National Vegetation Information System Taxonomic Review
Centre for Plant Biodiversity Research
Department of the Environment and Heritage, 2004
The CPBR assessed the taxonomic accuracy and currency of the 5,447 species used in Taxon_Lists table to create plant community descriptions in the NVIS Veg_Description table. Priority was on major, widespread groups such as eucalypts (Eucalyptus, Corymbia and Angophora), Acacia (Mimosaceae), the grasses (Poaceae) and dominant shrub families such as Proteaceae and Epacridaceae. Priorities were set (see Appendix C) as it was unclear at the start of the project as to whether all records could be examined in the time frame. All records were examined for taxonomy and comments were added in the new columns (Sections 2.1 and 2.2) where appropriate.
Along with taxonomic assessment, comment on the accuracy of authors was also provided for several major plant groups, family Proteaceae, Acacia and the eucalypts. This author assessment was carried out to check the accuracy and consistency of information between different State data providers. The issue of plant name authors is discussed in Section 3.2. The major plant groups are discussed in detail in Section 3.3.
Of the 5,447 records assessed a number of taxonomic issues were apparent. 896 (16.45%) records need to be altered to some degree. Table 3.1 shows record numbers, the types of taxonomic issues and their frequency within the Taxon_Lists dataset.
|NO. RECORDS||% RECORDS|
|Single (unique) occurrence names||2704||49.6|
|INFRASPECIFIC NAME ISSUES|
|Infra name in wrong field||1||0.02|
|Infra name missing||37||0.7|
|OTHER TAXONOMIC ISSUES|
|Alternate family name||415||7.6|
|Missing family name||1||0.02|
|Name qualifier present||51||0.9|
|Total taxonomic issues||896||16.45|
The issues shown in Table 3.1 will be discussed in greater detail in the following sections.
As discussed in Section 2.21 there was early evidence of many of the records present in the Taxon_Lists dataset being duplicates of other records. All taxa were assessed to separate those that had unique records from those with multiple records. Taxa with multiple records were marked with either “M” for master records and “D” for duplicate records, following the procedure outlined in Section 2.21. Overall, just over 50% of Taxon_List records involved some form of name duplication.
The identification of name duplication helped provide an accurate picture of how many taxa are listed in the Taxon_Lists dataset. The 5,447 records present relate to 3,684 individual taxa.
Duplication was based on the identification of multiple occurrences of an individual taxon, not necessarily an exact match of the taxon’s name. While an exact match of name was often the case, there were many examples of a taxon with multiple records with slight variations in the way authors were written. Despite these minor variations, records identified as duplicates were marked accordingly, with a CPBR_SOLUTION reference to the master record. Table 3.11 shows examples of how duplication was presented in the Taxon_Lists dataset.
|Taxdsc Id||Genus||Species||Species Author||Rank||Infra Species||Infra Author||Master record||Cpbr Solution|
|118726||Carpobrotus||glaucescens||D||Refer to 117982|
|137137||Carpobrotus||rossii||(Haw.) N.E.Br.||D||Refer to 2041|
|15723||Carpobrotus||rossii||(Haw.) Schwantes||D||Refer to 2041|
|16418||Disphyma||crassifolium||(L.)L.Bolus||ssp.||clavellatum||(Haw.) Chinn.||D||Refer to 2520|
Of the 5,447 NVIS taxon records, 538 (9.9%) relate to infraspecific taxa, typically subspecies and varieties. The NVIS database structure separates each name element ; this includes a separate column for the infraspecific rank component of the name. This “rank identifier” was populated by a considerable amount of typographic error. Of the 538 infraspecific names, 227 records (42%) had the rank component of the infraspecific name misspelt. Table 3.12 below shows examples of typographic errors found with reference to State data sources.
|Misspelt rank identifier||Correct spelling||Source State||Number of misspellings|
|ssp.||subsp.||ACT, NSW, SA, Vic||146|
The most common misspelt rank found was for subspecies, being abbreviated to “ssp.” as opposed to the more usual form of “subsp.” Other simple errors included the dropping of full stops at the end of abbreviations. The high degree of variation in rank spelling is a concern when trying to match up taxon records between States. The example below demonstrates how a minor variation in the spelling of the species rank identifier could potentially see the same taxon treated as two separate taxa.
Eg. Lomandra confertifolia subsp pallida NSW
Lomandra confertifolia subsp. pallida Qld
The other major infraspecific name related to some records possessing a rank identifier with no infraspecific name. Nomenclaturally this makes no sense; typically if the infraspecific taxon is unknown the rank identifier is not added. Two examples of this are shown in the example below:
Eg. Lomandra multiflora ssp.
Dianella revoluta var.
All 37 cases of this relate to SA provided records probably from the same single source. It seems likely that the infraspecific names were lost during the data transfer.
A total of 69 records were found to have synonymous (non-current) names. It should be noted that these names are considered non-current as compared to the APNI/WIN database. It should also be noted that in some cases, there were differences in names when comparing between States.
These were scored as two synonym types:
- Nomenclatural synonyms
- Taxonomic synonyms
The types and frequency of synonyms for each State contributor are shown in Table 3.13 below.
|State||Nomenclatural||Taxonomic||Total||% of synonyms|
See Appendix D for a full list of synonyms and what is listed as the correct name on APNI/WIN.
The majority of synonymous names proved to be nomenclatural synonyms. A nomenclatural synonym is a non-current name, which is based on the same type (specimen) as the current name concerned. Most nomenclatural synonyms are a simple case of a change in generic limits (see example below). In some cases, changes in taxonomy in recent publications has been picked up in APNI/WIN but not updated by the States in their respective databases.
Eg. Danthonia caespitosa to Austrodanthonia caespitosa
Cyathodes juniperina to Leptecophylla juniperina
In other cases, publication of new taxa results in the published name replacing a temporary phrase name.
E.g. Leptorhynchos squamatus subsp. A to Leptorhynchos squamatus subsp. squamatus
A total of fifteen taxonomic synonyms were found, spread fairly evenly between the States. A taxonomic synonym is a non-current name, which is based on a different type (specimen) to the current name concerned. Most taxonomic synonyms encountered in NVIS are the result of changed taxonomic concepts, where for example, two related taxa are reassessed and considered to represent one taxon, the older of the two names then assuming priority (see Greuter et al. 2000).
E.g. Hakea muelleriana to Hakea mitchellii
In other cases, taxonomic synonyms may be generated by the reduction of published infraspecific taxa to synonymy, where these are found to be unworkable or poorly defined.
E.g. Atriplex nummularia subsp. omissa to Atriplex nummularia.
Taxonomic synonyms may also be generated where there is a difference of opinion as to the taxonomic limits of a particular taxon. In the example below, there is a difference in taxonomic opinion between NSW and other State institutions.
E.g. Themeda australis (NSW) to Themeda triandra (other institutions)
Most institutions consider the Australian Kangaroo Grass to be part of a cosmopolitan species known as Themeda triandra. NSW recognises differences between the Australian populations and those found in Asia and Africa, recognising a separate species, Themeda australis. For the purposes of this example, APNI/WIN is following the consensus view and treats Themeda australis as synonym.
- 3.141 Alternate family name
- 3.142 Missing family name
- 3.143 Name qualifier present
- 3.144 Double epithets
- 3.145 Status unknown
- 3.146 Name misspelt
- 3.147 Phrase names
- 3.148 Non-plant taxon
A wide range of other taxonomic issues were encountered, these included:
- Alternate family name
- Missing family name
- Name qualifier present
- Double epithet
- Status unknown
- Name misspelt
- Phrase name
- Non-plant taxon
These issues are presented in Table 3.14 along with a State breakdown.
|Alternate family name||154||2||18||224||10||7||415|
|Missing family name||1||1|
|Name qualifier present||2||49||51|
415 records possessed family names that were different to those found in APNI/WIN. In accordance with the principles of taxonomic hierarchy, each family can be made up of many genera and many more species. However, each species can only belong to one family. As our understanding of plant evolution progresses, there is considerable debate and discussion as to the limits of higher taxa, especially families, and changes in family composition are frequent and rarely stable.
Some differences in family names are due to different institutions following different family treatments, some of which differ again to those adopted by APNI/WIN. The high total for NSW can be accounted for in this way, particularly with the legumes. APNI/WIN recognises three separate families for the legumes, whereas NSW recognises one family, divided into three subfamilies. South Australia on the other hand recognises yet another alternative for the legumes. Table 3.141a shows the variability in family names for some of the legume genera as compared to APNI/WIN.
|State||Genus||NVIS family name||APNI/WIN family name|
Whilst such differences of systematic opinion are common amongst botanical institutions, it does make it difficult to group families in the NVIS database.
Table 3.141a shows the use by South Australia of the alternate legume family name Leguminosae. South Australia data includes many other alternate family names, including Umbelliferae (Apiaceae), Gramineae (Poaceae), Compositae (Asteraceae), Labiatae (Lamiaceae) and Cruciferae (Brassicaceae). The International Code of Botanical Nomenclature, Section 18.5 (Greuter et al. 2000) makes it clear that such alternate family names are validly published and nomenclaturally acceptable. However for the purposes of linking different State’s data for genera such as Acacia, alternate family names such as Leguminosae make the process difficult.
Other occurrences of alternate family names in the Taxon_Lists dataset seem to be simple errors, with species being linked to families that they clearly have no relation to. Table 3.141b below shows examples of some of these family errors, all such occurrences relate to NSW data.
|Species||NVIS family||APNI/WIN family|
Two other records seem to have simple typographic errors in the FAMILY field; these are shown in Table 3.141c below.
|Species||NVIS family||APNI/WIN family||State|
There was one record that had no information present in the FAMILY field. This seems to have been a simple oversight with one of the NSW records, Linum marginale (Linaceae).
The process of trying to name and describe organisms is often problematic, living entities not easily fitting into simple hierarchical structures. Delimiting plant species from one another and naming variations has historically been difficult, particularly for closely related taxa. Some species, such as Acacia aneura are apomictic, that is, possessing an ability to reproduce asexually through seed. Apomixis and other biological complexities often result in complex clusters of microspecies, not easily dealt with in traditional taxonomic and nomenclatural systems. One method that is often used to help make allowances for such complexity is the use of name qualifiers. These qualifiers help provide additional information about the delimitation of a particular taxon. Unfortunately botanists have through time used different forms of qualifiers to describe similar forms of complexity. This is apparent in the Taxon_Lists data set, with “complex”, “sp. agg.” (species aggregate), “s.l.” (sensu lato = in the broad sense) and “nothossp.” (notho-subspecies) all being used to describe species complexes of different kinds.
As can be seen in Table 3.14, the NVIS dataset has 51 records that have some form of qualification. Table 2.22 lists all types of qualifiers present, generally occurring in the same field as the species epithet. The example below showing how qualifiers typically appear in the Taxon_Lists dataset:
E.g. Triodia irritans complex Pellaea falcata s.l.
In general the State contributors have avoided using name qualification, preferring to keep species names as simple as possible. Table 3.14 shows that name qualification is used almost exclusively by Victoria; with two “complex” qualifiers also used by South Australia.
While qualifiers are a valid descriptor in plant taxonomy, they can be problematic as far as databases are concerned. The fact that different data providers choose to use them and others avoid their use makes it difficult to compare records, even where the same taxon is concerned. Similarly, duplication could not be assigned to taxa with qualified names, the presence of the qualifier resulting in a name different enough to not be considered a true duplicate.
There were four of examples of what for the purposes of this project described as “double epithets”. These are cases where records do not have the typical single species epithet, but two epithets separated by a slash. The example below shows how these are displayed in the NVIS database:
Eg. Eucalyptus saligna/botryoides
Similar to species complexes described with qualified names in Section 3.143, these double epithets describe the complexity found with closely related or unrelated, but morphologically similar, taxa.
One case of likely hybridisation is shown in the example above. Eucalyptus botryoides and Eucalyptus saligna are known to form hybrid swarms (Harden 1991), with hybrid plants in these populations often showing features of both parents. This may account for the use of the double epithet in the case of this NSW record. Known hybrids are typically written as shown in the example below:
Eg. Eucalyptus saligna x botryoides
Where a hybrid is sufficiently frequent or well known, it may be formally named, for example Adenanthos x cunninghamii (Proteaceae) a hybrid between A. cuneatus and A. sericeus. These are comparatively rare and none were noted in the NVIS dataset.
Another explanation for the use of double epithets is difficulties with identification. For example, two related (or sometimes unrelated) species may be identical in many of their shared characters, but differ significantly in one character (for example fruit shape or flower colour), characters which ma y not be apparent for part of the year. This probably explains the presence of the Sclerolaena diacantha/uniflora record. These species are separated in Harden (1990) by fruit features, but are otherwise indistinguishable.
There were 72 records in the Taxon_Lists that related to taxa without identification to species level. This has been represented in the database by some form of “species qualifier”. For the purposes of editing the database, “status unknown” was added to the CPBR_PROBLEM field and “U” was added to the CPBR_WIN_CURR field. Whilst the status of these qualifiers is understood, there is no equivalent to these species qualifier names within precise name databases such as APNI/WIN. They are thus difficult to compare and were marked as unknown.
Table 3.145 shows some examples of these records, starting with records identified to genus level, and then followed by two examples identified to family level. All States use some form of these species qualifiers records as shown in Table 3.14.
As can be seen in Table 3.145, the qualifier present in the SPECIES field is quite variable, with “species”, “sp.” and “spp.” used. It is difficult to determine whether the use of these qualifiers is referring to a single species or multiple species of that genus or family. The correct use of the abbreviation “sp.” relates to a single species, however it seems likely the use of “Cyperaceae sp.” in Table 3.145 refers to multiple sedge species and should be written as Cyperaceae spp. The Nelumbo species example shown in this table could be referring to a single taxon or multiple taxa. There is considerable confusion with the use of the species qualifier as to exactly what the original data recorder was referring to.
There are a number of reasons that the species qualifier may be present in the Taxon_Lists. Table 3.145 above shows some of the more traditionally difficult taxa to identify, such as Poa, Juncus and Cyperaceae. Whilst these groups are easily recognisable, they can be difficult to identify to species level and surveyors often add the genus or family name to species lists with the species qualifier.
Another possibility may be a surveyed site included numerous species of a family, which individually were occasional, but collectively the family is dominant. The example below shows how this can describe the character of a landscape fairly simply:
E.g. Eucalyptus spp. woodland
Asteraceae spp. dominated herb field
There were six examples of records with names featuring minor typographic errors in either the SPECIES or INFRA_SPECIES fields. These were identified in the CPBR_PROBLEM field and listed as “N” (not current) in the CPBR_WIN_CURR field. The correct spelling of the name, according to the APNI/WIN database, was then added to the CPBR_WIN_NAME.
Twelve examples of phrase names were encountered in the Taxon_Lists dataset, all from Queensland sources, see Table 3.14. Phrase names are temporary names given to taxa that may warrant formal description at a later date. Providing phrase names helps distinguish between multiple undescribed taxa within the same genus (as can be seen with Eucalyptus in Table 3.147 below) and also provides a means of referring to a known, undescribed taxon, formal description of which may be some years away. The examples listed shown the typical way modern phrase names are presented, including some indication of where a taxon occurs (e.g. Buckland Tableland) and a single herbarium collection (e.g. A.R. Bean 799) used for reference.
The phrase names encountered were highlighted, as they are not present within APNI/WIN; these records thus received a “U” in the CPBR_WIN_CURR field. There appears to be no problem with the presentation of these phrase names and one could assume that Queensland data providers have a good understanding of what undescribed taxa are present within their State. The APNI/WIN database in this case should be updated with these phrase names, which it has the capability of listing.
|Myrtaceae||Austromyrtus||sp. (Main Range P.R.Sharpe 4877)|
|Myrtaceae||Callistemon||sp. (Boulia L.Pedley 5297)|
|Myrtaceae||Eucalyptus||fibrosa||subsp.||(Glen Geddes M.I.Brooker 10230)|
|Myrtaceae||Eucalyptus||sp. (Buckland Tableland A.R.Bean 799)|
|Myrtaceae||Eucalyptus||sp. (Caldervale D.Jermyn AQ582304)|
The presence of a fungi taxon from a Queensland data source was part of the Taxon_Lists dataset. This was highlighted in the CPBR_PROBLEM field with a “Non-plant taxon” identifier and listed as “U” in the CPBR_WIN_CURR field as the APNI/WIN database does not include fungal families.
As mentioned in Sections 2.22 and 3.0, a portion of the Taxo n_Lists dataset was examined for the third component of plant scientific names, authors. Authors are the names (frequently abbreviated) of botanists who are involved in the publication of a particular plant name. A standard list of authors is maintained with the International Plant Name Index (IPNI 2003) based at Kew Botanic Gardens. The IPNI author list is used by APNI/WIN as the standard for all plant names. Authors for species and infraspecies were examined, however authors for genera and families were not relevant to the NVIS database and thus were ignored.
Records from 29 selected families (approx. 2,200 records, 50% of total records) were examined for author accuracy as per APNI/WIN name authors. Appendix E lists the families examined for author accuracy. The two relevant Taxon_Lists fields examined were SP_AUTHOR and INFRA_AUTHOR. Three types of author issues were encountered:
- author missing
- author incorrect
- author name in wrong field
Table 3.2 below presents the author issue types, their numbers and proportions for the selected families examined. From this table it can be seen that of the records examined, one in four had author inconsistencies or errors. Authors as they relate to some of the major plant groups are discussed in Section 3.3.
|Author issues||Record numbers||Proportion of records*|
|Author in wrong field||14||0.6|
* proportion of 2,200 records
The records found in the Taxon_Lists provided by State contributors were inconsistent when it came to the provision of author data. Most providers included some form of author; however NSW omitted authors with many of their records. In these cases, the CPBR_PROBLEM field was annotated with one of three messages:
- · Author missing
- · Author missing in INFRA_AUTHOR
- · Author missing in SP_AUTHOR
“Author missing” was added to taxa without infraspecific ranks. The other two messages were added where infraspecific ranks were present to help indicate which of the two authors needed correction.
Whilst the problem with these records is a simple presence/absence issue, it does result in an inconsistent dataset. It also makes it difficult to compare a taxon listed by one State with records from other States.
These records all had some form of author present, but what was present was inconsistent with the author listed in APNI/WIN. Just over 10% of records examined had incorrect author data. This may take the form of typographic errors, the wrong author cited, missing base names, misplaced punctuation or incorrect author abbreviations. In these cases, the CPBR_PROBLEM field was annotated with one of four messages:
- Author incorrect
- Author incorrect in INFRA_AUTHOR
- Author incorrect in SP_AUTHOR
- Author incorrect in two fields
“Author incorrect” was added to taxa without infraspecific ranks. The other three messages were added where infraspecific ranks were present to help indicate which of the two authors was in error, if not both.
This error relates the positioning of the author in the wrong field. All fourteen of these cases were Victorian records and all related to autonyms. When an author names a infraspecific taxon, such as Blechnum patersonii subsp. queenslandicum T.C.Chambers & P.A.Farrant, it is because they have recognised features that distinguish the newly described infraspecific taxon from the typical or “type” form of the species. This type subspecies (or variety) is known as an autonym; when an autonym is created the species epithet is duplicated at the subspecies rank, as shown in the example below:
Eg. Blechnum patersonii subsp. patersonii
The author for an autonym is the same as it is for the species as it relates to the same taxon. The author error encountered with these autonyms relates to its positioning. An example of one of these erroneous records is shown below:
Eg. Blechnum patersonii subsp. patersonii (R. Br.) Mett.
As is shown in this example above, the authority is in the wrong position, appearing in the INFRA_AUTHOR field. Autonym authors are positioned after the specific epithet. For the context of the Taxon_Lists dataset this would appear in the SP_AUTHOR field, resulting in the correct order as shown below:
Eg. Blechnum patersonii (R. Br.) Mett. subsp. patersonii
Authors should be placed after the infraspecific epithet in all non-autonymic names, as in Blechnum patersonii subsp. queenslandicum T.C.Chambers & P.A.Farrant above.
Along with assessing all records present in Taxon_Lists, it was decided that it would be useful to examine some of the more prominent Australian plant groups in more detail. The groups chosen were family Proteaceae, the wattles (Acacia) and the eucalypts (Eucalyptus, Corymbia and Angophora). It was felt that as these groups are important land cover taxa in all contributing States, it would be useful to compare the data provided by each of these contributors, particularly with regard to duplication of records.
As for all Taxon_Lists records, these large groups were assessed for taxonomic accuracy. Most taxa were also assessed for author accuracy, making up three of the 29 families mentioned in Section 3.2. Each of these three groups in the following subsections is presented in tables broken down by the taxonomic and authority issues encountered and by State.
Proteaceae includes many genera and species that make up important components of many Australian vegetation communities. Genera such as Banksia, Hakea and Grevillea are typically present in the shrub strata of many of these communities. Table 3.31 shows the details of taxonomic and nomenclatural issues encountered with Proteaceae records in the Taxon_Lists dataset.
In this it can be seen that all States provided records, NSW contributing the most records. Overall 186 records are present representing 112 individual taxa, which indicates a record duplication rate of 37%. While record duplication between States was expected, it was interesting to note that there was a 63% duplication of records for NSW alone, clearly reflecting multiple data providers for that state.
Authors were completely checked for Proteaceae, the main issue relating to missing authors for most of the NSW records. Incorrect author citation also accompanied 9% of the Proteaceae records. Misspelling of infraspecific rank was the next most common error, with ten records. Other issues were encountered in relatively low proportions.
It should be noted that Western Australian species were not assessed for this study due to data from that State being unavailable (see Table 1.34). South-west Western Australia has a very high diversity of Proteaceae with an equally high level of endemism. This should result in low duplication of records when added to the other records in the Taxon_Lists dataset.
|Single occurrence names||11||5||24||12||10||12||74|
|Author in wrong field|
|Infra name in wrong field|
|Infra name missing|
|OTHER TAXONOMIC ISSUES|
|Wrong family name|
|Missing family name|
|Name qualifier present||1||1|
The wattles (Acacia) comprise the most diverse plant genus in Australia, with species distributed across the continent in nearly all vegetation communities, especially in arid Australia. Table 3.32 shows the details of taxonomic and nomenclatural issues encountered with wattle records in the Taxon_Lists dataset.
High levels of record duplication were encountered with the wattles (33.3%), reflecting perhaps the nature of many species possessing a distribution crossing several state boundaries. New South Wales once again showed high levels of duplication between its data providers.
Author accuracy was examined for many of the Acacia records. Authors missing and incorrect citations were found with many of the records.
|Single occurrence names||23||30||83||21||3||8||357|
|Author in wrong field|
|Infra name in wrong field|
|Infra name missing||1|
|OTHER TAXONOMIC ISSUES|
|Wrong family name||80||42||122|
|Missing family name|
|Name qualifier present||1||1||2|
There were relatively few problems encountered with the infraspecific names or currency of names (synonymy) with the wattles. The greatest issue of concern was the inconsistency with the family name. As discussed in Section 3.141, there were several different family names used for the legumes in the Taxon_Lists dataset. Table 3.32 shows that NSW and SA provided alternate names to those used by other States and the APNI/WIN database.
The eucalypts for the purposes of this study relate three genera, Eucalyptus, Corymbia and Angophora. Eucalypts are of similar importance as land cover species as the wattles, and dominate the tree strata of most Australian woodland and forest communities. Perhaps reflecting the importance of this group, 13.5% of all Taxon_Lists records are eucalypts. Table 3.33 shows the details of taxonomic and nomenclatural issues encountered with wattle records in the Taxon_Lists dataset.
|Single occurrence names||17||294||55||207||54||28||78||733|
|Author in wrong field||2||2|
|Infra name in wrong field|
|Infra name missing||4||4|
|OTHER TAXONOMIC ISSUES|
|Wrong family name||1||1|
|Missing family name|
|Name qualifier present||10||10|
The highest level of record duplication was encountered with the eucalypts, out of the three large groups profiled. 44.7% of the 733 eucalypt records were duplicates, when consolidated the records relate to a total of 381 taxa. Duplication was also encountered when comparing individual contributors data within a state. New South Wales and Queensland columns are shown in Table 3.33 as representing fewer taxa than there are actual records.
Approximately half the eucalypt records were checked for author accuracy. From this sample it was apparent that many NSW records were missing authors, showing consistency with the findings outlined in Section 3.31 and 3.32. One trend that did stand out as being somewhat different was the high numbers of incorrect author citations that accompanied records from Victoria.
Infraspecific rank misspellings were also relatively high, relating to nearly 6% of all eucalypt records. The high numbers of nomenclatural synonyms encountered with the Northern Territory records relates to alternate views on taxonomy. The Northern Territory maintains a broad view of the eucalypts, retaining the Bloodwoods and Ghostgums within Eucalyptus, as opposed to other States and the APNI/WIN database, which recognise the generic segregate Corymbia for these groups.
The only other issue of note for the eucalypts was the ten records from Victoria that possessed name qualifiers. These qualifiers are discussed in Section 3.143 in greater detail.