*SHOW Translate into KEY format, directives reworked June 1992 
 
*HEADING Grass genera of the world 
 
*LISTING FILE tokey.lst 
 
*COMMENT. 
----- READ THIS ----- 
 
The products of this directives file (KCHARS, KITEMS) are intended as a 
general purpose conversion, for use in association with special purpose 
KEY directives files (for an example, see the file KEYAUST1). 
Items and characters to be included in particular keys, 
and the various KEY parameters, should be specified in the latter files. 
When experimenting with key generation, it is quicker to run TOKEY once 
and make adjustments via these special purpose KEY directives files: your 
keymaking run will then be invoked by (e.g.) KEY KEYAUST1. 
 
The CHARACTER RELIABILITIES as generated here will default to 1-495,5 and 
will be over-ridden by those specified in the KEY directives file. The 
file KEYRELIA contains a set of reliabilities suitable for preliminary 
efforts at 'conventional' keys: they will usually need minor modification, 
depending on the generic sample. 
 
The following KEY STATES are more or less suitable for making 
'conventional' keys to fairly large suites of diverse genera. They may 
require modification when working with taxonomically restricted samples 
(e.g. temperate floras largely comprising pooids), in which case you will 
have to alter them here, then re-run TOKEY. 
 
If an appropriate regional generic list exists in the .TAX set provided 
with the INTKEY package, the easiest way to obtain your INCLUDE ITEMS for 
keymaking is via this. For example, to obtain it for China: in INTKEY, 
enter INPUT CHINA.TAX followed by INC T CHINA. Then FILE OUTPUT CHINA.TMP, 
followed by OUTPUT TAXA REM. The output file (CHINA.TMP) should now be 
edited, removing everything but the list of taxon numbers, which 
constitutes the INCLUDE ITEMS for the your KEY directives file. 
 
LIKELY SOURCES OF PROBLEMS WITH LARGE KEYS: 
 
1. Separate, species-level descriptions are appended to the ITEMS file for 
two widespread members of the Koeleria/Trisetum/Rostraria/Trisetaria 
(etc.) shambles: i.e. 'Lophochloa cristata' and 'Lophochloa pumila'. Where 
one of these is the only representative of this taxonomic mess in your 
region, keymaking will be much simplified by excluding the generic 
descriptions and using the appropriate species description(s) instead. 
 
2. Thellungia is inseparable via printed keys from some variants of 
worldwide Eragrostis. 
 
3. The world descriptions of Eragrostis and Diplachne are absolutely 
separable via printed keys only with reference to dubious characters. This 
exemplifies a situation where it may be worthwhile editing the world 
descriptions, to remove variants that are not represented in your region. 
Another example concerns Aristida, where for most regions it is 
appropriate to adjust the awn description so as to render the 'trifid' 
form diagnostic. However, experience with the present package has shown 
that few if any data modifications will required to obtain satisfactory 
results. 
 
4. In general, sensu stricto versions of genera have been 'accepted' for 
this database, wherever they seem at least to some extent defensible. The 
compelling reasons for this are detailed in associated publications. You 
can easily coalesce the existing DELTA descriptions into sensu lato 
versions, however, by using the SAVE SUMMARY directive of INTKEY (q.v.). 
To obtain information on descriptions which might usefully be coalesced, 
assign INTKEY, restrict the taxa to those of interest to you (e.g. those 
of your region), then DESC T REM NOMEN. You will need to delete the 
'comments' containing statistical data from the INTKEY-generated coalesced 
descriptions (their presence will cause DATA BUFFER problems); but a more 
serious problem is that you may find sensu lato (world) descriptions 
resulting in very ponderous keys, or even in incompletely separable 
genera. It may be better to leave the descriptions as they stand, and 
simply alter the names in the final product. Note that a description of 
Stipa sensu lato is provided, affording a key-making alternative to the 
more recently advocated segregates. If the latter are preferred, the sensu 
lato version must be masked out. 
 
5. A generic description of Sieglingia is provided here, which may 
facilitate and improve key-making in regions where S. (Danthonia) 
decumbens is the only representative of the Danthonia complex. Other 
segregates of danthonioid genera also included here, which may serve a 
similar purpose, are Erythranthera, Monostachya and Pyrrhanthera (all 
supposedly covered by the description of Rytidosperma sensu Linder); and 
Urochlaena (reduced by him to Tribolium). In all such cases, if the sensu 
lato description is employed, the segregates must be masked out for key- 
making (and vice versa). 
 
6. A few pairs of genera are absolutely separable only by using a 
character which is variable in one of them, followed by use of another 
character dependent on the first. The directive ALLOW IMPROPER SUBGROUPS 
will be required in order to resolve them. This sometimes results in very 
ponderous keys, with repetitious incomplete separations of such pairs. 
They will be improved by presetting the critical controlling character 
where appropriate. Beware of the following pairs in this connection: 
 
Panicum/Ichnanthus (use presence/absence of a conspicuously stipitate 
upper floret, followed when present by the form of the stipe) 
 
Eragrostis/Heterachne (use lemma carinate/not, followed when carinate by 
keel winged/not) 
 
*INPUT FILE specs 
 
*TRANSLATE INTO KEY FORMAT 
 
*COMMENT STOP AFTER ITEM 2 
 
*COMMENT USE NORMAL VALUES 13 141 189 208 248 272 
*COMMENT USE NORMAL VALUES 208 
*COMMENT USE NORMAL VALUES 189 208 248 272 
 
*COMMENT. ABSOLUTE ERROR 8,10 25,0.005 
*PERCENT ERROR 13,50 15,50 56,25 141,10 
 
*COMMENT World KEY STATES 
*KEY STATES 13,~2/2-5/5-25/25-75/75-150/150-300/300-1000/1000~ 15,~5/5~
25,~2/3~ 28,1/2/3/4~ 56,~1/1-3/3-5/5-12/12-25/25-50/50-100/100~ 69,1/2-3/4
71,~1/1-5/5-12/12-20/20-40/40~ 96,~2/3-5/6-12/12-20/20~ 135,1/2/3-5/6~
137,1/2/3-5/6~ 138,1/2/3/4/5/6/7~ 141,~0.9/0.9-2/2-5/5-15/15-30/30~
180,~0.3/0.3-0.75/0.75~ 187,0/1/2/3/4/5/6/7/8-9/10-12/13-19/20~
189,0/1/2/3/4/5/6/7-9/10-12/13-19/20~ 193,1/2/3~ 197,0/1/2/3~
204,0/1/2/3/4/5/6/7/8-9/10-12/13-19/20~ 208,1/2/3~
220,1/2/3/4/5/6-8/9/10-12/13-23 225,1/2/3/4/5/6-8/9/10/11/12/13-23/24~
248,0/1/2/3/4/5/6/7-9/10-12/13-19/20~ 265,1/2/3/4-5/6/7~
272,0/1/2/3/4/5/6/7-15/15~ 274,0.1/0.1-0.5/0.5-1/1-3/3-7/7-15/15~
283,1/2/3/4-5/6/7~ 314,1/2-5/6-20/20~ 339,~30/30-86/86-100/100~
342,~0.5/0.5-1.5/1.5-3/3-9.5/9.5~ 343,~20/20~ 344,~0.3/0.3-0.7/0.7~
346,~29/29-49/50-75/75~ 414,~4/4~ 469,~5/6-7/8/9/10/11-15/16~
470,~2/3-19/20-59/60~ 471,2/3/4/5/6~ 474,~2/2~ 475,~5/5~ 491,1/2/3~ 
 
*COMMENT Britain KEY STATES 
*COMMENT KEY STATES 13,~4/4-50/50-150/150-200/200-600/600~ 15,~5/5~
25,~2/3~ 28,1/2/3/4~ 56,~1/1-3/3-5/5-12/12-25/25-50/50-100/100~ 69,1/2-3/4
71,~1/1-5/5-12/12-20/20-40/40~ 96,~2/3-5/6-12/12-20/20~ 135,1/2/3-5/6~
137,1/2/3-5/6~ 138,1/2/3/4/5/6/7~ 141,~0.9/0.9-2/2-5/5-15/15-30/30~
180,~0.3/0.3-0.75/0.75~ 187,0/1/2/3/4/5/6/7/8-9/10-12/13-19/20~
189,0/1/2/3/4/5/6/7-9/10-12/13-19/20~ 193,1/2/3~ 197,0/1/2/3~
204,0/1/2/3/4/5/6/7/8-9/10-12/13-19/20~ 208,1/2/3-10/11~
220,1/2/3/4/5/6-8/9/10-12/13-23 225,1/2/3/4/5/6-8/9/10/11/12/13-23/24~
248,0/1/2/3/4/5/6/7-9/10-12/13-19/20~ 265,1/2/3/4-5/6/7~
272,0/1/2/3/4/5/6/7-15/15~ 274,0.1/0.1-0.5/0.5-1/1-3/3-7/7-15/15~
283,1/2/3/4-5/6/7~ 314,1/2-5/6-20/20~ 339,~30/30-86/86-100/100~
342,~0.5/0.5-1.5/1.5-3/3-9.5/9.5~ 343,~20/20~ 344,~0.3/0.3-0.7/0.7~
346,~29/29-49/50-75/75~ 414,~4/4~ 469,~5/6-7/8/9/10/11-15/16~
470,~2/3-19/20-59/60~ 471,2/3/4/5/6~ 474,~2/2~ 475,~5/5~ 491,1/2/3~ 
 
*COMMENT. INCLUDE CHARACTERS 
7-9 11 13-17 19 24 27-28 37-49 51-52 54-60 63-79 82-83 86-96 98-130 132-139
141-142 144-158 160-201 203-209 211-212 214-237 240-253 255-260 262-280
283-284 286-287 289-291 293-301 323 332-335 348-349 359-361 363-366 370-372
374 389-397 406-408 417-452 454-468 477 493-503 505-510 512-514 
 
*KEY OUTPUT FILE kchars 
*INPUT FILE charsi 
 
*KEY OUTPUT FILE kitems 
*INPUT FILE items 
