03002naa a2200289 a 450000100080000000500110000800800410001902400270006010000180008724501380010526000090024350000830025252020590033565000230239465000130241765000130243065300150244365300250245865300140248365300310249765300240252865300120255270000170256470000180258170000150259977300980261410610662021-03-10       2020    bl uuuu     u00u1 u    #d7 a10.1111/aec.129042DOI1 aQUIÑONES, A.  aWeed control in natural grasslandsbA case study using a perennial native forb from the South American Campos.h[electronic resource]  c2020  aArticle history: Accepted for publication April 2020. Published November 2020.  aAbstract Management of native weeds in natural grasslands is a challenging task. Often, recommendations are based on short?term studies of the response of weed cover and density to different control methods. However, perennial species well adapted to disturbances typically recover from commonly used control techniques. In this long?term study, we focused on a rosette native to the South American Campos : Eryngium horridum Malme (Apiaceae). This plant is strongly avoided by cattle due to its spiny leaves and tends to form dense patches, reducing the available grazing area. We aimed at understanding how key demographic processes, such as size?related plant survival and seedling establishment, are affected by different control treatments. For this, E. horridum cover, density and size structure were assessed over three years in response to mechanical, chemical and integrated (mechanical + chemical) control methods. In a field experiment, we used a weighted rim and a wiper applicator with 2,4?D + picloram for the mechanical and chemical control, respectively. Cover was reduced by control treatments (?control phase?), but this was not sustained in the long term (?recovery phase?). Regardless of the method used, control success was closely related to effects on population size structure. Mortality was high and rapid in large rosettes, which effectively led to a rapid and widespread cover reduction in all control treatments. However, only herbicide reduced rosette density delaying the recovery phase. Seedling density was low during the experimental period and scarcely affected by treatments. We conclude that cover reduction depends on removing all large rosettes, but recovery is related both to the size of the remaining pool of small rosettes and to the ability of buried rhizomes to resprout. Finally, we highlight the importance of finding a balance between productive goals and biodiversity conservation. In that context, integrated control successfully reduced cover, delayed recovery and minimised the amount of herbicide used.  aCONTROL DE MALEZAS  aFORRAJES  aPASTURAS  aDEMOGRAPHY  aESPECIES INDESEABLES  aGRASSLAND  aRIO DE LA PLATA GRASSLANDS  aUNDESIRABLE SPECIES  aURUGUAY1 aLATTANZI, F.1 aSALDAIN, N.E.1 aLEZAMA, F.  tAustral Ecology, November 2020, vol 45 (7), p. 909-920. Doi:https://doi.org/10.1111/aec.12904