America’s Caribbean islands have received much less attention that Hawai`i with regard to either their biological importance or level of biological invasion. I posted a blog three years ago about invasive tree species. I rejoice that the silence might be starting to change – thanks in large part to Julissa Rojas-Sandoval.
The West Indies
The Caribbean islands – including but not limited to Puerto Rico and the U.S. Virgin Islands – are considered one of 35 global BD hotspots on Earth (Castro, Quinones, and Gould 2016; Zimmerman et al. 2021). The most comprehensive descriptions are a decade old. Castro, Quinones, and Gould (2016) report that the Caribbean islands are home to ~14,526 plant and terrestrial vertebrate species, half of which are endemic to the region. At that time more than 900 species were on the IUCN Red List. The West Indies are characterized by exceptionally high levels of plant endemism, with more than 70% of seed plants considered endemic to the region (Rojas-Sandoval et al. 2017, citing earlier floristic assessments).
The Caribbean islands have a long history of anthropogenic disturbance – especially since Europeans arrived in 1492 — and intentional or accidental introductions of non-native species. Available sources differ on the number of species that should be categorized as “invasive”.
Potter et al. (2022) report that 17% of the flora of the islands of the Caribbean archipelago are not native. Rojas-Sandoval et al. (2017) identified 516 invasive non-indigenous plant species, representing about 4% of the total West Indian flora, and concluded that the incidence of invasive plants in the region is high compared with other island groups. Interestingly, each island group has its own suite of invasive species. More than 60% of these taxa occur on a single island. Thirty-eight percent belong to one of three families — Fabaceae, Poaceae, and Asteraceae. Thirty-four families are represented by a single species; 78 families by fewer than five.
They report that the most widely distributed species is river tamarind, Leucaena leucocephala (Fabaceae); it occurs on all nine islands. Another seven species are found on seven of the islands: Casuarina equisetifolia (Casuarinaceae); Eichhornia crasssipes (Pontederiaceae); Megathyrsus maximus (Poaceae); Melaleuca quinquenervia (Myrtacea); Ricinus communis (Euphorbiaceae); Spathodea campanulata (Bignoniaceae), Terminalia catappa (Combretaceae).
The density of invasive plants per square kilometer varies considerably among the islands. The Virgin Islands (US and British) rank second highest at 0.245 / km2. Puerto Rico doesn’t stand out. As noted above, each island has a distinct invasive flora. The pair of islands which share the highest number of species (122) is Puerto Rico and the Virgin Islands (Rojas-Sandoval et al. 2017). I suggest it might be informative to explore whether these islands’ century-long ties to the United States influenced trade pathways and planting choices, thereby leading to this similarity.
Rojas-Sandoval et al. (2017) report that 31% of the invasive plants (162 species) are herbs; 20% (102 species) are trees; 16% (80 species) are shrubs; 15% (77 species) are grasses; 13% (67 species) are vines; 3% (15 species) are succulents; 2.5% (13 species) are aquatics. Three-quarters are perennials. These plants occur primarily in human-altered habitats: 98% (506 species) in ruderal areas; 61% (314 species) in seminatural areas; 35% (138 species) in agricultural areas. Only 5% (27 species) are found in mature forests.
Nineteen of the species are listed among the “world’s worst alien invasive species”. For example, six of the eight most widely distributed species (above) are invasive in Florida also.
Seventy-five percent of the invasive plant species have escaped from cultivation. According to Rojas-Sandoval et al. (2017), half of these were introduced as ornamentals; another third were introduced for various agriculture and forestry purposes. Finally, 7% were deliberately planted in natural ecosystems for some purpose such as erosion control. They suggest that repeated introduction of some of these species – especially the ornamentals – might increase propagule pressure and thus their probability of establishing.
Generalities About Biology on Puerto Rico
Puerto Rico is the smallest island of the Greater Antilles at ~ 9100 km2 (5,325 mi² / 1,379,000 ha) (Rojas-Sandoval, Presley, and Willig 2026). It is 39% forest, 32% grassland, 13% woodland and shrubland, 11% urban, 3% herbaceous wetlands, 1% forested wetlands, 1% inland water, < 1% natural barrens (Castro, Quinones, and Gould 2016).
Puerto Rico is relatively poor in species richness compared to mainland tropical areas. In one example, a plot in a Puerto Rican wet forest had half the number of freestanding woody species as a comparably-sized area in central Panama. Puerto Rico lacks large herbivores and predators. Still, Puerto Rico is much more diverse than other more isolated islands, e.g., the Hawaiian Islands (Zimmerman, Rojas-Sandoval, and Shiels 2021). There are 2,780 species of plants and 361 native vertebrates, including 277 birds, 52 reptiles, 19 amphibians, and 13 mammals (Castro, Quinones, and Gould 2016).
This biodiversity reflects the island’s highly varied topography – elevations range from sea level to 1338 m – and rainfall. Areas in the south and southwest receive ~800 mm while the high-elevation areas in the northeast are drenched by more than 4000 mm.
Unlike in most tropical areas, forests on Puerto Rico are regrowing due to abandonment of agriculture. Forest cover has increased from less than 6% in the 1950s to more than 55% by 2014. These regenerating forests contain a mixture of native and non-native species in a mosaic of forest at various stages of succession (Rojas-Sandoval, Presley, and Willig 2026). Potter et al. (2022) reported that two-thirds of Puerto Rico’s forests comprised novel tree assemblages.
Invasive species
A decade ago, Puerto Rico had 176 invasive plant species – 32% of the island’s flora. They comprised 44 herbs, 35 grasses, 33 trees, 28 vines, 19 shrubs, and 5 succulents. In this category, Puerto Rico ranks higher than other islands in the Greater Antilles e.g. Jamaica (21%), Dominican Republic (18%), and Cuba (12%) (Rojas-Sandoval et al. 2017; Rojas-Sandoval and Acevedo-Rodríguez 2015).
Three studies have addressed invasive species on “la Isla del Encanto”. They differ in in geographic scope, data, and study objectives.
Zimmerman, Rojas-Sandoval, and Shiels (2021) addressed biological invaders in all taxonomic groups, not just vascular plants. However, their analysis was limited geographically to the Caribbean National Forest (CNF; popularly known as El Yunque). This study was based on published literature, museum and herbarium collections, local and academic experts’ experience, and personal field observations. They concluded that there were some worrisome trends but no strong evidence that the National Forest’s ecosystem was in danger of fundamental change as a result of invasive species. However, they admit data were limited and called for systematic surveys and detailed studies. This need is particularly acute because of the ecological importance of El Yunque, which is one of the few (and the most important) remnants of original native forest on Puerto Rico.
Potter et al. (2022) relied on data collected as part of the USFS Forest Inventory and Analysis (FIA) survey. The survey evaluates 341 permanent plots every 5 years. They assessed trees (no other kinds of plants) in forested areas across the island, not just in the CNF. They usually attribute invasive trees’ prevalence largely to the land-use history, i.e., reforestation of formerly agricultural lands. These scientists emphasize the importance of recent land-use history in determining where non-native tree species occur. They do not offer an overall assessment of the trajectory of invasions.
A more recent analysis by Rojas-Sandoval, Presley, and Willig (2026) was also island-wide and included all woody plants with dbh ≥ 2.5 cm, i.e., trees and shrubs. Noting that invasion intensity is steadily increasing in Puerto Rico’s forests, they concluded that the threat to Puerto Rico’s native biological diversity is greater than did the earlier studies.
Relying on the FIA surveys has the advantage of providing statistically valid data. However, this survey has two drawbacks as regards assessing plant invasions. First, the FIA inventory data don’t evaluate species in the understory, i.e., herbs, small shrubs, or vines. These species might also alter long-term successional trajectories (Rojas-Sandoval, Presley, and Willig (2026). Second, the FIA inventory is not designed to detect even woody plant species in the early stage of invasion, i.e., those that occur only as small seedlings on a few inventory plots (Potter et al. 2022).
Invasive Plants
Zimmerman et al. (2021) report that 168 non-native plant species from 135 genera and 57 families are established in the National Forest – at varying levels from “naturalized” to fully invasive. Of these, 77 species (~46%) had previously been listed as invasive on Puerto Rico. Twelve were hangers-on from crops planted in the past (e.g., coffee, mango, and breadfruit). They considered most of these non-invasive because they hadn’t increased strongly during a 10-year study. They named one possible exception, Simarouba amara, a timber tree, which had increased in numbers in areas of less intense land use after hurricanes in late 1990s. However, this species has not been mentioned as problematic by any other study.
El Yunque (CNF) is home to 37 of the non-indigenous plant species, including 14 vines, nine herbs, and eight grasses. These plants are described as colonizing primarily disturbed areas (e.g., landslides), wastelands, river edges, and roadsides.
The non-native species present in the regrowing forests vary by forest type – as would be expected given the large differences in rainfall. In the subtropical dry forest the principal invasive plants are river tamarind (Leucaena leucocephala) and Prosopis juliflora (Rojas-Sandoval, Presley, and Willig (2026). The former is one of the most widespread and common of the highly invasive non-native tree species throughout the Caribbean. On Puerto Rico, L. leucocephala is documented on 12.6% of 294 forested plots, 40% of plots in this biome. P. pallida (algarroba) is recorded on 10.9% of all survey plots, 27.1% of plots in the biome. Invasions in the dry forest are sometimes arrested by chronic disturbance – presumably fire (Potter et al. 2022). However, the authors do not report which species – native or introduced – come in after the disturbance.
The subtropical moist forest is presumably the biome most affected by previous agricultural efforts. African tuliptree (Spathodea campanulata) occupies 41.8% of survey plots in this biome. Rose apple (Syzygium jambos) is found on 36.4% of these survey plots (Potter et al. 2022).
Invasive trees in the subtropical wet and rain forests biome included Erythrina poeppigiana and two species also found in the moist biome – Spathodea campanulate, and Syzygium jambos. S. campanulata is found on 33.9% of the plots in this biome (Potter et al. 2022). Zimmerman et al. (2021) call our attention to an herb, Selaginella willdenowii. They call this ornamental spikemoss one of most aggressive plant invaders in the CNF.
There is a difference of opinion on the threats posed by Spathodea campanulata. Potter et al. report that S. campanulata is found on 6.1% of all plots surveyed – half or less of plots invaded by the dryland species Leucaena leucocephala and Prosopis juliflora (and the S. campanulata plots are spread across two much larger biomes – moist and wet forests). S. campanulata is also shade intolerant so Potter et al. believe it might decline in the future as other species overtop it. The conflicting opinion is offered by Rojas-Sandoval, Presley, and Willig (2026). They call Spathodea campanulata a key driver of native biodiversity loss and homogenization in moist and wet forests. They note that the species has broad ecological tolerance and can proliferate after hurricanes. They found the species to already be significantly influencing native forest distinctiveness and composition – and to be increasing in abundance.
Potter et al. (2022) considered rose apple (Syzygium jambos) to pose a higher threat since it is shade tolerant and can form dense, monotypic stands under closed canopies – which S. campanulata cannot. Zimmerman, Rojas-Sandoval, and Shiels (2021) agree that where it occurs in areas of secondary forest near stream beds, S. jambos outcompetes other species. Rojas-Sandoval, Presley, and Willig (2026) note the species might potentially contribute to homogenizing biodiversity. A new development might counter this threat: the accidental introduction of guava/ohia rust (Austropuccinia psidii); it has been highly fatal to rose apple populations in Hawai`i. Both Zimmerman et al. and Rojas et al. mention this possibility but I have not seen a more recent discussion of whether this is occurring.
Zimmerman et al. (2021) say several invasive vines are established in El Yunque. The vines respond quickly to the disturbances caused by the region’s frequent hurricanes. Seven of the 14 species in the CNF are morning glories in the Ipomoea genus. Other locally abundant non-native vines are Epipremnum pinnatum, Pueraria phaseoloides, Dioscorea alata, Thunbergia alata, and Thunbergia fragrans.m
Potter et al. (2022) defined invasive tree species as those exhibiting 60 – 75% of stems in the “small” diameter categories (less than 12.5 cm dbh). Fifty-seven non-native tree species in old-growth and regenerating forests it this definition. They considered 17 of them as highly invasive, 16 as potentially highly invasive, and two as moderately invasive. That is, 35 of 57 nonnative tree species, or 60%, are actually or potentially bioinvaders. While on the continent only seven non-native tree species occurred on at least 2% of FIA plots across the ecoregions in which they were inventoried, on Puerto Rico 21 species (38%) occurred on at least 2% of the FIA plots. Potter et al. (2022) could not assess the invasiveness of eight additional species that occurred only as small stems on a couple of survey plots.
Potter et al. (2022) note the possibility of conflicting views about the invaders: some of the moderately to highly invasive species provide timber and non-timber forest products. These include S. campanulata, L. leucocephala, Syzgium jambos, and Mangifera indica (mango).
Rojas-Sandoval, Presley, and Willig (2026) also analyzed FIA data to assess how non-native tree species affect the biodiversity and composition in three forest types: dry, moist, and wet forests. Unlike Potter et al. (2022), they found that non-native tree species are reshaping native tree assemblages in Puerto Rican forests, leading to significant declines in native species abundance, richness, evenness, and diversity.
The transformation is greatest in the dry forest; there, native richness has declined by more than 90% by some measures. Native communities have become depauperate, dominated by a few species. In subtropical moist forest, native richness peaked at low to moderate invasion levels, then declined. S. campanulata is the dominant non-native species; as noted above, it is found on 41.8% of survey plots in this biome. In subtropical wet forests, native species richness and diversity were highest in uninvaded and lightly invaded plots, followed by marked declines at higher levels of invasion. This is largely driven by that one species, S. campanulata (found on a third of the survey plots in this biome).
The dry forest suffered the greatest proportional loss of biodiversity, including the largest proportional declines in native species richness. Wet forests also experienced strong declines in native richness & abundance. However, since wet forests had much higher initial diversity & structural complexity, their proportional decline in biodiversity is less. Thus, invasion effects differ among forest types depending on initial diversity, resource availability, land-use history, successional stage, and community composition.
Invasion can make individual plots diverge locally through loss of some rare species or gain of different non-native species. If this happens across the study site, the compositional heterogeneity at broader scales is simultaneously reduced. This results in landscape-level biotic homogenization.
Rojas-Sandoval, Presley, and Willig (2026) say homogenization effects are particularly concerning in insular ecosystems, where high endemism and ecological specialization make native biotas especially vulnerable to displacement.
To their surprise, the key driver on Puerto Rico is the increasing abundance of the non-native trees, rather than non-native species richness (Rojas-Sandoval, Presley, and Willig. 2026).
RECOMMENDATIONS
Potter et al. (2022) advise limited efforts to eradicate or control the eight species that occurred only as small stems on a couple of survey plots. These are probably incipient invasives. They name specifically Schinus terebinthifolius (Brazilian pepper) – a species already recognized as moderately invasive in Hawai`i and seriously invasive in peninsular Florida. In 2019 USDA APHIS approved release of two biocontrol insects in Florida targetting Brazilian pepper.
I recommend that, in order to avoid having to consider importing those biocontrol agents to Puerto Rico, island conservation stakeholders should focus eradication efforts on this species now while it is still – apparently – not widespread.
Rojas-Sandoval, Presley, and Willig (2026) are bolder. They call for implementation of more effective policies to prevent additional introductions and spread of the invasives already on Puerto Rico or nearby islands. They suggest that officials focus on the species named in their dataset. In addition, scientists should also collect data on understory dynamics to get a more comprehensive assessment of bioinvasion in the forest. Rojas-Sandoval et al. (2026) also urge implementation of policies aimed at reducing anthropogenic disturbances in these fragile ecosystems. One component of this program is for those responsible for forest management, conservation, and climate adaptation to incorporate invasion dynamics into their strategies. They remind us that Caribbean islands host extraordinary species richness – many of which are endemic. So protecting these biological riches is of global importance.
Preventing anthropogenic disturbance is likely to aid control of invasive plants. Zimmerman et al. (2021) found that the better-protected forests of the Caribbean National Forest was less invaded by vascular plants than were similar areas that had higher levels of human disturbance. However, this was not the case regarding invasive mammals and some invertebrates. Other than the passing reference to Austropuccinia psidii on Syzygium jambos, none of the studies considered pests or pathogens hosted by trees or other plants.
Other Taxa
Zimmerman et al. (2021) also evaluated invasive animal species. They report that six species of vertebrates are invasive in Caribbean National Forest. Black rats (Rattus rattus), mongoose (Herpestes auropunctatus), and feral cats (Felis catus) are widespread through both disturbed and undisturbed mature forest. Each threatens many native species. The black rat also preys on seeds of several native tree species. House mouse (Mus musculus) is also very common, but restricted to roadside habitats. Cane toad (Rhinella marina) and green iguana (Iguana iguana) are also present. The green iguanas was introduced relatively recently and is rarely observed in CNF. Cane toads have been established much longer. Zimmerman et al. (2021) report no impacts.
Surprisingly, they report no introduced bird or fish species in El Yunque, even the species are common on the island and in parts of the Forest having a history of anthropogenic disturbance.
Two invertebrate bioinvaders are reported as ubiquitous in the National Forest: a hybrid honeybee and Aedes aegypti mosquito. The Africanized honeybee has evolved gentle behaviors and is no longer a great threat to cavity nesting birds, including the endangered Puerto Rico Parrot (Amazona vittata). Two species of introduced earthworms, Pontoscolex corethrurus and Ocnerodrilus occidentalis, are widespread in El Yunque; a third, Drawida barwelli, is found only in lower-elevation forest areas outside NF. The earthworms alter biogeochemical cycling in the soil, which might further influence plant community dynamics. Non-native slugs and snails do not appear to be spreading rapidly or damaging native plant species. Zimmerman, Rojas-Sandoval, and Shiels (2021) warn that additional invertebrates might be of concern in the future, since they can expand rapidly after forest destructions caused by hurricanes.
See also the brief summary of invasive species on Puerto Rico included in the Regional Summary Appendix for the Southeast in Poland et al. (2019). I note mentions of the South American Harrisia cactus mealybug (Hypogeococcus pungens, which is killing columnar cacti in the islands’ dry forests.
SOURCES
Lugo, A.E., J.E. Smith, K.M. Potter, H. Marcano Vega, C.M. Kurtz. 2022. The Contribution of Non-native Tree Species to the Structure and Composition of Forests in the Conterminous United States in Comparison with Tropical Islands in the Pacific and Caribbean. USFS International Institute of Tropical Forestry General Technical Report IITF-54.
Poland, T.M., Patel-Weynand, T., Finch, D., Miniat, C. F., and Lopez, V. (Eds) (2019), Invasive Species in Forests and Grasslands of the United States: A Comprehensive Science Synthesis for the United States Forest Sector. Especially the Appendix on the Southeast and Caribbean. Springer Verlag. Available gratis at https://link.springer.com/book/10.1007/978-3-030-45367-1
Potter K.M., Riitters, K.H. and Guo. Q. 2022. Non-native tree regeneration indicates regional and national risks from current invasions. Frontiers in Forests and Global Change Front. For. Glob. Change 5:966407. doi: 10.3389/ffgc.2022.966407
Rojas-Sandoval, J., R.L. Tremblay, P. Acevedo-Rodriquez, H.D. Soltero. 2017. IAS plant species in the West Indies: geog, ecological and floristic insights. Ecology and Evolution 2017; 1-12
Rojas-Sandoval, J., Presley, S.J. and Willig, M.R. 2026. Increasing non‑native tree abundance reshapes biodiversity and composition of tropical forests. Biol Invasions 28, 84 (2026). https://doi.org/10.1007/s10530-026-03802-w
Zimmerman, J.K., J. Rojas-Sandoval, and A.B. Shiels. 2021. IAS in PR: The View From El Yunque. Front. Ecol. Evol. 9:640121. doi: 10.3389/fevo.2021.640121
Posted by Faith Campbell
We welcome comments that supplement or correct factual information, suggest new approaches, or promote thoughtful consideration. We post comments that disagree with us — but not those we judge to be not civil or inflammatory.
For a detailed discussion of the policies and practices that have allowed these pests to enter and spread – and that do not promote effective restoration strategies – review the Fading Forests report at http://treeimprovement.utk.edu/FadingForests.htm
Or