INQUA Working Group on Data-Handling Methods

Newsletter 8: July 1992

TURN YOUR EXPENSIVE OLD PC INTO A DUMB POLLEN COUNTER

Louis Maher

Back in the late 1970's I recorded pollen identifications on a paper-roll adding machine as a way of recording pollen counts. It was one of the old fashioned models with 72 keys: 8 columns of keys 1 to 9. It devoted the two right columns for cents so the largest number you could enter was 999999.99. When I saw a pine grain I would ring up a 1000.01, a spruce would get 1000.02, and a sedge 1000.13. You did not have to enter the zero columns, so it was not much work. As you might guess a spruce bladder would be recorded as 500.02. You could get a count total just by touching the subtotal key. The sum of all the taxon identity numbers was gibberish, of course, but by putting an imaginary decimal point three places left of the machine's, you would get the total of all the 1's and 0.5's. This was a low-tech way of keeping a history of the pollen types so that later I could do a statistical study to see whether the different sizes or types were randomly distributed on the slide.

After accumulating several miles of paper tape, it will suddenly occur to you that somebody is going to have to go back through it all to get the total for each taxon. At the time my research computer was a "Commodore PET," with a 16K memory (upgraded from its original 8K), and I proudly worked up a program to simulate a bank of 100 counters, any one of which could be incremented by touching a two-key code. I gave the program to anyone who would take it; I do not know whether it was used.

Keith Bennett (1990a, 1990b) devised a clever aid to pollen-counting by programming a Psion "Pocket Organiser" to record taxa by a combination of one or two letters. The Psion is battery operated, relatively inexpensive, and its data can be up-loaded to a PC. I use one and enjoy it. The only problem with it is that once a count is started, you cannot do other programming with it without saving the data first. And then it is not easy to put the partial count back in the machine to continue the count. But if you think of the Psion as a dedicated counter, this is no problem; it simply shuts down and "sleeps" when you stop touching the keys.

I never bothered to use my IBM PC as a pollen counter; it is in constant use for other things. But then I noted Pierre Zippi (1992) had a program for turning a Macintosh into a microfossil counter, and I got to thinking that a lot of labs probably have old outmoded DOS PCs around that are being used mostly as furniture. So I reworked my old PET program by borrowing some of Keith Bennett's Psion ideas and added some save and retrieve features. Owen Davis comments in this issue (p. 14) that we ought to deal with counters for microfossils, so this is a first shot. If others let me know what they have developed, I will include them in the next issue.

POLCOUNT sets up 100 counters for 100 different pollen taxa. Think of these counters as labeled 00, 01, 02, ... 99; each is assigned to a pollen taxon. Each time POLCOUNT is run, a *.TAX file is loaded that associates the name of a real object (a pollen taxon, for example) with each of the counters. You can load TAXLIST1.TAX, a dummy taxon list that comes with POLCOUNT, in order to see how the program works. To record a grain of taxon 05 in counter 05, simply type the two numbers 0 and 5 in sequence. You will hear a click, and the actual name of taxon 05 will appear on the screen as confirmation. You do not need to touch the <Enter> key; the program knows that any two strokes on number keys should increment the named counter by one. You will note that keys other than numbers usually produce an irritating beep. (The <Esc> key will always end the program if you wish to do so.)

The number keys serve to increment the counters. One of the 100 counters you control is reserved for a special purpose: 00 is used for marker grains that you might add to the sample for determining pollen concentration. There is also a counter 100, but you cannot access it directly because 99 is the largest two-digit number you can enter. Counter 100 sums the counts in counters 01 through 99 (it does not include counter 00); you can think of it as keeping a rough record of the pollen sum.

POLCOUNT loads its Taxon List each time the program is run. It offers TAXLIST1.TAX as the default name, and you can choose that by merely touching the <Enter> key. The initial TAXLIST1.TAX file is a dummy sample that allows you to see how the program works. TAXLIST0.TAX is an 'empty' list consisting only in the names: Marker Grain, Taxon 01, Taxon 02, etc. When you decide the taxon order you wish to use, load TAXLIST0.TAX into your word processor and replace the dummy names with the actual taxon names you are going to use. The names can be short or long; they can contain spaces. When you wish to look at the taxon list while running POLCOUNT (see F7, below), only the first 8 letters will show. SAVE YOUR LIST IN ASCII TEXT FORMAT under a different name, but ending with the extension .tax. If you use TAXLIST1.TAX, it will load as the default. Different analysts can use different lists simply by giving them different names (But always with the same 3-letter extension: '.TAX') which can be entered manually from the *.TAX files shown in the directory listing.

The function keys F1 - F10 allow you to correct errors, seek information, edit the counters, and load and save data files. The purpose of the function keys is always displayed at the bottom of the screen while counts are recorded:

F1
Help gives a quick description of the program.
F2
Ignore the first key pressed if the mistake is noted before pressing the second key.
F3
Subtract 1 from the last counter changed; i.e. correct a mistake. If the counter ranged between 01 and 99, then the pollen sum counter would be debited as well.
F4
Subtract 0.5 from the last counter changed. This can be used to record half grains. If Picea were taxon 01, and a single Picea bladder were noted, the key sequence 0, 1, F4 would, in effect, add a half count to counter 01 (and to the pollen sum counter 100).
F5
Fix/Correct a counter sum. This allows you to make substantial corrections. If you have been recording a special unknown grain type and then find out what it is, you can press F5. The program will ask which counter you wish to change, and it will tell you what that counter now holds. You can then add (+ value) a number to one counter and subtract (- value) it from another, etc.
F6
Displays the contents of counters 00 (markers) and 100 (pollen sum).
F7
Displays for counters 01 - 99, their two-digit identification numbers, the first 8 letters of their taxon name, and the number of grains recorded. The number of marker grains and the pollen sum are indicated in the screen's title. This allows you to look up the 2-digit Number Code for a particular Taxon. F7 is especially useful for looking up the code numbers of taxa so rarely encountered that they are hard to remember. When you first start using POLCOUNT, it is best to use a key on a sheet of paper. You can make one by putting your printer on line; then run POLCOUNT, touch F7 to display the taxon list, and then press the computer's <PrintScreen> key.
F8
You can change your Taxon List file as you work although you should use this feature only on rare occasions. You should think carefully about the taxon list before you start a project and use a word processor to edit TaxList0.TAX that comes with POLCOUNT. If you save your file with the default name TaxList1.TAX, it can be loaded very easily when you start POLCOUNT. For many projects, you will not need all 100 counters, and the higher numbered ones can be left as dummy 'Taxon nn'. When you begin to find taxa that you had neglected to put on the taxon list, press F7 to find an unused counter; touch any key to return to the count screen; then touch F8 to assign it a taxon name. When you are finished, the changes will be immediately saved to your *.TAX file on disk.
F9
When you start POLCOUNT, you can recover a disk file from a previous session and go on counting an unfinished slide. When you press F9 you will be shown the files in the directory. Choose the one to load, and continue the counts.
F10
SAVE CURRENT FILE. This is the normal way to end a POLCOUNT session: Save the file and quit. When the file is saved you are given the option of continuing the count. Using F10 at intervals is a way of making sure your data are mostly saved in case of a power failure. The first time, you will be asked the name of your site and the sample number (usually the depth in cm). For reasons that will be obvious later, make the file names as short as possible. Level 302.5 from 'Blue Lake' could be saved as: BL302.5
POLCOUNT files for the individual samples can be combined into "Wisconsin" format files. POLFILE allows these to be imported into spreadsheets, or PALYPLOT, or TILIA and TILIA·GRAPH.

Two POLCOUNT sample files are shown below. POLCOUNT makes an ASCII text file, fashioning a title line from the site's name and the day's date. The second line is always 102, the number of categories in the file: counters 00 through 100, plus the sample's identifying number, usually its depth in centimeters. You will note that all data lines (except the last) consist of ten values, each separated by a space. The sample's depth in centimeters is the first value on the file's third line; the other numbers in the sequence are the sums in counters 00 through 100. The sum in Counter 100 will equal the total of all the previous categories EXCEPT THE FIRST TWO, which are Depth and the number of Markers.

Blue Lake, WI = 01-02-1992
 102
 .5 250 1.5 0 7 12 1 1 16 5
 0 33.5 13 172 5 3 5 10 5 1
 1 1 1 2 0 0 2 0 0 0
 0 0 0 0 0 24 11 5 153 4
 13 0 1 1 4 2 0 1 2 3
 0 0 0 0 0 7 5 2 21 38
 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0
 0 595

Blue Lake, WI = 01-11-1992
 102
 535 315 30.5 1 117 0 5 23 123 8
 44.5 153.5 44 59 10 7 11 4 0 0
 0 0 0 1 0 1 0 0 0 0
 0 0 0 0 0 6 13 5 3 2
 2 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 1 9 5 1 40
 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0
 0 729.5
The following is the partial list of taxa in TAXLIST1.TAX. The file has a title line (here, BlueLake Taxa), followed by the number 101, which is the number of categories to follow.
BlueLake Taxa
101
Marker
Picea
Larix
Fraxinus
...
Taxon 98
Taxon 99
Sum(01-99)
A "Wisconsin" format file has a site title for the first line. The second line records the number of taxon categories in the file, and the third line records the number of samples in the file. The sample counts then follow in sequence from top to bottom in the core, each starting with an identifier such as the sample's depth in centimeters. The 102 taxon categories are found at the end of the file, each on a separate line. Use a word processor to make a "Wisconsin" format file. Load each POLCOUNT sample in sequence into a single large file and append the TAXLIST1.TAX file at the end. Use the word processor to delete the first two lines of each sample (the title and the number 102). Put a main title at the top, followed by the number 102 on the second line and the total number of separate samples on the third line. From the appended TAXLIST1.TAX list, delete the first name (the title) and the second line (101). Insert the word Depth(cm) on a separate line just after the counts and just before "Markers." The following is an example of a "Wisconsin" format file made from the example material.
Blue Lake, Wisconsin
 102
 2
 .5 250 1.5 0 7 12 1 1 16 5
 0 33.5 13 172 5 3 5 10 5 1
 1 1 1 2 0 0 2 0 0 0
 0 0 0 0 0 24 11 5 153 4
 13 0 1 1 4 2 0 1 2 3
 0 0 0 0 0 7 5 2 21 38
 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0
 0 595
 535 315 30.5 1 117 0 5 23 123 8
 44.5 153.5 44 59 10 7 11 4 0 0
 0 0 0 1 0 1 0 0 0 0
 0 0 0 0 0 6 13 5 3 2
 2 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 1 9 5 1 40
 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0
 0 729.5
Depth(cm)
Marker
Picea
Larix
Fraxinus
...
Taxon 98
Taxon 99
Sum(01-99)
When the "Wisconsin" format file has been created with your word processor, save it as an ASCII text file with the extension .RAW, such as BLUELAKE.RAW. This file will undoubtedly contain some taxa that never received a count in any of the samples. You can remove the taxa with zero counts from the file by using the program REMZEROS. Run REMZEROS and load "BLUELAKE.RAW." You can save the file without the unused taxa either with the same name or another of your choice. But it is good practice not to use the same name; keep the original file intact in case you wish to count additional levels later. The new levels can be put in their correct positions with a word processor and REMZEROS run again to remove any taxa still with zero counts.

If you wish free copies of POLCOUNT and REMZEROS, I would be pleased to make them available both as compiled programs and as QuickBASIC source code. The latter will work with QBASIC that comes with DOS 5.0, and will allow you to modify the programs to fit your own needs and output file structures. These programs make no use of graphics and should run on any IBM compatible PC. The pollen counter package which contains the programs, sample files, and instructions for transferring the final counts into TILIA, is available in a self-unzipping format as POLCNTPK.EXE. This is also available for Anonymous FTP in directory /pub/inqua of geology.wisc.edu (see the article on the Internet on p. 17 of this issue).

Bennett, K. D. 1990a. Pollen counting on a pocket computer. INQUA - Commission for the Study of the Holocene, Working Group on Data-Handling Methods Newsletter 3:5.

Bennett, K. D. 1990b. Pollen counting on a pocket computer. New Phytologist 114:275-280.

Zippi, P. A. 1992. Scientific software for Apply Macintosh. INQUA - Commission for the Study of the Holocene, Working Group on Data-Handling Methods Newsletter 7:10-13.


Copyright © 1992 Louis Maher
Home page
Newsletter 8 index
Author index
Subject index
WWW pages by K.D. Bennett