Weeds are one of several harmful factors affecting the yield of cultivated plants. Weed flora and vegetation on plantations of medicinal plant crops have been sufficiently investigated neither in our country nor around the world. However, weed control in medicinal plant crops, and application of herbicides in particular, is associated with a tendency to provoke environmental and ecotoxicological effects. Considering the fact that medicinal plant production is currently undergoing expansion, there is a need for floristic and phytocoenological research of weeds and weed communities in cultivated medicinal plant species. The perennial weeds cause even more serious damage in medicinal plant crops, and therefore the comprehension of their appearance, development and control should be treated with special attention.

Qualitative and quantitative analysis of perennial weed plants in several medicinal plant crops was carried out in the fields of the Institute for Medicinal Plant Research "Dr Josif Pančić" in Pančevo near Belgrade. The Braun-Blanquet Scheme and transformed the data on weed abundance and coverage according to a scale by Weshoff and van der Maarel were applied for the purpose of an analysis of perennial weeds determined in medicinal plant crops. Our investigation showed that only two perennial weed species, such as Sorghum halepense and Convolvulus arvensis are represented in all medicinal plant crops investigated. Other perennial weeds could be considered as pretty frequent ones, as: Cirsium arvense, Sonchus asper, and Capsella bursa pastoris. The total of 24 perennial weeds has been identified in different medicinal plant crops.

The analysis of perennial weeds affecting the growth of medicinal plant species under the controlled conditions, has included the strategies of coping against them, where different forms of perennial weed control, such as agrotechnical, chemical and biological were being discussed.

Key words: medicinal plant crops, perennial weeds, weed control.

Perennial weed species, especially geophytes, have become more frequent, therefore causing serious problems in agrophytocoenoses. The increase of their appearance could be connected with permanent and of long application of herbicides resulting in reduction of floristic composition of agrestal communities (Ajder, 1992). Remarkable changes, as a consequence of long-term herbicide usage, may be summarized as: 1) withdrawal (retirement) or, in some cases, total extermination of sensitive species, like Agrostemma githago and others, which have disappeared from northern Serbia (Vojvodina); 2) expansion of tolerant perennial weed species, such as: Sorghum halepense, Cirsium arvense, Convolvulus arvensis, Sonchus arvensis, Cynodon dactylon, Agropyrum repens, etc.; 3) widening of resistant plants from the group of therophytes, which were at the beginning of herbicide utilization under the full control (Amaranthus retroflexus, Chenopodium album, Abutilon theophrasti, Datura stramonium, Hibiscus trionum), and 4) appearance and invasion of alochtone species that were at the first time ruderal ones, moving later into the cultivated areas because their competitive abilities (Ambrosia artemisifolia, Iva xantifolia, Erigeron canadensis).

Perennial weeds that belong to the geophytes, represent the most problematic category on cultivated land (Richard, 1998, Miller and Lym, 1998). Their main characteristics are developed rhizomes, strong roots and high reproductive and regenerative ability (Wiese and Bean, 1997).

Damages that are caused by perennial weeds in agro-phytocoenoses, as well as in medicinal plant plantations are serious and may be summarized, as follows: 1) their powerful and numerous subterraneous shoots interfering with mechanical cultivation; 2) they are better competitors for light, water, mineral elements and vegetative space comparing to crops; 3) weeds decrease yield significantly; 4) expenses in crop production are getting higher because of weeds (Radosevich et al., 1996).

The status and problems in perennial weed control is similar as in row crops, plantations and therefore in medicinal plant crops. In medicinal plant plantations the weed problem is even more complex, regarding non-chemical methods of weed controlling, such as: physical and biological ones. Unfortunately, on large medicinal plant plantations, is not always possible to keep pure crop production without herbicide application.

The problem of weeds in medicinal plant crops have not been enough investigated, although weeds are there the most harmful element. There are few recent data concerning weed control in medicinal plant crops (Ivanović et al., 1997, 1998, Vrbničanin et al., 1998).

Considering the significance of mentioned problem, the aim of this work was to establish quantitative and qualitative participation of perennial weeds in four different medicinal plant crops (chamomile, parsley, mustard and dill) of Institute for Medicinal Plant Research, in Pančevo, near Belgrade, for creating a proposal of weed control.

Floristic and phytocoenological surveys of perennial weeds in four different crops of medicinal plants: Sinapis alba L. (mustard), Petroselinum crispum L. (parsley), Anethum graveolens L. (dill) and Chamomilla recutita (L.) Rausch., (chamomile) were carried out over 1998 and 1999 vegetation season. It was applied Braun-Blanquet scheme (1964) and transformed the data on weed abundance and coverage according to a scale by Westhoff and van der Maarel (1973). Determination of weeds was based on Josifović (ed.) (1971-1980). Species life forms were determined according to Raunkiaer (1934) and Ellenberg (1974).

In plantations of chamomile, mustard, dill and parsley during two vegetation seasons an investigation has been carried out on agrestal flora and vegetation. The total of 69 weed species was detected. Perennial weeds, geophytes and hemicryptophytes, are presented with the total of 24 species (Tab. 1). The intermediate (transitional) category of biannual weeds (14 species) belonging to thero-hemicryptophytes, may be also joined to the perennial weed group.

Results of qualitative and quantitative participation of perennial weeds in medicinal plant crops, i.e. abundance and coverage, showed that the most frequent species were: Sorghum halepense (johnsongrass), Convolvulus arvensis (field bindweed), Cirsium arvense (Canada thistle), Agropyrum repens (common couch), Lolium perenne (perennial rye-grass) and Rumex crispus (curled dock). Simultaneously, this is the most harmful weed group concerning the aspect of their control on medicinal plant plantations. Regarding the individual medicinal plant crops, in chamomile was registered the highest weeding (abundance and coverage from 1 to 5). The fact that chamomile crop was the most weeded could be explained by the fact that it was grown for several seasons at the same area (place). The most frequent weed species: johnsongrass (2-5), field bindweed (3-5), Canada thistle (2-3) and curled dock (3-5). These species, with an exception of curled dock, belong to the geophytes, which are characterized by the presence of strongly developed rhizomes, extensive perennial root system, or stolons, all together serving for fast widening and vegetative reproduction. Therefore, perennial weeds have better competitive ability than annual plants, conquering new areas and even suppressing other weed species and crop (Ghosheh and Chandler, 1998). Especially harmful is johnsongrass which reproduces from both seeds and rhizomes and is more competitive in agricultural crops when originating from rhizomes (Lopez, 1988). Rhizomatous johnsongrass tends to escape complete control from mechanical and herbicide treatment because of the large number of rhizome buds capable of propagation (Millhollon, 1985). According to the weeding level, at the second place was parsley crop where also dominated johnsongrass (3), field bindweed (3) and Canada thistle (2-3). Comparing all medicinal plant crops investigated, johnsongrass, Canada thistle and field bindweed could be set apart as most frequent and harmful.

A biological spectrum based on qualitative (floristic) data alone cannot provide full insight into the state of plant community (Kojić et al., 1997). It is possible to clearly differentiate the biological spectrum of vegetation. Our results on the life spectrum of vegetation (Fig. 1 B), contrasted with the life spectrum of flora (Fig 1 A), also revealed a slightly different situation concerning conditions in the habitat. Regarding therophytes (50.54%) and therohemi-cryptophytes (19.67 %), the situation is nearly identical to that of the life spectrum of flora, but data on the qualitative presence of geophytes and hemicryptophytes are totally different. In the life spectrum of vegetation, geophytes are more dominant (17.69%) than hemicryptophytes (12.09%), which creates a somewhat different account of conditions existing in the habitat of the analyzed phytocoenoses. The diversion (from the life spectrum of flora) might reflect a high degree of crop cultivation, primarily by application of intensive agrotechnological measures. High participation of therophytes (44.29% and 50.54%) and therohemicryptophytes (20% and 19.67%) suggests discernible presence of anthropogenic factor, which is feature of agrestal plant communities (Ajder, 1996). Lower participation of geophytes (12.86% or 17.69%) and hemicryptophytes (22.86% or 12.09%) shows some instability of weed flora and vegetation, i.e. high degree of human influence.

On the basis of available data (Miller and Lym, 1998, Richard, 1998, Ghosheh and Chandler, 1998, Vogelgsong et al., 1998), as well to our own experience, we recommend that for ecologically healthy medicinal plant crop production, the following weed control techniques recommended: crop rotation, tillage, fall applied herbicides and applied pre-emergence bio-herbicide (Vogelgsang et al., 1998). In our opinion a possible way for weed controlling could be composting with residues of particular medicinal plant material, for example, after distillation. Anyway, this will be subject of our further research.

Figure 1. Life forms of weed flora and vegetation determined in medicinal plant crops
A)

B)

H-hemicryptophytes; G-geophytes; T-therophytes; TH-thero-hemicryptophyes

CONCLUSION

On medicinal plant plantations, such as chamomile, dill, mustard and parsley, there was the total of 24 perennial weed species identified. Among them the Sorghum halepense (johnsongrass), Convolvulus arvensis (field bindweed), Cirsium arvense (Canada thistle) and Rumex crispus (curled dock) were the most frequent ones. Chamomile and parsley crop were the most weeded comparing to the other two plantations, regarding the presence of perennial weeds. Geophytes, such as johnsongras, field bindweed, Canada thistle, common couch, perennial sow thistle, corn mint and others, are better competitors for water, light, mineral elements and vegetation space than medicinal plant crops. Highly competitive level may be due to the presence of good developed rhizomes and perennial roots which express great both reproductive and regenerative ability.

For the purpose of perennial weed controlling on medicinal plant plantations, it is necessary to apply an integral system of plant crop protection, including crop rotation, tillage, fall applied herbicides, rosette technique, composting with residues of medicinal plant material, as well as application of preemergence bioherbicide, like Phomopsis convolvulus.

1. Ajder S. (1992): The influence of some herbicides on phytocoenological changes of corn crop weed community. Review of Research Work of faculty of Agriculture 37 (1), 37-47.

2. Ajder S. (1996): Fitocenološke i ekofiziološke promene korovske vegetacije strnih žita u zavisnosti od visinskog gradijenta. Doktorska disertacija, Poljoprivredni fakultet, Beograd.

3. Braun-Blanquet J. (1964): Pflanzensociologie. Grundzuge der Vegetationskunde. 3. Aufl., Springer-Verlag, Berlin.

4. Ellenberg H. (1986): Vegetation Mitteleuropas mit den Alpen. Ulmer Stuttgart.

5. Gajić M. (1984): Florni elementi SR Srbije. In: Sarić M.R. (ed.) Vegetacija SR Srbije I - opšti deo, SANU, Beograd, pp.: 317-397.

6. Ghosheh H. and Chandler J. (1998): Johnsongrass (Sorghum halepense) control systems for field corn (Zea mays) utilizing crop rotation and herbicides. Weed Technology 12, 623-630.

7. Ivanović M., Stepić R., Vukša P. (1997): Zaštita kamilice od bolesti, štetočina i korova. In: Kojić M. (ed.), Kamilica (Chamomilla recutita (L.) Rausch). - monografska studija. Ins. za proučavanje lekovitog bilja "Dr Josif Pančić", Beograd, pp.: 120-127.

8. Ivanović M., Spasić R., Stepić R., Babović M., Petrović O., Kostić M., Krnjaja V., Jakovljević D. (1998): Zaštita pitome nane od štetnih bioagenasa. In: Kojić M. and Jančić R. (ed.), Pitoma nana i druge vrste roda Mentha L. - monografska studija. Ins. za proučavanje lekovitog bilja "Dr Josif Pančić", Beograd, pp.: 177-192.

9. Josifović M. (ed.) (1971-1980): Flora SR Srbije, SANU, Beograd.

10. Kojić M., Popović R., Karadžić B. (1997): Vaskularne biljke Srbije kao indikatori staništa. Ins. za istraživanja u poljoprivredi "Srbija" i IBIS, Beograd.

11. Lopez J.A. (1988): Biological aspects and control of johnsongrass (Sorghum halepense) in grain sorghum (Sorghum bicolor). M.S. thesis. Texas A&M University, College Station, TX, 118 p.

12. Miller B. and Lym R. (1998): Using the rosette technique for Canada thistle (Cirsium arvense) control in row crops. Weed Technology 12, 699-706.

13. Millhollon R.W. (1985): Progressive kill of rhizomatous johnsongrass (Sorghum halepense) from repeated treatment with dalapon, MSMA or asulam. Weed Science 33, 216-221.

14. Radosevich S., Holt J. and Gherse C. (1996): Weed ecology, Implications for management. John Wiley and Sons, Inc., New York.

15. Raunkiaer C. (1934): The life forms of plants and statistical plant geography. Calderon Press, Oxford.

16. Richard E. (1998): Control of Perennial Bermudagrass (Cynodon dactylon) and Johnsongrass (Sorghum halepense) in sugarcane (Saccharum ssp. hybrids). Weed Technology, 12, 128-133.

17. Vogelgsang S., Watson A.K., Di Tommaso A. (1998): Effect of soil incorporation and dose on control of field bindweed (Convolvulus arvensis) with the preemergence bioherbicide Phomopsis convolvulus. Weed Science 46, 690-697.

18. Vrbničanin S.,DajićZ.andJevdjovićR.(1998):Preliminaryfloristic and phytocoenological investigation of weeds in medicinal plant crops. Acta Herbologica 7 (1-2), 81-90.

19. Wiese A. and Bean B.W. (1997): Economic evaluation of field bindweed (Convolvulus arvensis) control. Weed Science 45, 288-295.

20. Westhoff V. and Van der Maarel E. (1973): The Braun-Blanquet approach. In: Whittaker R.H. (ed.). Handbook of vegetation Sci., 5, 616-726.