PlantNET Home PlantNET Home | Search NSW Flora | Contact Us  

CERTAINTY AND UNCERTAINTY IN PLANT IDENTIFICATION

J.R. HoskingA, G.R. SaintyB and S.W.L. JacobsC

A NSW Agriculture, RMB 944, Tamworth, New South Wales 2340, Australia
B PO Box 1219, Potts Point, New South Wales 2011, Australia
C National Herbarium of New South Wales, Royal Botanic Gardens, Sydney, New South Wales 2000, Australia

Originally published as Hosking, J. R.; Sainty, G. R.; Jacobs, S. W. L. (1996). Certainty and uncertainty in plant identification. Proceedings of the Eleventh Australian Weeds Conference, 30 September - 3 October 1996, Melbourne, ed. R. C. H. Shepherd, pp. 464-467.

A few minor errors have been corrected in this electronic version.

SUMMARY
Many plant specimens are not identified correctly or many of the names used for them are not applied correctly. There is a need for plant identifications to be checked by specialists and specimens to be stored for future reference. This is particularly important for survey data and the only adequate long term storage appears to be in herbaria. Identifications are made in the light of current taxonomic knowledge and this is constantly being revised. Without access to the original specimens, results of previous surveys and research may not be of much use. Voucher specimens should therefore be a requirement for all plant research and this also applies to all weed research. Problems associated with use of databased collections and long term storage of specimens, particularly the expense of maintaining collections, are discussed.

INTRODUCTION
Even for those species that we now regard as being nomenclaturally stable or biologically well known, we have no idea what the future may bring in regards to new information and subsequent improvement in our biological understanding. These principles apply to weed species possibly more than any other. Weeds are often first recorded as a problem in the field, passed on in the form of an inadequate specimen to an identification authority, with little or no idea of their origin. They are frequently either identified with species that have proven troublesome elsewhere or identified from inappropriate publications from wrong geographical areas.

For species that have had a simple change of name there is not much problem. The synonymy can be quite straightforward and easily transferable. Where there has been a misidentification, at whatever level, or our knowledge has changed such that we now recognise two or more species in what was formerly one species (a good example is the ten species and a number of hybrids of blackberry that were formerly all called Rubus fruticosus) then it can be very difficult, if not impossible to track down what was the species actually being referred to.

Collecting and lodging relevant voucher specimens in recognised herbaria is the only process that allows the biological integrity of any particular survey or study to be checked or updated. We present examples below of some of the commonly confused species and some examples where confusion has clouded the literature, we discuss the processes involved in storing and maintaining the specimens and some of the techniques or information sources that can be misused. Species names used here are as in Harden (1990-1993) except where that species is not covered, and then the authority is given.

CONFUSION
There are many species that have been confused in the past and at present. Some examples of weed species commonly confused are given in Table 1.

Table 1. Weed species commonly confused and often incorrectly identified
FamilyGenusSpeciesComments
AsteraceaeAstersubulatus
Conyzaspp.
Erigeronspp.
Bidensspp.
Carduustenuiflorus
Carduuspycnocephalus
Cassinialaevis
Cassiniaquinquefaria
Centaureamelitensis
Centaureasolstitialis
Euchitonspp.prev. part of Gnaphalium
Gamochaetaspp.prev. part of Gnaphalium
Gnaphaliumspp.
Hypochaerisglabra
Hypochaerisradicata
Onopordumacanthium
Onopordumillyricum
Seneciolautussubspecies
Seneciomadagascariensis
Xanthiumcavanillesii
Xanthiumitalicum
Xanthiumoccidentale
Xanthiumorientale
CaryophyllaceaeArenarialeptoclados
Stellariamedia
Stellariapallida
Cerastiumspp.
ConvolvulaceaeCuscutaspp.
CyperaceaeCyperusbifax
Cyperusrotundatus
Cyperusvictoriensis
EuphorbiaceaeChamaesycespp.
Fabaceae - FaboideaeCytisusscoparius
Genista (or Teline)monspessulana
Genista (or Teline)stenopetala
Viciaspp.
Fabaceae - MimosoideaeProsopisspp.
FumariaceaeFumariaspp.
IridaceaeHomeriaspp.
Watsoniaspp.
JuncaceaeJuncusspp.
MalaceaeCotoneasterspp.
Crataegusspp.
Pyracanthaspp.
OleaceaeLigustrumsinense
Ligustrumvulgare
OxalidaceaeOxalischnoodes
Oxaliscorniculatus
Oxalisexilis
Oxalisperennans
Oxalisradicans
Oxalisrubens
Oxalisthompsoniae
PapaveraceaeArgemonemexicana
Argemoneochroleuca
Argemonesubfusiformis
PoaceaeAvenaspp.
Digitariaciliaris
Digitariasanguinalis
Echinochloaspp.
Eragrostisspp.
Panicumspp.
Phalarisspp.
Vulpiaspp.
PolygonaceaePersicariaspp.
Polygonumarenastrum
Polygonumaviculare
Rumexspp.
RosaceaeRubus'fruticosus'
Rubuschloocladus
Rubusdiscolor=R. procerus
Rubusleightonii
Rubuspolyanthemus
Rubuspyramidalis
Rubusradula
Rubusrosaceus
Rubusselmeri
Rubusulmifolius
Rubusulmifoliushybrids
Rubusvestitus
SalicaceaeSalixspp.
ScrophulariaceaeOrobanchespp.
Strigaspp.
SolanaceaePhysalisspp.
Solanumelaegnifolium
Solanumesuriale
ThymelaeaceaePimeleaspp.
TyphaceaeTyphaspp.
VerbenaceaePhylaspp.
Stachytarphetaspp.
Verbenabonariensis
Verbenaincompta P.W. Michael
Verbenacaracasana Kunth
Verbenalitoralis
Verbenaofficinalis
Verbenaquadrangularis Vell.= V. brasiliensis misapplied
ZygophyllaceaeTribulusmicrococcus
Tribulusminutus
Tribulusterrestris(introduced)
Tribulusterrestris(native)

In some cases confusion occurs between families, for example between Cuscuta spp. (Convolvulaceae) and Cassytha spp. (Lauraceae or Cassythaceae). It is also interesting to note that due to difficulty in identifying species of Cuscuta, all species in this genus have been declared noxious in many States of Australia (Parsons and Cuthbertson 1992). Native Cuscuta spp. are not considered to be a problem whereas C. campestris is considered to be a major problem (Parsons and Cuthbertson 1992). There are still a number of examples where species limits still need to be defined, for example at least two distinct entities are covered by the name Tribulus terrestris in Australia (Morrison and Scott 1993).

Misidentifications have resulted in delays to control programs for various weeds. This occurred in South Australia where Solanum elaeagnifolium was collected by J. M. Black in 1918 but it was believed to be Solanum esuriale at the time. In 1947 Black sent specimens to Kew and they were identified as S. elaeagnifolium (R. Carter personal communication). Co-ordinated control did not start until 1958 when the South Australian Department of Agriculture started to refer to the species as introduced (R. Carter personal communication).

Rapid spread of weeds may also occur through misidentifications. A recent example is the rapid distribution of alligator weed, Alternanthera philoxeroides, by the Sri Lankan community in Australia. The species was distributed in the mistaken belief that it was the vegetable, mukunawanna, Alternanthera sessilis (J. Dellow and R. Carter personal communications).

Herbaria The only way to minimise the problems of misidentifications or subsequent classification changes is to collect voucher specimens and to lodge them in a herbarium where there is some chance of the collections being maintained in the long term.

It is difficult for generalist Identification Officers, who are not specialists in any particular group, to correctly identify large numbers of specimens accurately. Such Officers deal with large numbers of enquiries for little or no charge, and are often very skilled. Mind reading, however, is not one of their skills and if you have some critical voucher specimens that should be retained then this information needs to be communicated. Most herbaria will not retain poor quality collections for any reason. If you have a research project where the lodging of voucher specimens is relevant, then you should arrange for the collaboration of an appropriate specialist beforehand. These days this often means including funds for identification.

There is also a cost associated with storage of plant specimens and this needs to be recognised. This cost should particularly be built into projects where many specimens will be collected and stored for future reference. The need for constant curation of collections is also necessary as anyone who has looked at specimens in herbaria will realise. It is difficult for Identification Officers who are not specialists in particular groups to be able to give the correct identification when a number of distinct species are included under the same name in collections. This is a common occurrence in herbaria and results in a number of misidentifications, but it is also how progress is made in understanding the group. The need to constantly update names and identifications in the light of current taxonomic knowledge and to increase funding to maintain collections cannot be overemphasized.

Collection of specimens There is a need for high quality plant specimens to be lodged in herbaria. In most cases this will mean flowering and fruiting sections of plants and in some cases other parts such as roots and bulbs. In some cases it is also desirable for collection of vegetative stages. This is particularly important for identification of forms of Chondrilla juncea (R. Groves personal communication). In this case natural enemies such as the rust fungus, Puccinia chondrillina Bubak & Syd., and the chondrilla gall mite, Aceria chondrillae (Canestrini) show specificity to particular forms of Chondrilla juncea (Groves and Cullen 1981).

Databases are no substitute for looking at the specimens Databasing of collections is increasing around Australia and this is desirable but it is no substitute for checking the specimens. Plants have often been misidentified, details from the collection typed in incorrectly and the location vague (could be applied to many areas). If using databases at least check outlying locations as these are most likely to be incorrect. It also pays to check whether there have been any problems with the database. This may mean that changes have been made in the collection but do not appear on the database.

Use of voucher specimens Voucher specimens can be used to check previous identifications in the light of current taxonomic knowledge. For example the photograph of Verbena bonariensis in Auld and Medd (1987) was redetermined as Verbena incompta (Michael 1995) because voucher specimens were lodged at the NSW Herbarium.

Many species are not sent to herbaria for identification because people think that they know the species they are dealing with. This was the case with Chromolaena odorata (L.) R. M. King & H. Robinson, from the Tully area, which locals called giant billy goat weed in the mistaken belief that the species was a large form of Ageratum conyzoides.

Many plant surveys, including weed surveys, have been published in the past where voucher specimens have not been lodged in a herbarium. We often have difficulty in believing some of the names on lists but there is no way to check the accuracy. Without specimens many of these records have to be disregarded. Good voucher specimens take time to collect but are essential. No survey should be published without vouchers being lodged in a designated herbarium.

Some plant books have excellent voucher specimens for the species photographed. For example Cunningham et al. (1981) and Auld and Medd (1987). A number of lists of plant species for various areas also have large numbers of voucher specimens lodged at various herbaria, for example McBarron (1955), Williams (1979) and Hosking (1990).

Misuse of voucher specimens It is essential that the policy of a herbarium with regard to specimens is understood. In some cases the number of specimens collected over time has been used to indicate whether a weed problem is increasing or decreasing. This is of little use if a herbarium considers that they have plenty of specimens for a particular area of the State and no longer retain additional collections. Most specimens sent in for identification are not retained by herbaria. Presence or absence of a species from an area based on herbarium specimens is also fraught with danger. So-called well known species are often rarely sent in for identification resulting in absence of specimens from various locations.

Importance of correct identifications In a number of cases the correct plant identification, and an understanding of its taxonomy and biogeography are important. These are particularly important for biological control programs. For example:

1. Various strains of the blackberry rust, Phragmidium violaceum, are likely to be more effective than others on different introduced Rubus spp. in Australia (Bruzzese and Hasan 1986, Bruzzese 1995).

2. The seed-feeding weevil, Erytenna consputa Pascoe, used to control Hakea sericea in South Africa was collected from Wilsons Promitory Peninsula in the mistaken belief that this was the same plant as the one causing the problem in South Africa (Kluge and Neser 1991). Recent taxonomic study has shown that the plant from Wilsons Promitory is Hakea decurrens R. Br. (Barker 1996). Populations of this weevil collected from H. sericea from south-eastern New South Wales, from the correct plant taxon, have successfully established on this plant in South Africa.

3. Early attempts to control this Salvinia molesta were not successful because the weevil, Cyrtobagous singularis Hustache was collected from Salvinia auriculata Aubl. in the mistaken belief that the plant species were the same (Room 1986). Salvinia weevil, Cyrtobagous salviniae Calder & Sands, collected from Salvinia molesta now successfully controls this water fern in many areas around the world.

Correct identifications may also be important for chemical control of weeds. For example various Fumaria spp. appear to have different susceptibilities to herbicides (McQuinn 1990). Another example is where irrigation managers at Emerald in the 1970s noted that Vallisneria gigantea was not being controlled by the accepted concentration of acrolein (C. Julian personal communication). An investigation concluded that the 'form' of Vallisneria present in the Emerald channels had a thicker than usual leaf and required a higher dose rate. This 'form' has been known as Vallisneria spiralis var. denseserrulata Makino.

CONCLUSION
Collect voucher specimens and others will know with certainty the species being referred to. Do not collect vouchers and you may as well not publish your results.
ACKNOWLEDGMENTS
We would like to thank Bill Barker for information on Hakea species in Australia. Suggestions and comments received from Richard Carter, Jim Dellow, Richard Groves, Rick Roush and Andy Sheppard have been noted and their assistance is acknowledged.
REFERENCES
Auld, B.A. and Medd, R.W. 1987. 'Weeds.' (Inkata Press: Melbourne and Sydney).

Barker, W.R. (1996). Novelties and taxonomic notes relating to Hakea Sect. Hakea (Proteaceae), mainly of eastern Australia. Journal of the Adelaide Botanic Gardens 17, 177-209.

Bruzzese, E. (1995). Recent status of biological control of European blackberry (Rubus fruticosus agg.) in Australia. Proceedings of the Eighth International Symposium on Biological Control of Weeds, pp. 297-299.

Bruzzese, E. and Hasan S. (1986). The collection and selection in Europe of isolates of Phragmidium violaceum (Uredinales) pathogenic to species of European blackberry naturalized in Australia. Annals of Applied Biology 108, 527-25.

Cunningham, G.M., Mulham, W.E., Milthorpe, P.L. and Leigh, J.H. (1981). 'Plants of Western New South Wales.' (NSW Government Printer, Sydney).

Groves, R.H. and Cullen, J.M. (1981). Chondrilla juncea: the ecological control of a weed. In 'The ecology of pests', eds. R.L. Kitching and R.E. Jones, pp. 6-17. (CSIRO, Melbourne).

Harden, G. (1990?1993). 'Flora of New South Wales.' Volumes 1-4. (New South Wales University Press, Sydney).

Kluge, R.L. and Neser, S. (1991). Biological control of Hakea sericea (Proteaceae) in South Africa. Agriculture, Ecosystems and Environment 37, 91-113.

McBarron, E.J. (1955). An enumeration of plants in the Albury, Holbrook and Tumbarumba districts of New South Wales. Contributions from the New South Wales National Herbarium 2, 89-247.

McQuinn, D.J. (1990). Information on Fumaria species distribution in cereal crops in southern Australia and practical guidelines to aid identification. Proceedings of the Ninth Australian Weeds Conference, p. 89. Michael, P.W. (1995). A new name for a widespread and misunderstood species of Verbena (Verbenaceae). Telopea 6, 181-183.

Morrison, S.M. and Scott, J.K. (1993). Assessment of the origins of Tribulus terrestris in Australia. Proceedings of the 10th Australian Weeds Conference and 14th Asian Pacific Weed Science Society Conference, pp. 388-391.

Parsons, W.T. and Cuthbertson, E.G. (1992). 'Noxious Weeds of Australia.' (Inkata Press, Melbourne).

Room, P.M. 1986. Salvinia molesta - a floating weed and its biological control. In 'The Ecology of Exotic Animals and Plants.', ed. R.L. Kitching, pp. 164-86. (John Wiley & Sons, Brisbane).

Williams, A.R. (1979). A survey of natural pastures in the North-west Slopes of New South Wales. Technical Bulletin No. 22. (Department of Agriculture, Sydney).

  Privacy | Copyright | Disclaimer | About PlantNET | Cite PlantNET