home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Business Shareware
/
Business Shareware.iso
/
start
/
edu
/
plntgro1
/
readme.2nd
< prev
next >
Wrap
Text File
|
1992-07-20
|
26KB
|
514 lines
INTRODUCTION
During the late 1800's, large numbers of people
gained the ability to read and began to have money in
their pockets. This made it possible for companies to
produce books in large numbers, and amongst those books
of course were ones about plants.
That change brought about a revolution in the
distribution of information about plants. Currently we
are near the beginning of another revolution in the use
of plant data. This will arise from the new literacy -
i.e. computer literacy - and the fact that large numbers
of people now have access to computers.
Until recently, most computerised information about
plants has been held by institutions and corporations.
PLANTGRO provides a system where the general public can
have data too and can make further data-files for the
package.
What PLANTGRO does is to provide a way of recording
the environmental likes and dislikes of hundreds of
plants. It then stores that information so that it can
be used by a large number of people to predict the growth
of those plants.
This has been made possible by using methods of
describing plants' responses to environmental factors
which are natural to human beings. The methods are
supported in the package by scientific concepts and
procedures but untrained people can understand what is
happening and can see whether the results are making
sense or not. PLANTGRO therefore enables both well-
trained and little-trained people to record their
knowledge and share in the use of it.
No package like this has been produced before.
Databases exist which allow one to sieve information
about plants, and large models exist which predict growth
by simulating the effects of factors on physiological
processes such as photosynthesis, but there has been no
general prediction system which works in the way the
human mind works. Since human beings have been using
natural methods for thousands of years and have a vast
collective knowledge of plants, PLANTGRO should make it
possible to document the environmental relationships of
at least a thousand plants and make that information
available worldwide for routine purposes.
To help that spread of data come about, the programs
and data-files supplied with PLANTGRO have been made as
simple as possible, which has enabled us to allow all the
software included in the package to be freely copied so long
as the source is acknowledged and the copying is not done
for profit. The handbook is being held under copyright
to maintain public trust in our advice and to help CSIRO
recover some of the costs involved in producing PLANTGRO,
but to further promote collaboration, purchasers of the
handbook will be entitled to a year's free membership of
the PLANTGRO network. This will bring users a newsletter
carrying information about new files and programs which
have been produced, and it will help them find other
users who have similar interests.
Provided here are all the PLANTGRO programs in
compiled form, nearly all of the 130 starter datafiles
provided in the full package, and enough guidance to help
you see how PLANTGRO works and what it is capable of.
The full package provides the source code (GWBASIC), all
the data-files, the 240-page handbook, and Eric Meyer's
VDE word processor, which can be useful for editing data-
files. The handbook is a key component because it
provides the expert systems for recording personal
knowledge of plants' responses to environmental factors.
The handbook has been written in a friendly style and has
a skill rating system to help newcomers to science to
find their feet and feel valued.
The programs are available too in QuickBASIC. For
all information about versions, prices, etc., please
see `Ordering' below.
If you would like to have a trial run now with
PLANTGRO, please jump to the `Operations' chapter. If
you would like to read first about the components of the
PLANTGRO package and their capabilities, please move now
to `Description of PLANTGRO', which follows directly
below. The file called README.PG is the routine update
file which tells users of developments with the package
since mid-1991.
DESCRIPTION OF PLANTGRO
Describing soils, climates, and plants
PLANTGRO deals with 22 environmental factors
including light, nutrients, salinity, temperature, water
supply, waterlogging, and wind. Effects of any one of
these factors can also be linked to plant quality - e.g.
the taste of a fruit.
There is a program each for making soil, climate, and
plant files. The soil program asks for routine soil
survey data, but if you can't obtain such data, you can
use the option which accepts description of the soil in
everyday language. The program then makes an intelligent
guess at what the data would probably be.
The climate program also asks mainly for routine
data. Data on solar radiation can be hard to find and
there is a help system for this work which enables
guesses to be made by analogy with data-sets for well-
known places. Note that the program will accept data
based on weekly, tenday, or monthly time-units and it can
use actual or average data. However, for all types of
data, the number of time periods handled is limited at
present to twelve. This is not as bad a limitation as it
sounds because the prediction programs allow use of the
monthly average data to wrap around at the end of the
year. You can even use actual monthly data in this way
by entering them as average data, but be very careful how
you interpret the results if you do that.
The bulk of the plant data stored is made up of what
are called notional relationships. These take the form
of curves or straight lines which describe how the plant
is thought to respond to particular factors. For each
factor, the handbook provides Figures showing options to
choose from for that factor. The particular relationship
chosen is then entered by its number in the Figure and
the program records the formal details automatically.
To help the user interact with this system very
easily, the relationships are made up from straight-line
segments, e.g. a curve which first goes up, then across,
and then down may be composed of a rising straight
segment, a level segment, and then a falling straight
segment. What is recorded in the plant file is merely
the x,y co-ordinates of the ends of each segment. Thus
if you find that a relationship is producing errors, you
can change it easily by altering the x,y values. You
don't have to know any maths to make such changes: all
you need is a feel for how the plant actually responds
and an understanding of the units used for the x-axis.
Because little is known about many plants of interest
and because units such as tonnes per hectare are
meaningless for many of them, the y-axis in these
notional relationships is just a simple rating system
from 0 to 9. These are called suitability ratings.
The handbook describes seven starting points for
making a plant file. One of these is scientific data,
another is data on the range of conditions the plant
experiences in the field, and the rest start with
different types of personal knowledge of about the
plant's responses to factors in the environment.
As you can imagine, you will need the handbook to
make a real plant file, but you can run the program on a
trial basis by entering 1 virtually everywhere. That way
you'll get a feel for what is asked for. For example,
every time you are asked for an option number, the
program is referring to the sets of options offered in
the handbook for that factor.
(Note, by the way, that when the program asks whether
you want to make a general file or a phasic one, it is
referring to the fact that if a plant is quite well
understood, you can make a file for each phase of its
development - e.g. establishment, leaf growth, flowering,
and maturity. This way PLANTGRO can give more reliable
predictions of the plant's performance in particular
locations.)
Making predictions
As for making predictions of performance, there are
three programs to choose from here. The first - called
GRO.BAS - is the workhorse program which handles just one
prediction at a time, using either a general or a phasic
plant file. The second - called MULTIGRO.BAS - accepts
orders for predictions for combinations of soil, climate
and general plant files - e.g. 3 soils x 2 climates x 3
plants.
The third - called PHASEGRO.BAS - deals only with
phasic plant files and will estimate performance across
all the phases described for the plant chosen (note that
unlike with MULTIGRO.BAS, only one run at a time is
permitted at present with this program). To see what
happens, for example, you can use PHASEGRO.BAS with the
set of files for spring wheat provided here and predict
the crop's growth in Sydney, the Galapagos Islands, Rio
de Janeiro, Hong Kong, or anywhere else that takes your
fancy. (If you do follow this suggestion, make sure that
you have at least 7 pages of paper available in your
printer before you start.)
Note that with all the prediction programs, the main
results from the predictions are given in terms of
limitation ratings from 0 to 9, where 0 = no limitation,
8 = slow death and 9 = rapid death. These ratings are
the direct inverse of the suitability ratings used to
make the plant files. At first sight it seems silly to
make this changeover when producing predictions, but this
is what human beings usually do. They describe plants'
responses in positive terms, but when they predict the
plants' performance, they look for the biggest problem
because that largely determines the overall yield. Since
PLANTGRO uses that method too, it made sense to use 9 to
indicate the biggest problem rather than suitability
rating zero.
Brief specification
PLANTGRO has been successfully operated using MS-DOS
versions from 3.2 to 5.0 and may run on versions less
than 3.2. As indicated earlier, the language used is
GWBASIC (this is because computers have often been
delivered over the years with `free' copies of this
language), and total access is given to the GWBASIC programs
and data-files. This means that users can improve data-
files, modify the programs, or make new versions for
special purposes such as teaching. Whatever work of this
kind is done can be freely sold. There is no licencing
whatever on the GWBASIC software CSIRO has provided in the
package. Upcoming versions for Macintosh and Windows will
have to be covered by licences to ensure proper returns for
the company investing in their production.
The handbook is composed of two `books'and eight
appendices. The first book - called Getting Started -
has 22 chapters and introduces the concepts and basic
methods. The second book - Help with Making Plant Files
- has four chapters. All chapters and appendices are
given a skill rating in the range 1 to 5, where 1 is for
people who have a strong feel for plants but have little
or no formal training in plant science or computing and 5
is for people who are able to change the programming and
the science in the package.
As a further encouragement to shy users, an effort
has been made throughout the handbook to convey the
feeling that a real human being is talking to them. Most
people have had the `shakes' at some time in their life
with new methods, and users are shown that the author
knows what it feels like.
To help users to get a feel for the way the package
works, starter data-files are provided for 30 generalised
soil types, 40 climates, c. 30 crops, and c. 30 timber
trees. Among the crops are banana, cashew, cassava,
cocoa, coffee, kenaf, maize, potato, soybean, sugarcane,
sweet potato, taro, wheat and yam. Among the trees are
several Acacia species and some tropical hardwoods.
Networking
As indicated above, one aim behind PLANTGRO is to
encourage collaboration and sharing of ideas and data.
To help this develop, a network will be established,
first on a paper basis and later via email also.
Purchasers of the handbook who return the card at the
back of the book will be given free membership for a
year, which will bring them a newsletter containing
information about new programs and data-files and
discussion about developments with PLANTGRO. They will
also be helped, if requested, to find other users with
interests similar to their own. Those who have not
bought the handbook will be able to join for $A25 in
Australia and $US25 elsewhere.
OPERATIONS
Preliminaries
These notes assume that you have pulled down PLANTGRO
onto a hard disk. If you have acquired it somehow on a
1.4 Mb diskette, please be sure to make a backup copy
before proceeding. Also be aware that you may quickly
run out of space on the diskette as you work with
PLANTGRO.
Before starting work with PLANTGRO for the first
time, ensure that PLANTGRO will have access to the DOS
SORT.EXE program. If it doesn't, modify the PATH
statement in your AUTOEXEC.BAT file. If you don't know
how to do this, get help. Don't play with that file
without knowing what you are doing.
To call up PLANTGRO, simply type in PGRO. This will
produce an opening screen which asks you a yes/no
question. The question basically says "Do you want fresh
alphabetical lists made of the data-files currently
accessible to PLANTGRO?". This question is important
for the first run and also later if you bring in files
made by other people. So, say `yes' the first time
around.
If PLANTGRO has successfully linked up with your DOS
SORT.EXE program, you will see indications on the screen
that the alphabetical lists are being made. When
finished, the package moves to the main menu.
Newcomers
For newcomers to predicting plant growth on
computers, probably the best way to start is to choose to
make a soil file. Then, when that program comes up,
choose the verbal method, which uses everyday language.
Then think of a soil you know - perhaps that in your
backyard - and answer the questions on the screen as best
you can. The set of questions at the end are for
background information and are not compulsory. Blanks
may be entered there if you need to. (By the way, it is
a good idea to make the soil's formal file name start
with the same letters as in the soil's general name which
you entered. This will ensure easy selection of your
file later on when you want to grow a plant in it. You
can even use the same name if it is short enough.)
When you've finished entering your soil description,
PLANTGRO will put your soil's filename into the list of
soil filenames which was produced when you entered BEGIN
and said `yes'. It will insert it in its correct
alphabetical position according to the filename. This
means that your soil will be offered to you on the screen
and you can therefore make a prediction for that soil.
Now use the main menu to say you want to make a
prediction. Then choose the first type of prediction
method offered, i.e. no. 1. (At this stage, make sure
that your printer is connected and that it is set to
produce compressed print. Don't select the print-to-file
option you'll be offered later because the file overflows
the screen and you won't understand it like that.)
When asked, select a `general' plant file to give you
lots of plants to choose from to grow in your soil. Then
select the soil file you made and then choose average
monthly climate data. When you've chosen a climate,
PLANTGRO will interface the plant with the soil and
climate to see how the plant might grow in those
conditions.
From there on it is up to you. You can stay with the
same program and choose a different plant or a different
soil or climate, or you can go back to the main menu and
do something from there. If, however, you skipped the
preceding chapter which gave an overview of PLANTGRO and
how it works, please go back and read it fairly soon.
You will then understand a lot more about what you have
seen on the screen and on your printer.
Experts
If you are familiar with computers and plant science,
your biggest problem may be a feeling of distaste due to
the simplicity of PLANTGRO, its use of semi-quantitative
data, and the lack of reference to physiological
processes. This is understandable but I ask you to
postpone judgement until you know more about PLANTGRO's
foundations and its uses.
Formal explanation of the basis of PLANTGRO is given
in the publications listed below under `Background
Information'. A simple analogy with what those
publications say can be obtained by thinking about an
atlas of your country. Since childhood, you have been
used to seeing maps of different scale in that atlas, and
I doubt that you have ever criticised the coarser-scale maps
for not showing the detail in the finer-scale maps. You
have instead accepted instinctively that the coarse maps
cover a lot of ground and are good for showing broad
features and that the finer maps cover a small amount of
ground and are good for showing detail. All are useful,
and each complements the other.
Likewise, your method of predicting plant growth is
part of a hierarchy of methods of predicting plant
growth. PLANTGRO is a step coarser than yours and
because of that, it can cover lots of plants and lots of
factors, but to do that it has had to give up detail.
To say, therefore, that a prediction system at one scale
is better than a prediction system at another is like
saying that a district map in an atlas is better than the
national map. Each can be bad or good as judged by the
standards for its scale, but they cannot be directly
compared except in terms of qualities such as the use of
colours etc. So, please think about these things when
looking at PLANTGRO.
Finally, if you wonder where the magic is in
PLANTGRO, you won't find much in the software, and as an
expert, you may not find much in the handbook. Where it
really is is in the minds of users who have found that
PLANTGRO helps them to do things which they have wanted
to do for many years. That warms their hearts, and for
them to find too that they can use the package to record
and mobilise their knowledge of plants and then
distribute it, makes them happier still. To see eyes
light up when those discoveries are made is a great
feeling.
To you as a person experienced in predicting plant
growth, there is little more to say. All you have to do
is to use the menu system and see what emerges. I hope
that after some exploration of the package and some
thought about what it offers, you will agree that it is
not a poorman's version of what you do but something
which substantially expands the range of prediction
methods available to scientists and the general public.
Perhaps it may even prove helpful to you in a way you
didn't appreciate at first sight.
BACKGROUND READING
Hackett, C. (1988) Matching Plants and Land:
Development of a General Broadscale System from a Crop
Project for Papua New Guinea. CSIRO Div. of Water and
Land Resources, Natural Resources Series No. 11,
Melbourne, c. 80 pp. (Available from author's
address, $A 16 plus p&p)
Hackett, C. (1991a) Mobilising environmental information
about lesser-known plants: the value of two neglected
levels of description. Agroforestry Systems 14, 131-
43.
Hackett, C. (1991b) PLANTGRO: A Software Package for
Coarse Prediction of Plant Growth. CSIRO, Melbourne,
242 pp + disks.
ORDERING
Copies of the complete PLANTGRO package, the handbook
alone, the original disks alone, and an upgrade disk
carrying the programs in QuickBASIC without data-files
can be obtained from:
CSIRO Publications, PO Box 89, East Melbourne, Vic. 3002,
Australia. Ph 61-3-418-7217. Fax 61-3-419-0459.
Mastercard and Visa are accepted, plus Australian
Bankcard.
The prices for delivery within Australia (including
postage and packing) are:
Full package, GW version (3.5" and 5.25" disks) $A 100
Handbook only $A 65
Original disks, including data-files $A 40
Upgrade disk 1 (programs for QuickBASIC) $A 25
For delivery outside Australia, the prices are
numerically as above but in $US. The difference covers
air mail and currency charges.
The source of this file is:
Dr Clive Hackett, CSIRO, 306 Carmody Road, St Lucia, Qld
4067, Australia. Ph 61-7-377-0209 Fax 61-7-371-3946
