EXOTIC PALMS THREATENS NATIVE PALMS: A RISK TO PLANT BIODIVERSITY OF ATLANTIC FOREST 1

– Invasive plants can profoundly modify physical and biological characteristics of their new environments, especially when such habitats are already fragmented and reduced by anthropogenic pressure, such as the Atlantic Forest of Brazil. Here, we hypothesized that exotic palms successfully establish among the natural Euterpe edulis populations through a continuous propagule input by avifauna, high germination rates, and rapid growth. As a result, the native palm is experiencing decline and may be threatened with extinction. Beginning in 2007, we conducted a continuous forest inventory (FCI) every three years in the primary and secondary forest fragments of Viçosa, Minas Gerais. We use a Markov matrix to project future distributions of palm trees. The secondary forest contained three exotic palm species: Archontophoenix alexandrae , Livistona chinensis , and Arenga caudata . The first palm is a serious risk to natural E. edulis populations in the Atlantic because of frequent interactions with birds, rapid germination, and aggressive colonization in the lower to medium vertical forest strata. Currently, natural E. edulis populations are viable and sustainable, capable of regeneration, growth, and fruiting, their communities maintain continuous gene flow, dominating vertical forest strata compared with exotic palms. However, exotic palms should be monitored and control measures should be analyzed, especially in areas with A. alexandrae populations.


INTRODUCTION
The Atlantic Forest is among 25 critical biodiversity points worldwide (Cincotta et al., 2000). Threatened by land use changes and decreasing land cover (Soares-Filho et al., 2014), only 12.5% of the original forest area remains (SOSMA/INPE, 2014). The Atlantic Forest comprises ecosystems that have lost 70% of their original vegetation but together still possess more of 60% of all terrestrial species on the planet (Chediack and Baqueiro, 2005). Much of the current forest is extremely fragmented by human activity and located within cities (Gastauer et al., 2015). Furthermore, approval of the new Brazilian Forest Code (2012) increased the vulnerability of these regions (Soares-Filho et al., 2014).
Along with fragmentation effects, the invasion of exotic species has led to plant biodiversity big losses in Atlantic Forest remnants (Christianini, 2006). Biodiversity loss begins with reduced environmental and genetic variability (along with any interactions), eventually culminating in local extinction (Santos et al., 2016). Through competitive exclusion, invasive terrestrial plants can profoundly modify the physical and biological characteristics of their surroundings (D'Antonio and Vitousek, 1992), altering soil properties, shading, primary productivity, and susceptibility to fire (Christianini, 2006;Mengardo and Pivello, 2012).
During the 21 st century, the Portuguese royal family introduced exotic palm trees to Rio de Janeiro for use in gardens and botanical parks (Araujo and Silva, 2010). Avifaunal seed dispersal subsequently facilitated establishment in native ecosystems (Chapple et al., 2012;Almeida et al., 2015). The negative effects of biological invasions on local biodiversity can occur through multiple levels, altering individuals in a community, genetics, population dynamics, and ecosystem processes (Parker et al., 1999;van Wilgen and Richardson, 2014). Among exotic and invasive palms in Brazil, the genus Archontophoenix exerts a particularly strong impact on natural Juçara or açaí palm (Euterpe edulis M.) populations in the Atlantic Forest (Dislich et al., 2002;Christianini, 2006).
Identifying illegal extraction is extremely difficult (Galetti and Fernandez, 1998), because exploitation is selective and undetectable via Landsat satellite images (Asner et al., 2005). During the past decade, E. edulis was declared a species at risk of extinction in Argentina and vulnerable in Brazil and Paraguay (Chediack and Baqueiro, 2005). In this study, we hypothesized that E. edulis decline and extinction risk is closely tied to the establishment of exotic palm trees through continuous avifauna-promoted propagule entry, high germination, and rapid growth.

Study area
The study area comprises two forest fragments of semidecidual seasonal mountain forest in Viçosa, Minas Gerais, Brazil. The first area (A1) is a secondary forest fragment of 75 ha, located at the University Federal of Viçosa (UFV) (20º 45'2 00''3 S; 42º 51'2 00''3 W). Deforested in 1922 for coffee planting, area A1 is currently in the process of natural regeneration. The second area (A2) is a preserved (Gastauer et al., 2015) primary forest fragment of 36 ha (20º 47'2 43''3 S; 42º 50'2 47'3 W) on a private rural property, used as control ( Figure  1). The municipality encompassing both forests fragments with 620-820 m in altitude, intersected by Doce River tributaries and exhibiting a highly mountainous topography. Red and red-yellow Oxisols dominate the region. The climate is humid subtropical wetland, with rainy summers and cold, dry winters (Ferreira Júnior et al., 2007).

Data collection
We performed continuous forest inventory (CFI) every three years starting from 2007 until 2013, to evaluate forest dynamics of exotic palm invasion into natural E. edulis populations. Each forest fragment (A1 and A2) contained five randomly distributed, permanent plots (10 × 25 m, 250 m²), totaling a sampled area of 0.25 ha. Two types of data were collected. Level I included all palm trees with diameter at breast height (DBH) > 5.0 cm, preferably measured at 1.30 m from the soil. A Vertex IV hypsometer (Haglöf Sweden) was used to estimate total height (H) of palms. Level II involved the A1 fragment only; 10 randomly placed subplots (1 m 2 ) were first designated within each plot (50 subplots

CONDÉ TM et al.
total), and then all palm seedlings with circumference at ground level (CGL) < 10 cm were measured.
During CFI, plants were identified with the support of the Forest Engineering Department at UFV. A sample from each species was then collected and sent to the Botany Herbarium at UFV. Updates to any scientific names were performed at Tropicos (http:// www.tropicos.org/), and species were classified by the Angiosperm Phylogeny Group III.

Data analysis
Forest vertical structure was analyzed across three levels (Condé and Tonini, 2013) where Y nθ = column vector of tree count by diameter class, after n time periods;G θ = Markov matrix of the first period (2007-2010); Y (n-1)θ = column vector of tree count by diameter class, earlier in n time periods; R txθ = column vector of ingrowth trees by diameter class in a given period, dependent on rate or probability of ingrowth (scenarios 1 or 2);M txθ = column vector of dead trees by diameter class in a given period, dependent on rate or likelihood of death (scenarios 1 or 2).
Column vectors of ingrowth and mortality were calculated with the following equations: Where N (n-1)θ = living tree count by diameter class, earlier in n time periods;N θ = living tree count by class diameter, in a given periods;R θ = ingrowth count by tree diameter class in a given period; M θ = dead tree count by diameter class in a given period.
Predictions of exotic and native palm dynamics were made for four time intervals: 1) 2007-2010, 2) 2010-2013, 3) 2013-2016, 4º) 2013-2021 (A1 only). Distribution frequencies of individuals per diameter class were compared using the chi-squared test. Significance was set to p < 0.01 and p < 0.05. In scenario 1, frequencies were estimated with entry and mortality rates for 2007-2010. In scenario 2, frequencies were estimated with entry and mortality rates for 2010-2013. Statistical and graphical analyses were performed with Microsoft Excel 2007 and the "ggplot2" package in R (R Core Team, 2018).

RESULTS
In A1 during 2007, we measured E. edulis, Archontophoenix alexandrae H. Wendl. & Drude, and Livistona chinensis (Jack.) R. Br. (latter two species are exotic). Density of individuals was highest for E. edulis (920 ind. ha -1 ). Species composition was unchanged in 2010, but the E. edulis population had decreased, while exotic populations remained stable. In 2013, we found an additional exotic palm, Arenga caudata (Lour.) H. E. Moore, while E. edulis population further declined and A. alexandrae population increased. In A2, exotic palm trees were not found during either CFI period (Table 1). Both DBH and H were higher in E. edulis of A2 than of A1. Additionally, the A2 E. edulis population increased by 2% from 2007 to 2010 but decreased by 6% from 2010 to 2013 (Table 1).

Eq3
Exotic palms threatens native palms... reduction in progress (Table 1). In contrast, A. alexandrae exhibited a +300% rate of change, indicating substantial increase in regeneration. Livistona chinensis regeneration was also declining. In 2007-2010, 8% of studied areas did not exhibit natural palm regeneration, but by 2010-2013, the amount had risen sharply to 22%. Figure 3 depicts morphometry, regeneration, fruiting, and establishment data from A. alexandrae and L. chinensis within E. edulis populations of A1.

CONDÉ TM et al.
We observed high population density (776 ind. ha -1 ) among A. alexandrae, with individuals concentrated in the 7-23 DBH classes. This density is associated with rapid vertical growth, leading to dominance of the LS and MS strata. These results suggest that A. alexandrae had already been successfully established in the past, and current populations continue to proliferate rapidly. By comparison, a high-density population of A. cunninghamiana (DBH = 9.5-25 cm and stem height = 12 m) was found in an Atlantic Forest reserve of São Paulo, Brazil (Dislich et al., 2002). The presence of this invasive species in the reserve is unlikely to be from past disturbances in fragmented forests; instead, the close proximity of afforested urban squares and gardens to forest remnants results in a continuous supply of exotic seeds via avifaunal interactions (Dislich et al., 2002;Mengardo and Pivello, 2012).
Dynamics of E. edulis in A1 reveals that natural regeneration is in progress, mainly due to the dominance of reproductive adults in the middle and superior vertical strata. However, E. edulis populations were strongly and negatively affected in plots that also contained A. alexandrae, which exhibited intense regeneration associated with large-diameter individuals. On average, each E. edulis adult contributes to the regeneration of 98 seedlings per year (Matos et al., 1999;Reis et al., 2000;Fantini and Guries, 2007).Therefore, E. edulis survival and growth depends strongly on seedling density and adult presence.
Several researchers had previously suggested that E. edulis regeneration would remain stable regardless of local environmental conditions (Matos et al., 1999). Our results do not support such a conclusion; reduced E. edulis regeneration in our study fragments during 2010-2013 appear to be associated with precipitation and humidity fluctuations (INMET, 2014) linked to global climate change (IPCC, 2014). Indeed, multiple studies have demonstrated that E. edulis has very specific water requirements (Reis et al., 2000;Fantini and Guries, 2007;Corrêa Júnior et al., 2008).
The continuous increase of A. alexandrae may create a genetic flow barrier among E. edulis populations through mechanisms such as resource competition, as well as differential pollination, seed dispersal, and seedling survival efficiency (Carvalho et al., 2015;Santos et al., 2016). In contrast to A. alexandrae success, we found a low number of L. chinensis seedlings, despite previous reports of high germination rate (96-99%) regardless of soil and climatic conditions (Kobori, 2006). This outcome may be because L. chinensisis poorly adapted to the shaded conditions in a closed-canopy forest.
Although we did not evaluate how forest fragmentation influences growth and gene flow dynamics of palm trees (Santos et al., 2016), we note that fragmentation probably had direct and indirect effects on plant biodiversity in both study sites. Currently, the Atlantic Forest is heavily fragmented, and evidence shows that species composition responds strongly to the degree of isolation or connectivity, fragment size and shape, as well as characteristics of surrounding matrices and edges (Bierregaard Jr. et al., 1992;Brasil, 2003). Thus, monitoring and management of invasive palms among natural E. edulis populations should be based on understanding how regeneration, growth, reproduction, and mortality are associated with phenology, seed production, and dispersal.
Various methods of protection and conservation of E. edulis have been developed (Orlande et al., 1995, Martins andLima, 1999;Reis et al., 2000;Fantini and Guries, 2007;Corrêa Júnior et al., 2008), but we have made very little progress in conserving existing populations. For conservation efforts to be effective, we need to advance public policies that focus on education, training, and environmental awareness. Specifically, people must be taught sustainable extraction techniques as well as how E. edulis production and trade relate to the species' regeneration, growth, mortality, and cutting cycle. Environmental organizations must combat illegal harvesting, foster sustainable practices, and place value on those who extract non-timber forest products in ways that preserve forest environmental services. Moreover, the government should include legislative representatives committed to nature conservation.
In agreement with Dislich et al. (2002), we recommend the partial or total removal of young and adult A. alexandrae, L. chinensis, and A. caudata within secondary fragments and their surroundings, thus hindering the propagation of these invasive species within forests. We also advise the enforcement of Law 11.428, which delineates the use and protection of native vegetation in the Atlantic Forest Biome (Brasil, 2006). Continued monitoring under this law should help conserve the genetic inheritance of E. edulis populations and maintain native plant biodiversity. Currently, the majority of afforestation and landscaping species are Exotic palms threatens native palms... not screened for possible environmental damage or the potential of biological invasion. Finally, caution and careful planning must accompany any introduction of landscapes with exotic trees.

CONCLUSION
The exotic palm A. alexandrae represents a major risk to native E. edulis populations in the Atlantic Forest because of strong avifauna propagation, high germination rate, and dominance in the lower and middle forest vertical strata.
Natural E. edulis populations are actually viable and sustainable if communities maintain continuous gene flow through regeneration, growth, and fruiting. However, control and monitoring measures should be developed for exotic palm trees, especially in areas where A. alexandrae is already present.
Finally, planning of afforestation and landscaping projects near Atlantic Forest remnants should carefully consider whether selected plants have invasive potential that could harm native plant biodiversity.