques in GenStat the researchers discovered that seedling emergence could continue for 3 – 11 years (depending on the depth of the seed) after weed management programmes have removed the actual Lantana plants. The details and information discovered on the ecology of the Lantana suggest that to effectively control or eliminate this weed farmers and agricultural producers would need to employ on-going management techniques and intensive follow-up on the weed management programmes, as the seed survival and persistence of the Lantana appears to be far greater than first thought.

Our thanks to Dr Gabrielle Vivian-Smith and her colleagues at Alan Fletcher Research Station for their help in producing this story.

The full paper is available here.

We are covering a range of sectors, events, giving training courses and workshops and visting countries ranging from Hungary to Vietnam.

If you are attending one of the events we’ve listed and wish to meet up with us – please email Carey Biggs or telephone: +44 (0)1442 450230.

With at least 1.3 million described species of insect, accounting for at least two-thirds of known organisms, it’s no wonder that entomology fascinates the human race. And yet understanding and studying insects is not just for those with an interest in the creatures; information and knowledge on insects affects farmers and agricultural producers who may wish to be able to control the populations living and feeding on their crops. With such huge numbers of species, it’s easy to imagine the potential damage to crops, and why over the centuries humans have become fascinated with the control of these creatures, particularly in an agricultural setting.

Crop pests can take on different forms through their lifecycle, from larvae through to the final adult stage, and can cause havoc for farmers both above and below ground. The damage caused by insects living above the ground is often more easily identified and treated than the insects living below ground on the roots of plants, so crop losses from soil insects are often underestimated, and sometimes even not recognised as being caused by insects. White grubs, the larvae of chafer beetles, are one group of species for which this is true, with below-ground damage to plants often being extensive and in some cases causing the death of the plants. This complexity is further compounded by the fact that some soil pests feed on different plants at the different stages of their lifecycle, both above and below the ground. Scientists in India noted devastating effects on teak (Tectona grandis) seedlings in the nursery beds at the Ramdongari Forest Nursery in central India associated with the presence of white grubs, with the plant destruction causing severe economic effects. Little is often known about the biology and ecology of such white grubs, making them one of the toughest pests to manage, often leading to the high use of chemical insecticides, rather than Integrated Pest Management (IPM) schemes.

Teak is an important part of India’s economy by virtue of its timber quality, popularity and ever-increasing demand; it is highly prized by ship builders and manufacturers of outdoor furniture (amongst others) due to its ability to withstand most weather conditions. Combine this with the difficulty of managing white grubs, and it’s no surprise that Dr Nitin Kulkarni from the Tropical Forest Research Institute in Jabalpur, India chose to investigate the effect of the larvae of the chafer beetles Holotrichia rustica and H. mucida on teak seedlings. The area studied (Ramdongari Forest Nursery) accounts for around 500,000 to 800,000 teak seedlings per year and supplies seedlings for state forest departments and non-governmental organisations, including small farmers, across central India, so is of significant economic importance to the region.

The researchers observed and recorded data on the adult beetle emergence, their host preferences and populations, as well as the grub populations and damage to the teak seedlings. They found that typically the seedlings wilt due to the roots having been eaten, and often die. The insect count and plant damage data collected over the four years of the study were analysed in GenStat Discovery Edition 3, the free version of GenStat available to the developing world, using analysis of variance followed by the multiple comparison of means. Dr Kulkarni said “GenStat covers most of the statistical analyses and yet, is very user friendly with appropriate demonstrations.”

It was noted that the adult beetles emerged from the teak beds and became active at the onset of the monsoon season, when humidity had risen rapidly after pre-monsoon showers, and remained active for approximately 18 days. Importantly, they observed that whilst the adult beetles fed and mated on the leaves of jujube and other Ziziphus species, immature grubs of both species fed on the fine rootlets and main roots of teak seedlings. It has therefore led to the suggestion that it may be possible to control this pest with an IPM approach by targeting adults rather than applying chemicals against the larvae. With adult activity triggered by measurable weather conditions – falling temperatures and increased humidity – manual collection of the adults on the shrubs growing on nursery field edges appears one viable option for managing the pests without use of toxic chemicals, as part of an IPM program.

Information and results from research such as this has wider reaching impacts for entomologists looking at other white grub species, many of which are pests of food crops across the developing world. As well, the research progresses research into effective pest-management techniques, in this case on the chafer beetles in Indian teak nurseries.

Imagine needing to conduct research of this nature but with limited access to the appropriate tools, specifically appropriate data analysis software. Unfortunately researchers and scientists in the developing world don’t always have access to the tools they need for their work; however data analysis software is something they do have access to. With GenStat Discovery edition 3 these researchers were able to use high quality, respected and appropriate software to conduct their analysis, and so provide farmers – and in this case, teak nurseries – with relevant advice and assistance on the cultivation and protection of their crops.

GenStat Discovery edition 3, is widely used across the developing world to assist in analysing and visualising data. It is free to non commercial establishments across the developing world. For more details or to download look at the VSNi website.

Our thanks to Dr Rogers of Research Connections and Consulting and to Dr Kulkarni of the Tropical Forest Research Institute. Dr Rogers provides pesticide-efficacy data analysis services to the crop-protection industry and also assists authors with the preparation and submission of scientific publications.

The original paper (abstract) is available from Wiley Interscience

Technical tip – registration of GenStat Discovery

The GenStat Discovery licence is valid for one calendar year, but a new licence key will be automatically sent to you in December, irrespective of when you originally downloaded it if you register, meaning you never need be without it, nor do you need to download another version, at the end of the year.

Users who have obtained their copy of GenStat Discovery through an old or a colleague’s CD, may find they need to request a new licence key from support, if the GenStat Discovery CD was made prior to the current year.

To register your copy of GenStat Discovery visit the registration webpages.

Latest training courses

VSNi run training courses throughout the year on a variety of topics, and we are always updating the schedule. Our next courses on GenStat are scheduled for April 2010.

Introduction to GenStat will take place on Monday 19th April 2010 at Apsley Paper Trail in Apsley, Hertfordshire, UK. This course is aimed for those who need an overview of GenStat and the sklls to start them analysing and visualising data.

For users who are confident with the menu system and wish to develop their skills in programming in GenStat, the Introduction to the GenStat Command Language is being held on Tuesday 20th April 2010, again at the Apsley Paper Trail.

For more details or to book your place on either of these courses please register on line, or email support.

Courses at the Statistical Services Centre, University of Reading

We’re pleased to tell you of additional courses in statistics run by colleagues at the Statistical Services Centre at the University of Reading; all of which allow time for practicals for the attendee to put into practise what they have learnt:

E-SMS – a popular online course about understanding variability (next run scheduled 22nd March-10 May). It includes the essentials needed to describe data well (descriptive statistics) and to make good generalisations (statistical inference). It also covers the concept and uses of p-values, and establishes the groundwork for statistical modelling based on simple linear regression. Find out more at e-SMS website.

Review of Basic Statistics (19-20 April) – this course covers the basic ideas of statistcs needed in research systematically and quickly. Mathematical details are kept to a minimum, but concepts such as hypothesis tests, t tests, different distributions are all covered.

Regression Analysis: A Hands-on Approach (21-22 July) – Regression is a powerful technique for studying relationships between quantitative variables. Summarising relationships by the most appropriate equation (modelling) is very quick using Genstat. Find out more about the details on this course.

Introduction to ANOVA – (21-22 April) is a key technique popularly used in analysing research data. Every serious researcher needs to acquire skills in using this technique, and to have a good appreciation of the underlying concepts.

What Sample Size do I need? (1 December) Choosing an appropriate sample size is a common problem and should be given due consideration in any research proposal, as an inadequate sample size invariably leads to wasted resources. This course aims to give a practical introduction to sample size determination in the context of some commonly used significance tests.

More details on these courses and the Statistical Services Centre is available at the SSC website.

Out and about with VSNi

From 11-16 July 2010, VSNi’s CEO, Stewart Andrews will be presenting GenStat for Teaching at ICOTS8 in Slovenia. ICOTS8 in Ljubljana in Slovenia.

If you would like to meet with Stewart at one of this event please email Stewart Andrews to arrange the details.

We’re always updating the list of events we can support and sponsor – so please send us details of any events you are organising or involved in, and as we decide on more events for the future we’ll list them on our website.

Statistical analysis software designed for Bioscientists, GenStat has been a fundamental part of Science at UKZN for many years. It is ideally suited for agricultural research initiatives and has features specifically designed to assist plant researchers and breeders.

According to UKZN Statistician, Dr Henry Mwambi, it is imperative to increase capacity in Biometry in KwaZulu-Natal and in the sub-Saharan region. He said the University was fortunate to secure Dr Payne who has a busy schedule and is in demand for his expertise and experience with GenStat.

Read more here.

Thankfully neither our livelihood or our next meal was dependent on the success of our vegetable crop, but seeing the speed and extent of the destruction caused by these caterpillars made me realise the very real problem that these creatures can cause farmers, whose livelihood does depend on the success of their crops. Without adequate pest control measures farmers can face either crop destruction, or tough decisions regarding the application of pesticides to their crops.  In our family instance our total crop (admittedly only around 10 Brassicas) was wiped out by the caterpillars, but not all pest infestations will have such a dramatic effect. In some instances plants can tolerate small “injuries” (physical harm by a pest) but no “damage” (monetary value lost as a result of pest injury, such as spoilage, or yield reduction etc), or will channel more energy into growth or fruiting. Therefore the effects of some pest populations may not cause enough of a loss to justify the time or expense of pest control operations. So the question is, at what point do farmers decide to employ some form of pest control measures?

Entomologists refer to a break-even point, or economic injury level (EIL) where the pest populations density is at a level where the cost to control it equals the amount of damage it inflicts. The EIL is usually expressed as the number of insects per unit area, or sampling unit. Often, due to the lag time in the implementation of a pest control strategy and its effects, pest control measures need to be put into place before a pest population reaches outbreak status: this moment is below the EIL and called the economic threshold. The economic threshold is the decision point for action to ensure the pest population does not reach the EIL.

Recent research carried out by Dr Rogers and Dr Brier at the Queensland Department of Primary Industries and Fisheries, in Australia, has looked at the effect of the cotton bollworm (Helicoverpa armigera) on soybean and dry bean.  The cotton bollworm is a major pest on both crops in Australia, Asia, India and Africa. However there are limited details on the EIL’s or economic thresholds for these plants. In the past, attempts to develop economic thresholds have used artificial defoliation or depodding techniques which may not accurately simulate the effect of the caterpillars’ eating, and hence lead to inaccurate results for EILs. This latest research has shown that the cotton bollworm does more than just eat the leaves; it eats the leaf veins, petioles, apical growing points and hence limits the plant’s ability to compensate for the damage caused by the caterpillar: this means that previous estimates for the economic threshold from research using artificial defoliation and depodding are artificially higher. Two papers were published in Crop Protection 29, detailing studies on pest damage relationships for the cotton bollworm on vegetative state (Crop Protection 20, 39-46) and on soybean and dry bean during pod fill (Crop Protection 29, 47-57).  The experiments were carried out at the Queensland DPI research station at Kingaroy, Queensland over 3 years. The collected data were analysed in GenStat using regression analysis; simple linear, multiple or split-line regressions were fitted depending on the relationship and data available.

“GenStat provided us with the security of knowing our statistical analysis was based on sound and reliable software,” says Dr Rogers, “We especially appreciated the support provided by GenStat staff.  They were able to provide us with an extension of the R2LINES procedure (a GenStat procedure that fits a model of 2 straight line segments to the data) that enabled us to get the most out of our data.  This was especially important to us as the data collection process in these trials was both painful and tedious – we spent months each year on our hands and knees in the dirt counting caterpillars so we wanted to get as much out of our data as we could!  With the enhanced R2LINES procedure we were able to model the variation in the data and produce parameter estimates that were biologically meaningful, both in terms of yield effects and how and when the plant’s compensatory responses cut in and for how long.”

The results of the studies have highlighted discrepancies with other studies where artificial defoliation has been used to mimic the behaviour of the caterpillar; the yield loss figures from Rogers’ and Brier’s studies are considerably lower than previous studies suggesting that the yield loss figures from previous studies are misleading. These studies have provided farmers with more accurate details on the effect of these caterpillars on soybean yield meaning more effective pest control measures can be employed.

Our thanks to Dr Rogers for his help in this piece.

Dr Rogers provides pesticide-efficacy data analysis services to the crop-protection industry and also assists authors with the preparation and submission of scientific publications.

Links to papers (abstracts):

Pest-damage relationships for Helicoverpa armigera (Hübner) (Lepidoptera: Noctuidae) on vegetative soybean

Pest-damage relationships for Helicoverpa armigera (Hübner) (Lepidoptera: Noctuidae) on soybean (Glycine max) and dry bean (Phaseolus vulgaris) during pod-fill

Any farmer will tell you of the importance of vaccination in controlling disease, and this is no different for fish farmers, or specifically Atlantic salmon farmers.  Vaccinations against diseases such as furunculosis, vibriosis, cold-water vibriosis, winter ulcer and infectious pancreatic necrosis, which could otherwise decimate a farm, are vital in protecting the fish and farms and controlling these diseases in farmed Atlantic salmon.

However, vaccinations can cause reactions at the injection site such as adhesions and discoloration, reduced appetite and growth, and in severe cases a downgrading of the carcass at market. Not only can this have an economic impact on the salmon farms, but also may create a negative perception of the farms with the general public. As with all aspects of farming the public is concerned not just with the food in question, but also of the welfare of the production animals, which means that breeding companies need to be sure of the actual physical health and safety of their fish.

As with many aspects of farming and agriculture, all aspects of aquaculture are the subject of scientific research to enable better understanding and improved planning for the future, including prevention of disease and improved fish health. The results from challenge tests have indicated substantial genetic variation in disease resistance to a variety of bacterial and viral diseases (including some of those listed earlier), significant but relatively low genetic correlations among the different diseases and a high genetic correlation between the survival of non-vaccinated pre-smolts (smolts are young salmon as they migrate from freshwater to the sea) and the survival of older non-vaccinated full and half sibs under natural break-outs of the diseases. With this in mind Atlantic salmon breeding companies are looking at both the development of efficient vaccinations and the breeding of fish with higher disease resistance.

The results from the challenge test research may suggest that genetic variation in immune systems might lead to genetic variation in innate disease resistance and variation in reactions to the vaccinations. One question researchers in this area need an answer to is whether there is significant genetic correlation between vaccine reactions and innate resistance, as this could mean a change of vaccine dose or a breeding programme to produce fish with the lowest reaction rate to vaccines and the highest natural resistance to diseases.

And it is these issues that a recent project, undertaken by researchers from Nofima Marin, the Norwegian School of Veterinary Science and Aqua Gen AS and funded by the Research Council of Norway (published in the Elsevier journal Aquaculture 287 (2009) 52-58) has looked at.

Atlantic salmon were taken from farmed strains and from offspring of wild salmon, and reared at different farms. Data was collected on sex, size (at different stages), adhesion scores and melanin scores of fish at harvest size and survival data from challenge tests of their full- and halfsibs. The data collected was analysed using ASReml to fit a multivariate linear mixed animal model to obtain estimates for the variance and covariance components for the random effects of the studied traits. A single trait sire and dam model, which included the same fixed effects was used to obtain across year-class heritability estimates (using an alpha version of ASReml 3).

Results from the project indicate that the severity of vaccine injuries in Atlantic salmon could be reduced through selective breeding. Additionally the relatively high correlation that was found between adhesion scores and melanin scores show that there is unlikely to be a need to select against both of these traits. The researchers warned that including an additional breeding trait objective would reduce the genetic gain for other traits, and therefore recommended that other solutions to the vaccine injuries be found, such as improved vaccines or vaccination procedures. The results also indicated a negative genetic correlation between body weight and adhesion or melanin score; however, these estimates are most likely biased towards the negative end of the scale, since vaccine injuries are expected to reduce growth rate for non-genetic reasons.  The study also indicated “no genetic association between resistance to furunuculosis and ISA (infectious salmon anaemia) and vaccine injuries”, which suggest that the vaccine injuries are caused by other aspects of the immune system, not associated with the genetic variation in innate resistance to these diseases.

Research of this kind is extremely important to the aquaculture industry to enable it to continue to develop, expand and improve, not just the industry, but the procedures, processes and systems associated with it. ASReml is widely respected and used in situations such as this; the recently released ASReml 3 provides users with even greater ability to analyse large datasets or complex models, quickly and efficiently.

Our thanks to Dr Bjarne Gjerde of Nofima Marin for his help with this story.

And there can be no doubt of the importance of such programmes to the Mongolian Przewalski horse, which for around 30 years was bred in captivity under a project led by the Institute of Zoology at the Zoological Society London, and has been successfully reintroduced into the wild. Without the captive breeding programme this wild horse would have become extinct, indeed for many years was classed as such.

Since the 1970’s captive breeding programmes within zoos have become more important as wildlife populations decreased and the shipping of endangered animals became more restricted. Zoos needed to be able to successfully breed their own animals to maintain animal stocks, leading to the creation of “studbooks” listing information on the captive species and enabling zoos to recommend pairings and help maintain genetic diversity by avoiding “in-breeding”. This data can be very detailed, with birth, sex, parentage, source and the animal’s movement between institutions, as well as size and weight information.

Despite the increasing importance of captive breeding programmes to manage zoo stocks and animal conservation, very few studies have been carried out to understand selection and adaption to captivity. Admittedly quantitative genetic studies can be labour intensive which may help to explain this, however a recent study published in “Trends in Ecology and Evolution” suggests that the use of zoological records obtained from captive bred animals (such as data found in the “studbooks”) may be useful in studying selection and adaptation to captivity. The researchers concentrated on using a statistical technique known as the “animal model” -  a mixed effects model and a form of linear regression, where the explanatory terms are both fixed and random.

The animal model is already widely used in studies for the genetic improvement of livestock, and is increasingly being used in evolutionary studies, so conservation is the next logical step.

What is the animal model?

The animal model allows a phenotypic trait to be split into its additive genetic and other non-genetic variance components; it requires data that already exists or is easy to obtain for captive populations, for example information on the pedigree and phenotypic data on quantitative traits, such as body weight. Additional factors which highlight the suitability of this approach include the ability of the animal model to analyse unbalanced datasets, common with captive populations due to missing phenotypic data or pedigree links.  Using the data in existence and appropriate statistical tools could lead to a far greater understanding of the effects of captivity on the genetic variation of different traits. Although there are a host of other issues of interest to evolutionary biologists and conservationists, in terms of a captive breeding programme the heritability of reproductive traits, such as litter size or birth mass would be of vital importance.

As an example the researchers looked at the heritability estimate for the litter size in Sumatran tigers. Using ASReml, they were able to deduce that environmental factors appeared not to play a part in litter size; hence if female tigers producing large litters are bred by zoos, then it’s likely that large litters will continue to be produced in captivity, because this is a heritable trait.

To take this to the next step and achieve the goals of the reintroduction of endangered species to the wild, scientists need to be able to understand which traits are beneficial to both captive and wild populations. The use of existing data and statistics means that we could be closer to this. Certainly the researchers behind this study suggest that fine-tuning the data collection for studbooks could assist with very detailed analyses and understanding of genetic traits within captive populations, leading to more successful captive breeding programmes.

ASReml is a highly respected animal model software package and is perfectly suited to researchers working in genetics and animal breeding; it is already widely used in animal breeding programmes, due to its flexibility, speed and accuracy. As well as being specifically designed for the fitting of linear mixed models using the Residual Maximum Likelihood (REML), it is ideally suited to large datasets with complex variance methods. Please email VSNi with your name and organisation details quoting reference ZOOCB09 to obtain a free 60 day trial.

Our thanks to the researchers, especially Fanie Pelletier for their help in this piece. For more information on their research see the pages at the Université de Sherbrooke.

The original paper can be accessed from Trends in Ecology & Evolution, Volume 24, Issue 5, 263-270, 06 March 2009.

Already widely used by scientists and researchers in the biosciences the new procedures and functionality added at version 3 will further extend ASReml’s ability to analyse large datasets and complex statistical models. Since its original launch ASReml has enabled scientists to investigate and understand relevant factors in breeding and genetics leading to for example, the identification of superior grain varieties for farmers in Australia (based on the National Variety Trials system) and generating herd heritability estimates (based on research at the Pennsylvania State University).

For full details see the main ASReml page.