Islands are isolated land masses that frequently exhibit simplified ecological systems containing locally adapted and endemic species, often with small population sizes, low reproductive rates, and a lack of predator defenses compared with mainland counterparts (Wilson and MacArthur, 1967; Moser et al., 2018). These attributes make island ecosystems more susceptible than mainland ecosystems to human-related impacts, such as the introduction and establishment of non-native species. The theory of island biogeography (Wilson and MacArthur, 1967) states that isolation and island size are critical factors determining the number of species that can colonize and establish on island ecosystems. Depending on species’ vagility, islands near the mainland may have similar native species composition and ecological characteristics and higher species richness than remote islands. Thus, islands with a biogeographic history of lying near continental areas may be more resistant to the establishment of new species than those isolated far from mainland (Moser et al., 2018). Islands across the Caribbean show strong species affinities with both North and South America because of their shared biogeographic history, which largely results from their proximity to these two continental regions (Roncal et al., 2020). Therefore, we might expect Caribbean islands to be more diverse and therefore more resistant to the establishment of introduced species, including those that are potentially invasive (i.e., species spreading rapidly into new areas; terminology follows Blackburn et al., 2011) than more isolated islands in ocean basins.

Humans are the ultimate ecosystem engineers, altering habitats for shelter, provision, and resource acquisition. Via their activities, they often introduce new species to areas they colonize, either intentionally for provision or ornamentation or unintentionally by bringing along parasites or commensals. Thus, isolation is effectively reduced. Studies have shown that in addition to isolation and island area (typical factors of the island biogeography theory), factors such as levels of anthropogenic disturbance, human activities (e.g., economic development, human population size, trade and transportation rates and pathways), and propagule pressure are key drivers explaining the diversity of introduced species and invasion success on islands (e.g., Gallardo et al., 2015). Human-mediated introductions of non-native species to islands may lead not only to increases in the number and distribution of species that may colonize islands but also to the displacement and/or extinction of native species. In this regard, human-related activities may lead to a breakdown of the “classical” biogeographic theory and may “redefine” species diversity on islands (Capinha et al., 2015; Blackburn et al., 2016; Rojas-Sandoval et al., 2020).

Caribbean islands have a long history of humans intentionally or accidentally introducing non-native species, and a subset of these have become invasive (e.g., Kairo et al., 2003; Reynolds and Niemiller, 2010; Wilson et al., 2011; Borroto-Páez and Mancina, 2017; Rojas-Sandoval et al., 2017; Shiels et al., 2020). Islands within this region have been exposed to pronounced anthropogenic disturbance, which has included ecosystem degradation and natural resource overexploitation, particularly since the beginning of the European colonization (Maunder et al., 2008; Acevedo-Rodríguez and Strong, 2008). These disturbances often facilitated the introduction and establishment of non-native and invasive species (Lugo, 2004; Rojas-Sandoval and Acevedo-Rodríguez, 2015; Rojas-Sandoval et al., 2017, 2020). On the other hand, biodiversity on Caribbean islands is so remarkable that it is considered one of the most important hotspots of diversity in the world (Mittermeier et al., 2004). Elton’s biotic resistance hypothesis proposes that areas of high diversity should be resistant to biological invasions (Elton, 1958). While this may be true at small scales, it is difficult to find evidence for it in larger regional scales (Stohlgren et al., 2003, 2006; Ackerman et al., 2017). Ackerman et al. (2017) evaluated Elton’s theory at different scales using a dataset of plant species in Puerto Rico. They found that there was a strong, positive correlation between invasive and native species richness at large scales (municipality scale), yet the relationship between native and invasive species was weak at local forest scales (forest reserves), suggesting some biotic resistance is present at local forest reserves in Puerto Rico.

The island of Puerto Rico is the smallest of the Greater Antilles, and at the whole-island scale, these dual processes of biotic resistance vs. direct or indirect anthropogenic facilitation of species introductions are present. Occupied first by the indigenous peoples beginning 4,000 years BP, then colonized by the Spanish and eventually annexed by the United States, 95% of the island of Puerto Rico was converted to agriculture, mining, and urban areas (Birdsey and Weaver, 1987). Human activities were absent at the summits of the Luquillo Mountains, and these mountains have remained as the largest forested area on the island (Figure 1). Located in the Luquillo Mountains, El Yunque National Forest (EYNF) is currently the largest and the most well-conserved forest area in Puerto Rico. The lower flanks are a mosaic of former human land use, mostly farming and forestry, where forest has recovered naturally or has been partly restored by humans. The summits and higher elevation areas are relatively well-conserved remnants of native forest, away from roads that extend to two of the three dominant peaks.

In this study, we asked the following question: What is the current status of introduced species within El Yunque National Forest (EYNF), the largest and most well-conserved forest area of Puerto Rico? We intended to answer this question by first reviewing the literature on non-native species of vascular plants, invertebrates, and vertebrates that have been deliberately and/or accidentally introduced in EYNF, and then asking local experts to review the species list and identify and provide justification for those species that they considered to be currently behaving as invasive in EYNF. Based on previous studies showing that areas with higher anthropogenic disturbance often provide more opportunities for the introduction and establishment of non-native and invasive species (Lockwood et al., 2009; Blackburn et al., 2016; Dawson et al., 2017; Rojas-Sandoval et al., 2020), we hypothesized that areas within the EYNF with long-standing continuous native forest would have fewer non-native and invasive species than areas subject to high levels of anthropogenic disturbance such as those in the lowlands and areas near roads and other historical land-use sites.

For the purposes of this review, we used the term “introduced” to signify non-native species brought to Puerto Rico intentionally or accidentally via human activities and have already established wild populations. We reserved “invasive” for non-native species that have been shown to spread rapidly and widely and are suspected or shown to have negative ecological or economic effects in the EYNF (Blackburn et al., 2011). Data presented here are based on published literature, museum and herbarium collections, first-hand experience from local and academic experts, and personal field observations. Using all this information, we first compiled a list of all the introduced species occurring within the EYNF including vertebrates, invertebrates, and vascular plants. Then, we identified invasive and potentially invasive species based on the abundance and the risk of causing negative impact on native communities within the EYNF. We found that some groups are better studied (e.g., vascular plants, earthworms, and vertebrates) than others (e.g., arthropods) and few species have been thoroughly evaluated for their invasiveness (i.e., validated ecological or economic harm). Thus, our conclusions are preliminary in part and, where information is lacking, serve as a guide for future research.

As the theory of island biogeography predicts, the island of Puerto Rico is relatively poor in species richness compared to mainland tropical areas, with concomitantly reduced biological complexity. For example, among comparable-sized areas (16–25 ha), in a forest dynamics plot (FDP) in Puerto Rico (in wet forest), there are as many as 150 freestanding woody species, while one plot in central Panama has ∼300 species and one in Ecuador has > 1,000 species (Condit et al., 2005; Ostertag et al., 2014). Similarly, the food web at El Verde, a field station located within the EYNF, shows a complete absence of large herbivores and predators, low faunal richness compared to the tropical mainland, and an abundance of frogs and lizards (Reagan and Waide, 1996).

Nevertheless, Puerto Rico is much more diverse than other more isolated islands from other regions. For example, the 4-ha FDP in Palamanui, Hawaii has only 15 tree species (Ostertag et al., 2014). Does this roughly 10-fold difference in plant species richness between EYNF and Palamanui impart any resistance to species introductions for EYNF? A widely recognized pattern in island biogeography is the species–isolation relationship (SIR), in which a decrease in the number of native species on oceanic islands will occur with increasing island isolation, linked to lower rates of natural dispersion and colonization on the remotest islands (Wilson and MacArthur, 1967). While the negative SIR has been well-documented for native species, the response of non-native species to geographic isolation is less clear and remains an open question. In this regard, Moser et al. (2018) tested the SIR for a large dataset of native and non-native species on islands worldwide and they found that the number of introduced species increased with island isolation for all taxa studied except for birds, which is a pattern that is opposite to the widely recognized negative SIR for native species. Moser et al. (2018) argue that this pattern is due to reduced native diversity and greater ecological naiveté of native biota on more remote islands. Furthermore, their analyses removed the influence of factors such as island size, climatic and topographic heterogeneity, and socioeconomic development [using per capita gross domestic product (GDP) as a proxy]. Thus, the expectation is that, all else being equal, the close biogeographic history of Puerto Rico and nearby mainland areas should provide a relative degree of biotic resistance to invasions of new species into the island’s biotic communities.

Among the important elements set aside by Moser et al. (2018) was anthropogenic disturbance, which has factored considerably in Puerto Rico, and the Caribbean in general (Kueffer et al., 2010; Rojas-Sandoval et al., 2017, 2020). Puerto Rico currently has one of the highest GDPs in the Caribbean and a high population density (Kueffer et al., 2010; Rojas-Sandoval et al., 2017) relative to many other islands, and its original forest cover was reduced to ∼5% (Birdsey and Weaver, 1987), all indicating a high degree of human disturbance. All of this has led to a high rate of introduction and naturalization of non-native and invasive species. For example, Rojas-Sandoval and Acevedo-Rodríguez (2015) found that about 32% of the total flora in Puerto Rico is introduced, a percentage relatively high when compared to other islands in the Greater Antilles such as Cuba (12%), Dominican Republic (18%), and Jamaica (21%). For Puerto Rico and other Caribbean islands, previous studies have shown that successful establishment of non-native plant species is more likely to occur in human-modified habitats than in pristine habitats. On these islands, disturbance and human-related activities seem to be major drivers influencing and facilitating the introduction and establishment of non-native plant species (Kueffer et al., 2010; Rojas-Sandoval and Acevedo-Rodríguez, 2015; Ackerman et al., 2017; Rojas-Sandoval et al., 2017, 2020).

To understand the diversity and richness of invasive species in EYNF, it is necessary to take into account the socioeconomic history of human-modified areas where farmland and logging occurred until the 1930s (e.g., Thompson et al., 2002). Our initial hypothesis was that undisturbed natural areas within EYNF would harbor fewer non-native species and be more resistant to invasion than disturbed areas away from protected areas. The information we garnered in our literature reviews and surveys of experts lends only partial support of this hypothesis. In general, vascular plants seem to hold to the pattern quite well, and a large number of the non-native plant species were more common in disturbed areas with a history of agriculture or other human land use. We found no evidence from managers suggesting that those species are rapidly spreading into better-conserved forest areas (Thompson et al., 2007; EYNF, 2008). On the other hand, the non-native vertebrates and invertebrates currently invading EYNF were not restricted in their distributions, for the most part. For example, introduced earthworms were found in both disturbed and undisturbed forests and at elevations covering all forest types described for EYNF. Similarly, invasive mammals such as black rats, cats, and mongoose roam freely throughout EYNF, and at least, the rats do not appear to be more active or abundant in disturbed habitats such as roads, landslides, riparian areas, or treefall and hurricane gaps relative to the interior mature forest (Shiels et al., 2018, 2019). Birds and fish were exceptions, with introduced species not found even in areas with long histories of human disturbance. One outcome that is clear from our review is the lack of detailed information on the distribution and abundance of non-native species within EYNF, as well as the concomitant lack of information on ecological or economic impacts of non-native species and their classification as invasive species. One of the caveats of this work is that the information that we were able to obtain was for the most part too coarse in both space and time to fully evaluate our hypothesis.

Few species have been fully studied in EYNF for their invasiveness as we have defined it. Six plant species, S. jambos and five bamboos (four Bambusa spp. and D. strictus), appear to ecologically dominate riparian areas where they were initially planted and are slowly spreading across the area (O’Connor et al., 2000; Brown et al., 2006). Additional plant species such as S. wildenowii, Ipomoea vines, and the grasses C. purpureus, M. maximus, and P. fasciculatum are behaving as invaders and forming dense patches that outcompete and smother native vegetation, and thus, they have been on the radar of managers (EYNF, 2008). Introduced earthworms are altering soil processes (González et al., 2007) and could potentially influence other ecological processes. Africanized honeybees and A. aegypti mosquitos threaten human visitors to EYNF. Introduced black rats are a nuisance and may cause significant harm to native vegetation (Shiels et al., 2019) and birds, while mongooses, rats, and feral cats may prey on critically endangered wildlife (Engeman et al., 2006). In sum, we find some worrisome trends in the effects of invasive species but no strong evidence that the EYNF ecosystem as a whole is challenged by these invaders such that a novel ecosystem will take over.

Our combined results showed that many non-native and invasive species in the EYNF are yet confined to the lowlands and to areas with high levels of anthropogenic disturbance. However, accurate data on their distribution, abundance, and impacts are very limited. Therefore, systematic surveys and detailed studies monitoring non-native and invasive species are needed to draw conclusions on the invasive potential and social, economic, and/or ecological impacts caused by non-native invasive species within this protected area. Considering that EYNF is one of the few (and the most important) remnants of original native forest in Puerto Rico, the potential of significant impacts by invasive species on its unique native biodiversity is high. The control and management of the current and potentially invasive species within EYNF should remain a high priority.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

JZ initiated the study, provided guidance, wrote much of an initial draft, and finalized the text. JR-S analyzed plant distribution data and contributed to synthesis and writing. AS synthesized information on animals and contributed to the writing. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.