2ND ANNOUNCEMENT : WORKSHOP ON MULTIVARIATE STATISTICAL ANALYSIS IN ECOTOXICOLOGY

[Tim Kedwards]

Please find attached below finalised details of the forthcoming workshop. If you require any further information feel free 
to contact me directly.

Regards,

Tim
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dr Tim Kedwards
ZENECA Agrochemicals
Ecological Risk Assessment Section
Jealott's Hill Research Station
Bracknell, UK
RG42 6ET

Tel:            +44 (0) 1344 414107
Fax:            +44 (0) 1344 414124
Email:          Tim.T.J.Kedwards at gbjha.zeneca.com
Disclaimer:-    'The opinions expressed herein are my personal opinions
                and do not necessarily represent those of my employer'

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

A WORKSHOP ON MULTIVARIATE STATISTICAL ANALYSIS IN ECOTOXICOLOGY

Saturday, 16 November 1996, 8:30 - 5:30, Washington Hilton and Towers
Presented in conjunction with the SETAC 17th Annual Meeting, 
Washington, D.C.

Introduction

Frameworks for ecological risk assessment often culminate in the evaluation of effects under simulated or actual 
environmental conditions.  A fundamental objective of such semi-field and field studies is to investigate effects at the 
community and ecosystem level.  To this end, data are collected on a large number of ecological variables. Study types 
include manipulative experiments (artificial streams, mesocosms, terrestrial field trials), environmental monitoring 
(before/after, upstream/downstream), and ecological surveys (plants, bird census).

Data from these studies are usually evaluated by using univariate statistics (e.g. analysis of variance, 
concentration-response regression).  These techniques have limitations because it is only possible to look at a small number 
of variables within the ecosystem at any one time, limiting effects determination to the population level, whereas the 
objective is normally to understand effects at higher levels of organisation.

The advent of greater computing power and recent developments of new multivariate statistical tools has made available a 
variety of techniques (e.g. CANOCO, RIFFLE, PRIMER) with the potential to overcome some of these previous limitations.  
Consequently, there is growing interest among ecotoxicologists in applying such techniques to field studies in order to be 
able to generate true community and ecosystem endpoints.  This workshop will provide an excellent opportunity for interested 
individuals to meet and exchange views on these approaches and allow developers of the statistical approaches to discuss the 
development and application of the techniques.  

The objective of this workshop is to provide information on and  generate discussion of multivariate statistical techniques 
in ecotoxicological field studies, focusing on the application of techniques for generating community and ecosystem level 
statistics, rather than mathematical derivation.

Attendance will be limited to 100 people to maximise information exchange. 

Registration is first-come, first served.  The registration fee is $225, and forms can be obtained from the SETAC Office.

Workshop proceedings will be provided to all registrants at no additional cost, and will be available for purchase by others 
after the workshop.

SETAC Office, 1010 North 12th Avenue, Pensacola, Florida  32501
T +1 904-469-1500    F +1  904-469-9778    
E setac at setac.org,  http://www.setac.org
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
FORMAT OF WORKSHOP

The format of the workshop has now been finalised.  It will be in two Sessions, the first comprising three presentations 
addressing the current situation concerning the analysis of ecotoxicological field studies. Professor Jim Kennedy, from the 
University of North Texas, will be speaking on designed experiments and the relative strengths and weaknesses of current 
analytical procedures. This will followed by a presentation from Dr Mike Harrass, AMOCO Corporation, outlining monitoring 
and survey work in ecotoxicology. The last presentation of the session will be given by Dr Tony Maciorowski, United States 
Environmental Protection Agency, and will provide a regulatory perspective of ecotoxicological field studies. 

The second Session aims to address the multivariate statistical methods currently used in ecotoxicology.  It will also 
include a review of new and novel methods currently being used elsewhere and which may prove beneficial in the 
ecotoxicological arena. The first presentation will be given by Tim Sparks from the Institute of Terrestrial Ecology in the 
UK.  Tim will review traditional multivariate methods and will give examples of their uses in ecology and ecotoxicology. The 
second presentation will be given by Professor Geoff Matthews (University of West Washington) addressing, amongst others, 
the non-metric clustering program RIFFLE which has successfully been used in ecotoxicological studies. This will be followed 
by a talk by Mr Paul van den Brink (SC-DLO, Netherlands) outlining the recent application of the CANOCO package to 
freshwater field testing. Dr Bob Clarke (Plymouth Marine Laboratory, UK) will present the application of the PRIMER package 
to similar studies. These latter three talks form the centre of current multivariate statistics expertise in the 
ecotoxicology. After a suitable discussion period the final talk will be presented by Professor Wojtek Krzanoswki 
(University of Exeter, UK), who will talk about new and novel methods of potential value to ecotoxicology but not yet 
implemented.