ARTICLES RIA / Vol. 43 / N.º 2

Mechanical and chemical control of dodoneifolia (Hook. Et Arn.) Cabrera (sweet chilca)

TOLOZANO, B.1; PISANI, J.M.1; PURICELLI, E.C.2

ABSTRACT The encroachment of native or exotic woody species in rangelands results not only in cover changes of some species, but also in changes of the amount and availability of forages for livestock. Tessaria dodonei- folia (Hook. et Arn.) Cabrera (sweet chilca) is a native shrub species in , which in recent years has been mentioned as an invader of north-eastern Argentinean rangelands. The objectives of this research were to study in sweet chilca: 1) the efficacy of mechanical (shoot cutting up to 30 cm high) and chemical control treatments (picloram+triclopyr and picloram+2,4D) and the combination of both types of control, 2) the optimal control timing (onset or end of the weed season growth), 3) the optimum dose of picloram+2,4D on of different sizes, and 4) the emergence of new seedlings of sweet chilca one year after herbicide application. Chemical treatments alone or in combination with with mechanical control showed a very good control. Piclo- ram+2,4D was the most effective treatment at applied at the beginning of active growth, while other treatments did not differ between application times. Sweet chilca plants did not survive the recommended dose for other shrub species. No differences in control on seedling establishment one year after herbicide application was observed.

Keywords: herbicides, optimal dose, rangelands, Argentina.

INTRODUCTION Kunst et al., 2015), biologic, chemical and mechanical con- trol; and the combination of chemical and mechanical con- Natural grasslands account for nearly 60% of the Ar- trols (Masters and Sheley, 2001; Van Wilgen et al., 2001; gentine area (160 million hectares) (Fernandez and Busso, 1999) and are basic forage resource used by livestock during Vitelli and Pitt, 2006). most of the year. With the increase in the coverage of native Among them, mechanical and chemical controls are the and exotic woody species in these natural systems there is a most effective and selective. Mechanical control allows the replacement of grasses and a reduction in accessibility and elimination of the unwanted species by cutting the aerial forage supply for cattle (Graz, 2008; Archer, 1990; Moleele et part or extracting the from the root. However, in places al., 2002; Tighe et al., 2009). where the species has occupied large areas or is forming Woody species can be controlled in natural grasslands large patches, mechanical control can be costly and labori- using controlled fire (Willard, 1973; Vitelli and Pitt, 2006; ous (Van Wilgen et al., 2001).

1Instituto Nacional de Tecnología Agropecuaria (INTA), Estación Experimental Agropecuaria (EEA) Rafaela, Ruta 34 km 227, CP 2300, Rafaela, Santa Fe, Argentina. Correo electrónico: [email protected], [email protected] 2Universidad Nacional de Rosario, Facultad de Ciencias Agrarias, CC N.º 14, CP S2125ZAA, Zavalla, Santa Fe, Argentina. Correo electrónico: [email protected]

Received June 8, 2016 // Accepted November 29, 2016 // Published online July 19, 2017

Mechanical and chemical control of Tessaria dodoneifolia (Hook. Et Arn.) Cabrera (sweet chilca) August 2017, Argentina 5

Selective herbicides can control the undesired species The experimental units were plots of 56 m2. Treatments without negatively affecting the other species, that are part were application of the herbicides was the treatment: piclo- of the natural pasture community (Van Wilgen et al., 2001). ram + triclopyr (4.4 + 8.35%) 5 l p.c./100 l diesel; and piclo- The herbicide efficacy can be conditioned by the pheno- ram + 2.4D (24 + 6.415%), 5 l p.c./ha (chemical control); logical momentum (DiTomaso, 2000), the height and the shoot cutting at 30 cm of height with grass trimmer (me- number of stems (Jacoby et al., 1990) and the stage of chanical control); combinations of both (herbicide with cut- growth of the plants (Agbakoba and Goodin, 1969). While ting) and a control without application of control methods. herbicide application can effectively control woody species, In the mechanical control treatment, herbicide application it may also have medium- and long-term side effects on started one week after the cut of the plants. We used a 16 liter grassland communities, which should be taken into account compression backpack with a flat fan type Lurmark tip. Herbi- when evaluating control techniques. cides were applied on individual plants to the point of runoff. In Argentina there are several examples of increased cov- At the beginning and at the end of each trial we counted the erage of woody species on natural grasslands and aban- density of adult plants (height above 0.5 m) in the plot. We doned fields such as Prosopis ruscifolia (vinal) (Cabral et evaluated the effectiveness of control methods by means of al., 2003), Acacia sp. (Cozzo, 1995), Geoffroea decorticans two indicators: 1) plant tolerance and 2) mortality. (chañar) (Echeverría and Giulietti, 2002). Tolerance was defined as dry biomass of remnant green Tessaria dodoneaefolia () (sweet chilca) is a na- foliage per unit area after treatment. Dry biomass was esti- tive woody species that in recent years has been reported as mated by harvesting the foliage of two live plants per plot at an invader of pastures in the northeastern Argentina. It is com- 30 and 90 days after application (DAA). monly associated with saline soils of northern Argentina, Par- The biomass was dried in an oven at 60 °C to constant aguay, and . It reaches a height of up to 3 m. weight. Mortality was defined as percentage of average Its leaves are not eaten by cattle, except in times of extreme weed mortality in the plots of each treatment one year droughts when some tender basal shoots may be eaten. It after the application. In turn, before applying the control flowers from October to March, and its flowers produce large methods and one year after that we counted young seed- numbers of small fruits with seeds of anemocoric dispersion lings (height: less than 0.5 m) in all the plots along a strip (Burkart, 1974). of 2 x 8 m. To know the control methods that lead to a reduction in Once we defined which herbicide achieved the highest the coverage of sweet chilca would provide effective tools plant mortality, we determined the optimal application rate for the producers of weed-encroached areas. on plants of three ages. We used a statistical design of di- The aims of this work were to study in sweet chilca: 1) vided plots with an arrangement of the subunits in stripes the effectiveness of mechanical control treatments (cutting and distributed in random blocks. The main plot consisted of of shoots at 30 cm in height), chemical control treatments the age of the plants (1, 2 and 3 years). In the strips, 7 dos- (herbicide application) and the combination of both; 2) the es of herbicide (1/4x; 1/2x; x; 2x; 4x; 8x and control without optimum moment to apply control methods (onset or end application: 0x) were applied, where x is the recommend- of active growth); 3) the optimal dose of the most efficient ed dose for the control of shrubs commonly found in natural chemical treatment among those previously studied in grasslands in the northern region of the province of Santa plants of different sizes and 4) the emergence of new seed- Fe: picloram, 2.56 g of active ingredient/l and 2,4 D, 9.6 g of lings per year after the application of the treatments. active ingredient/l. Two days before herbicide application, we recorded the maximum height with green foliage, the number of live stems, MATERIALS AND METHODS the diameter of the aerial part with green foliage and the diam- The tests were conducted at the Cattle Rearing Demon- eter of the crown of the treated plants (Table 1). stration Unit [Unidad Demostrativa de Cría Bovina] La Palmi- The crown of sweet chilca has stems with basal buds and ra of the EEA INTA Rafaela (29° 46’ 40.19” S; 61° 13’ 53.21” is located on the ground, almost at the surface level. The di- W) in the district of Las Avispas, province of Santa Fe, Argen- ameter of the crown is defined as the average of the length tina. This unit has 600 hectares affected to a large extent by of two perpendicular lines that join the outermost stems of sweet chilca. The soil is natric (natracualf or natracuol), ag- the plant at ground level. The optimum dose was evaluated ricultural land suitability 6 or 7. Average annual rainfall over by counting the dead plants one year after the application the last 50 years was 1103 mm, with a minimum of 11 mm of the herbicide. during August (dry season) and a maximum of 241 mm in The live biomass of the treated plants and mortality were February (wet season). analyzed by ANOVA and the means were compared with a We conducted the trials in areas highly covered with Tukey test (Infostat, 2008) with a significance level of 5%. sweet chilca coverage of this (2857 plants/ha) in 2008/2009, The normality of the variables was tested using the Shap- at the beginning (spring) and at the end (autumn) of the iro Wilks test and the homogeneity of variances using the growth season of this plant. We used a randomized com- Levenne test. The variables that did not comply with the plete block design with 4 replicates per treatment. ANOVA assumptions were transformed by square root.

TOLOZANO, B.1; PISANI, J.M.1; PURICELLI, E.C.2 ARTICLES RIA / Vol. 43 / N.º 2

Age (years) Height (m) Number of stems Diameter of the crown (cm) Diameter of the canopy (cm) 1 1,1 (0,05) 1,4 (0,13) 0,0 (0,00) 15 (0,03) 2 1,3 (0,12) 2,8 (0,64) 19 (0,04) 38 (0,06) 3 2,3 (0,06) 23,6 (1,66) 27 (0,01) 90 (0,03)

Table 1. Age, height, number of stems, crown and canopy diameter (mean + SE) of the plants of Tessaria dodoneifolia before being ex- posed to 7 doses of picloram + 2,4D.

180 c 160

) 140 -1

a *

h 120 . g k

( 100 80 60 b Biomass b 40 * 20 a a a a a a a 0 CM CM+Pic+Tric CM+Pic+2,4D Pic+Tric Pic+2,4D Treatments

30 days 90 days

Figure 1. Biomass of plants of Tessaria dodoneifolia at a) 30 and b) 90 days after the application of treatments at the beginning of the growth season. The bars on the columns correspond to the standard error. Different letters indicate significant differences (p <0.05; Tukey’s test) among the treatments for each harvest date. Asterisks on the columns indicate significant differences (p <0.05; t Test) between the means of each treatment at 30 and 90 days after the application of the treatments. CM= mechanical control; CM + pic + tric= mechanical control + picloram + triclopyr; CM + pic + 2,4D= mechanical control + picloram + 2,4D; pic + tric= picloram + triclopyr; pic + 2,4D= picloram + 2,4D.

A comparison of the two application times for herbicides The biomass of the remaining plants was greater at 90 (picloram + triclopyr and picloram + 2,4D) (Infostat, 2008) DAA than at 30 DAA. In the remaining treatments the bio- was performed using a t-test. mass was the same in both dates. At 30 DAA biomass was greater for mechanical control, while at 90 DAA the maxi- The number of young sweet chilca plants that settled a year mum biomass was observed in mechanical control followed after the application of the control treatments was analyzed by by mechanical control combined with picloram + triclopyr. ANOVA with a Tukey mean comparison test (Infostat, 2008). The data were transformed by square root. One year after application of treatments at the beginning of active growth plant mortality was not the same in all treat- ments (p <0.05) (Figure 2). RESULTS No dead plants were found under mechanical control. We report an interaction between the treatments per- Mortality was higher (p <0.05) in plants treated with me- formed at the beginning of the active growth and in the chanical control in combination with picloram + 2,4D, and tolerance to them (Figure 1a and b). Adequate control of individually with picloram + 2,4D. the biomass with every treatment - except for mechanical As a consequence of successive frosts, in the treatments control - was obtained at 30 DAA, and at 90 DAA in every applied at the end of the active growth it was only possible treatment except mechanical control and the combination to harvest the biomass at 30 DAA (Figure 3). The greatest of mechanical control and picloram + triclopyr. biomass was reported under mechanical control, which sig-

Mechanical and chemical control of Tessaria dodoneifolia (Hook. Et Arn.) Cabrera (sweet chilca) August 2017, Argentina 7

100 c c 90 bc 80 70 b 60 50 40

Mortality (%) 30 20 10 a 0 CM CM+Pic+Tric CM+Pic+2,4D Pic+Tric Pic+2,4D Treatments

Figure 2. Mortality of Tessaria dodoneifolia plants a year after the application of the treatments at the beginning of active growth. The bars on the columns correspond to the standard error. Different letters indicate significant differences (p <0.05, Tukey’s test). CM= mechanical control; CM + Pic + Tric= mechanical control + picloram + triclopyr; CM + Pic + 2,4D= mechanical control + picloram + 2,4D; Pic + Tric= picloram + triclopyr; Pic + 2,4D= picloram + 2,4D.

160

140 c

) 120 -1 a h

. 100 g k 80 bc

60

Biomass ( 40 ab 20 a a 0 CM CM+Pic+Tric CM+Pic+2,4D Pic+Tric Pic+2,4D Treatments

Figure 3. Plant regrowth biomass of Tessaria dodoneifolia, 30 days after the application of treatments at the end of the growth season. Bars on the columns correspond to the standard error. Different letters indicate significant differences (p<0.05, Tukey’s test). CM= mecha- nical control; CM + pic + tric= mechanical control + picloram + triclopyr; CM + pic + 2,4D= mechanical control + picloram + 2,4D; pic + tric= picloram + triclopyr; pic + 2,4D= picloram + 2,4D.

nificantly differed from the rest of the treatments, except When we compared the mortality caused by the treat- for the combination of mechanical control and picloram + ments applied at the beginning and at the end of the active triclopyr. Plants were less tolerant to the combination of me- growth, mortality was higher at the beginning of the growing chanical control and picloram + 2,4D, and picloram + 2,4D season for the combination of mechanical control and piclo- ram + 2,4D and did not differ between seasons for the rest individually. of the treatments. One year after the application of the control treatments at At the beginning of active growth, the combination of me- the end of the growth cycle, no dead plants were recorded chanical control and picloram + 2,4D resulted in a mortality in mechanical control (Figure 4). The highest mortality oc- rate of 90%. The same occurred for the application of this curred with picloram + triclopyr and picloram + 2,4D without herbicide individually (control: 88%). The mechanical con- mechanical cutting. trol without the application of herbicides did not resulted in

TOLOZANO, B.1; PISANI, J.M.1; PURICELLI, E.C.2 ARTICLES RIA / Vol. 43 / N.º 2

120

c 100 c

80

60 b

40 b Mortality (%)

20

a 0 CM CM+Pic+Tric CM+Pic+2,4D Pic+Tric Pic+2,4D Treatments

Figure 4. Mortality of Tessaria dodoneifoli plants a year after the application of the treatments at the end of the growth season. Bars on the columns correspond to the standard error. Different letters indicate significant differences (p<0.05, Tukey’s test). CM= mechanical control; CM + pic + tric= mechanical control + picloram + triclopyr; CM + pic + 2,4D= mechanical control + picloram + 2,4D; pic + tric= picloram + triclopyr; pic + 2,4D= picloram + 2,4D.

All control plants (0X) of different ages survived. One- Treatments Seedlings density ± (SE) year plants did not survive any of the remaining doses. T 0,1 (0,09) Survival was observed in plants two years old to 1/4X, and CM 0,0 (0,03) three years old to 1/4X and 1/2X. At higher doses, no plants CM+Pic+Tric 0,3 (0,25) survived. CM+Pic+2,4D 0,2 (0,07) Pic+Tric 0,3 (0,13) Pic+2,4D 0,2 (0,13) DISCUSSION

Table 2. Density of Tessaria dodoneifolia seedlings (plants/m2) a At the end of the active growth, plant mortality under piclo- year after the application of treatments ± standard error (SE). T= ram + 2,4D and picloram + triclopyr was above 90%. As at control; CM= mechanical control; CM+ Pic + Tric= mechanical con- the beginning of active growth, no mortality was observed in trol + picloram + triclopyr; CM + Pic + 2,4D= mechanical control plants under mechanical control alone, and the plants pre- + picloram + 2,4D; Pic + Tric= picloram + triclopyr; Pic + 2,4D= picloram + 2,4D. sented the highest remnant live biomass. This indicates that the cutting of individuals at 30 cm does not reduce the sweet chilca cover at any time. The plants have the stocks stored in the roots which allows the regrowth after the mechanical control. In another study, mechanical control over other As- teraceae species proved laborious and costly, and spawned the mortality of the plants, and the plants exposed to this regrowth after cutting. treatment presented the highest remnant live biomass. However, it was possible to control the species by cut- No young seedlings (height below 0.5 m) were found be- ting them twice a year when regrowth reached 0.5 meters fore application of the treatments. At one year, it was pos- in height. Results were obtained after approximately three sible to observe young seedlings, but no treatment effect years of successive cuttings (Marchesini, 2003). was found (p> 0.05) (Table 2). The density of seedlings be- Under chemical control the mortality was high both at the low 0.5 m was not associated with the application of control beginning and at the end of the active growth of sweet chilca. treatments. One year after application of picloram + tric- Although these treatments were not more effective in relation lopyr and picloram + 2,4D, we found the same density of to the time of application, there was a trend towards greater young seedlings. control for by the end of active growth (early autumn) (Fig- Both of the herbicides tested in this work were effective to ures 2 and 4). This was also reported in the control of Ru- control sweet chilca. Although the mortality was not statisti- bus ulmifolius, where the application of herbicides at the end cally different, picloram + 2.4D achieved a higher mortality of summer-early autumn resulted in a greater control of the rate and was the herbicide selected for the dose optimiza- species when compared with spring applications (Mazzolari tion test on plants of different ages. et al., 2011).

Mechanical and chemical control of Tessaria dodoneifolia (Hook. Et Arn.) Cabrera (sweet chilca) August 2017, Argentina 9

This could be due to the fact that at that moment the by Harrington and Miller (2005), it was shown that triclopyr plant starts a translocation of the assimilates towards the would have no activity in the soil and would present little roots and along with them it would be also transporting the risk to the associated vegetation. herbicides. In this way, more herbicide would arrive to the One to five years old plants of Rubus fruticosus did not roots, resulting in a greater mortality. Similar results were survive the recommended dose of 2,4,5 T and to picloram obtained by Harrington and Miller (2005) when studying or higher doses (Amor, 1974). Older plants were more tol- the control of Ligustrum sinense. These authors reported erant than younger plants. Similarly, the translocation of a lower control of the species during the spring because 2,4D and picloram in Convolvulus arvensis plants at 3, 7 the plants are actively growing and the translocation of as- and 16 weeks was higher in the seedling stage (Agbakoba similates would flow to the aerial parts of the plant and not and Goodin, 1969). towards the roots. These results are in agreement with those reported by Mechanical treatments allow reducing the height and Jacoby et al. (1990) and Jacoby et al. (1990) in Prosopis coverage of sweet chilca and thus facilitate a more uniform glandulosa var. Glandulosa, who observed that the plants application of the herbicides. Usually the combination of with the highest number of stems were less controlled by the mechanical treatment and the subsequent application the herbicides. In this study, three-year-old plants were high- of herbicides resulted in a lower mortality (<50%) than the er and had a greater number of stems, and consequently a individual application of herbicides. This could be attribut- larger diameter of crown and canopy. ed to the lesser remnant foliage remaining in the pruned plants when compared to those unpruned. Nevertheless, The larger diameter of the canopy could reduce the pen- this does not explain why the mechanical cut and picloram etration of the herbicide in the plant (Jacoby et al., 1990a + 2,4D treatment at the beginning of the growing season and b). In this way, the greater tolerance to herbicide doses was so effective. could be explained by the larger size. On the one hand, in Solanum glaucophyllum the cutting of the plants favors the action of systemic herbicides be- CONCLUSIONS cause the leaf/stem ratio is increased and this results in a greater receptivity and penetrance of the herbicide in the There was no mortality in plants under mechanical control plant (Bertín and Cepeda, 2007). On the other hand, the either at the beginning or at the end of the growth season. cutting of the plants could induce the mobilization of the re- Chemical treatments were the most effective, both at the serves to recover the lost biomass, thus generating a trans- beginning and at the end of the growth season. location towards the aerial part and not towards the roots. No differential effect of control treatments on the sub- Gonzaga (1998) suggests that the combination of me- sequent establishment of sweet chilca seedlings was ob- chanical and chemical control is more efficient than the served. application of herbicides without mechanical control. In this study, at the beginning of the growing season the efficacy of There was no plant survival at the recommended dose the combination of mechanical control and picloram + 2,4D for shrubs of picloram + 2,4D. One-year-old plants did not was not different from the one achieved by applying the two survive any of the tested doses, while 2-year and 3-year-old herbicides without mechanical control. plants survived at lower doses than the ones recommended. The cutting of sweet chilca plants prior to application of herbicides may be less effective because there is a reduc- ACKNOWLEDGEMENTS tion in the herbicide absorption surface. The combination of mechanical control and glyphosate on Eupatorium boniifo- To the National Scientific and Technical Research Coun- liun, E. buniifolium, Baccharis spicata, B. dracundifolia, B. cil (CONICET) for the fellowships granted to Barbara pingraea, B. punctulata and E. laevigatum plants, with a Tolozano. To the National Institute of Agricultural Technolo- regrowth of 0.5 m, reduced the coverage by almost 100% 5 gy (INTA) for the economic support to conduct this research months after the cutting (Marchesini, 2003). (Projects SANFE05, AEFP 263051). To Alejandra Cuatrín for her advice on statistical analysis. After annual applications of picloram (0.56 kg/ha) for five years Lym and Messersmith (1987) did not find accumulation at the soil surface and the persistence of the herbicide would BIBLIOGRAPHY be related to the rainfall occurred after the application and to the type of soil. In this way, greater amounts of picloram per- AGBAKOBA, C.; GOODIN, J. 1969. Effects of stage growth of 14 sisted in the soil in the driest years. On this basis, this herbi- field Bindweed on absorption and translocation of C-Labeled 2,4- D and Picloram. Weed Science 17: 436-438. cide would not have an effect on the establishment of new seedlings in the pasture community. AMOR, R. 1974. Ecology and control of blackberry (Rubus fruti- cosus L. agg.). Weed Research 14 (4): 239-243. In this study, rainfall in 2008 was 339 mm, while in 2009 it ARCHER, S. 1990. Development and stability of grass/woody was 886 mm. The greatest rainfall occurred during 2009 fa- mosaics in a subtropical savannaparkland, Texas, U.S.A. Journal vored the establishment of new weed seedlings. In a study of Biogeography 17: 453-462.

TOLOZANO, B.1; PISANI, J.M.1; PURICELLI, E.C.2 ARTICLES RIA / Vol. 43 / N.º 2

BERTÍN, O.; CEPEDA, S. 2007. Defoliación y control químico de JACOBY, P.; MEADORS, C.; ANSLEY, R. 1990b. Control of Ho- duraznillo blanco (Solanum glaucophyllum) en pastizales natura- ney Mesquite with Herbicides: Influence of Plant Height. Journal of les. Revista Argentina de Producción Animal 27(2): 67-74. Range Management. 43 (1): 33-34. BURKART, A. 1974. Flora ilustrada de Entre Ríos (Argentina). KUNST, C.; LEDESMA, R.; BRAVO, S.; DEFOSSÉ, G.; GODOY, Parte VI: Dicotiledóneas metacamideas (gamopétalas) B; Rubia- J.; NAVARRETE, V.; JAIME, N. 2015. Dinámica del contenido de les, Cucurbitales, Campanulales (incluso Compuestas). Colección humedad de pastos y su relación con la ecología del fuego en re- Científica del INTA, Buenos Aires, p. 554. gión chaqueña occidental (Argentina). Revista de Investigaciones CABRAL, A.; DE MIGUEL, J.; RESCIA, A.; SCHMITZ, M.; PI- Agropecuarias 41:83-93. NEDA, F. 2003. Shrub encroachment in Argentinean savannas. LYM, R.; MESSERSMITH, C. 1987. Leafy Spurge Control and Journal of Vegetation Science 14: 145-152. Herbicide Residue From Annual Picloram and 2,4-D Application. COZZO, D. 1995. Interpretación forestal del sistema fachinal de Journal of Range Management 40: 184-198. la Argentina y faxinal del Brasil. Quebracho 3: 5-12. MARCHESINI, E. 2003. Control de chilcas. Sitio argentino de DITOMASO, J. 2000. Invasive weeds in rangelands: Species, producción animal. Disponible: http://www.produccion-animal. impacts, and management. Weed Science 48(2): 255-265. com.ar/produccion_y_manejo_pasturas/pasturas_naturales_ ECHEVERRíA, J., GIULIETTI, J., 2002. Incidencia del Chañar especies/26-control_chilcas.pdf verificado: mayo de 2016. en la producción bovina en San Luis. Revista de Investigaciones MASTERS, R.; SHELEY, R. 2001. Invited Synthesis Paper: Prin- Agropecuarias 30: 59–66. ciples and practices for managing rangeland invasive plants. Jour- FERNÁNDEZ, O.; BUSSO, C. 1999. Arid and semi-arid range- nal of Range Management 54: 502-517. lands: two thirds of Argentina. En: ARNALDS, O.; ARCHER, S. MAZZOLARI, A.; COMPARATORE, V.; BEDMAR, F. 2011. Con- (Eds.). Case studies of Rangelands Desertifications. Agricultural trol of elmleaf blackberry invasion in a natural reserve in Argentina. Research Institute Report N.° 200. Reykjavik Iceland, pp: 41-60. Journal for Nature Conservation 19(3): 185-191. GONZAGA, S.S. 1998. Controle de Plantas Invasoras (melho- MOLEELE, N.; RINGROSE, S.; MATHESON, W.; VANDER- ramento do campo nativo visando o aumento na capacidade de POST, C. 2002. More woody plants? The status of bush encroa- suporte da pastagem natural, através de práticas de manejo). En: chment in Botswana’s grazing areas. Journal of Environmental EMBRAPA. Produção de Carne de Qualidade para o Rio Grande Management 64: 3-11. do Sul, Santa Catarina e Paraná. Bagé: CPPSUL, pp.78-94. GRAZ, F. 2008. The woody weed encroachment puzzle: gathe- TIGHE, M.; REID, N.; WILSON, B.; BRIGGS, S. 2009. Invasive ring pieces. Ecohydrology 1: 340-348. native scrub and soil condition in semi-arid south-eastern Australia. Agriculture, Ecosystems y Environment 132: 212-222. HARRINGTON, T.; MILLER, J. 2005. Effects of Application Rate, Timing, and Formulation of Glyphosate and Triclopyr on Control of VAN WILGEN, B.; RICHARDSON, D.; HIGGINS, S. 2001. In- Chinese Privet (Ligustrum sinense). Weed Technology 19: 47-54. tegrated control of invasive alien plants in terrestrial ecosystems. Land Use and Water Resources Research 1: 1-6. INFOSTAT. 2008. Manual del usuario. Grupo InfoStat, FCA, Universidad Nacional de Córdoba. Versión 2008, Editorial Brujas VITELLI, J.; PITT, J. 2006. Assessment of current weed control Argentina, p. 318. methods relevant to the management of the biodiversity of Austra- JACOBY, P.; ANSLEY, R.; MEADORS, C.; CUOMO, A. 1990a. lian rangelands. Rangeland Journal. 28: 37-46. Control of honey mesquite with herbicides: influence of stem num- WILLARD, E. 1973. Effect of wildfires on woody species in the Mon- ber. Journal of Range Management. 43 (1): 36-38. te region of Argentina. Journal of Range Management 26: 97-100.

Mechanical and chemical control of Tessaria dodoneifolia (Hook. Et Arn.) Cabrera (sweet chilca)