New records of the invasive macrophyte, Urochloa arrecta extend its range to eastern Brazilian Amazon altered freshwater ecosystems

Invasive species influence the structure and functioning of ecosystems, as they affect native species, significantly decreasing their diversity. Aquatic ecosystems harbor a great biodiversity, and invasive macrophytes significantly affect the native plant communities, causing a cascade effect on other trophic levels. Among invasive macrophytes, Urochloa arrecta is cause for concern in the Neotropics and is found in several regions of Brazil, specially in the southeastern and southern regions. So far the species had been recorded only in the northern state of Amazonas. We report the first record of the species in the state of Pará, in the eastern Brazilian Amazon. We emphasize that identifying sites where this species is invasive is the best strategy to prevent its spread, aiming at the protection and conservation of Amazonian freshwater ecosystems.

Freshwater ecosystems have the highest biodiversity per unit area in the planet (Balian et al. 2008), and thus, more species are threatened by invasions in aquatic than in terrestrial ecosystems (Moorhouse and Macdonald 2015). The invasion VOL. 50(2) 2020: 133 -137 ACTA AMAZONICA of rivers, lakes, wetlands, and reservoirs by macrophytes is especially harmful due to negative effects on native macrophytes (Madsen et al. 1991;Kovalenko and Dibble 2011) and their role in providing high habitat complexity and keeping biological diversity (Thomaz and Cunha 2010). By affecting native species, invasive macrophytes may cause cascade effects on other trophic levels (Kovalenko and Dibble 2011), reducing animal, plant, and microbial diversity (Lougheed et al. 2008;Thomaz and Cunha 2010), and by affecting structural ecosystem properties (Bunn et al. 1998;Cuassolo et al. 2016).
Grasses can be classified as highly invasive plants and a model group to understand invasion processes, given their resistance to stress and widespread distribution (D'antonio and Vitousek 1992;McKinney and Lockwood 1999). Poaceae are actively moved around by humans, can effectively compete with native species in a wide range of ecosystems, and can reduce native diversity and change ecosystem processes ranging from nutrient cycling to regional microclimate (D'antonio and Vitousek, 1992;McKinney and Lockwood 1999;Amorim et al. 2015). The Poaceae genus Urochloa has already shown great invasive potential, both in terrestrial (Williams and Baruch 2000) and aquatic ecosystems (Thomaz et al. 2009;Michelan et al. 2010).
Urochloa arrecta (Hack. ex T.Durand & Schinz) Morrone & Zuloaga is an aquatic species native to Africa, which has infested tropical and subtropical zones around the world (Amorim et al. 2015). It is a perennial grass with long, floating branches, forming thick mats with accumulated stems, leaves and roots. Details for the identification of the species are provided by Lorenzi (2001) under the synonymous name Brachiaria subquadripara. The species can propagate by stolons (Amorim et al. 2015), and through rhizomes or any other fragment that can be carried by water flow (Pott et al. 2011;Michelan et al. 2017). It is a source of concern in hydrodam reservoirs such as Itaipu, in southern Brazil (Thomaz et al. 2009) and natural aquatic environments, such as the Pantanal (Pott et al. 2011), where it reduces native biodiversity (Michelan et al. 2010). The occurrence of U. arrecta has already been reported in anthropized areas of several Brazilian states, specially in the northeastern, southeastern, and southern regions, in different biomes, such as the Cerrado savanna and the Atlantic Forest, but also in the northern state of Amazonas, the only one so far in the Brazilian Amazon region (Flora do Brasil 2020; Figure 1).
During a field survey in July 2017, we found extensive U. arrecta mats along stream shores and small lakes at several sampling sites in the state of Pará, in Paragominas  (Figure 2), and in Belém (01º27'21"S, 48º30'16"W), in the reservoirs inside Utinga Park, which supply most of the drinkable water to Belém (Figure 2; Table 1). We collected and herborized some plant samples for species identification according to standard herbarium techniques, and all the material is deposited in the Herbarium Felisberto Camargo at Universidade Federal Rural da Amazônia. T.S. Michelan identified the material.
Most sites where we found U. arrecta are anthropisized (e.g., pasture, agriculture, mining, and urban areas), and had shorelines cleared of riparian vegetation (Table 1). Considering the geographical location of our records (Figure 1), it is possible that the species spread into Pará from the neighboring state of Maranhão, as deforestation for pasture and agriculture in Pará created suitable conditions for its dispersal and establishment. Human activity, particularly multiple land use, has likely facilitated the introduction, spreading and stabilization of U. arrecta in the surveyed aquatic ecosystems. The expansion of the grass was likely facilitated by decreased shading due to deforestation or forest thinning, increased nutrient availability (e.g. phosphorous and nitrogen) in water and soil, and altered water depth and turbidity. These impacts can disrupt biological processes in native macrophyte species (Vitousek et al, 1997;Kowarik 2003;Lougheed et al. 2008;Gołdyn 2010;Sass et al. 2010;Quinn et al. 2011;Gallardo et al. 2016). The shading by native riparian vegetation can be a limiting factor for the establishment and spreading of U. arrecta (Evangelista et al. 2017;Fares et al. 2020), though in Australia the congeneric U. mutica can occur in sites with dense riparian vegetation (Mackay et al. 2010). This means that land-use change (i.e. loss of riparian vegetation) could increase the potential spread of U. arrecta, but that the species might also potentially pose a threat to undisturbed habitats, specially in wider rivers whith undisturbed riparian vegetation, where the light penetrates in between the margins. In this case, if fragments/propagules reach freshwater ecosystems that are naturally unshaded, such as floodplain lakes or wetlands connected to streams and rivers that can act as dispersal corridors (Säumel and Kowarik 2010), they can become a threat to native macrophyte communities there.
We emphasize that our records are a matter for concern, considering that the Amazon biome shelters so many important freshwater ecosystems and biodiverse communities, many of which are under constant land use pressure, and therefore already highly vulnerable to biodiversity loss due to  human impact. Early detection is the best strategy to prevent damages by invasive species, as it implies higher probability of eradication and prevention of spread. We recommend eradication of U. arrecta from aquatic ecosystems in Pará due to the potential ecological damage this grass may represent to these highly diverse ecosystems.