Gwilym Lewis

Dioecy has evolved many times independently within the angiosperms. The distribution, frequency of occurrence, and floral morphology of dioecious angiosperms constitute the foundations for comparative studies of dioecy, yet for many groups they are still poorly characterized. We assessed species of Miconia for the presence of dioecious reproductive system, characterized the floral morphology for staminate and pistillate flowers, and used herbarium records to analyze patterns of distribution and elevational range. We find that dioecious Miconia represent an uncommon case of mismatched stage of organ abortion between staminate and pistillate flowers, with functionally pistillate flower morphology largely consistent across species, and morphological expression in functionally staminate flowers varying from near absent to slight reductions in gynoecia. We identify 58 dioecious species and 15 putatively dioecious species within Miconia that are distributed primarily in montane habitats between 1000 m – 3500 m in the Andes, parts of Central America, and the Caribbean. Our results double the last known count of dioecy in Miconia and highlight the gradient of vestigial morphology in staminate flowers. Lastly, we provide discussion on the significance of dioecy in relation to floral development, pollination, and ecology in Miconia.

Introduction Tropical mountain rain forests (TMRF, natural forests at > 300 m asl) are globally important for biodiversity and ecosystem services and are believed to be highly vulnerable to climate change. But there are no specific approaches for rigorous assessment of their vulnerability at the landscape and local scales necessary for management for adaptation. We address the challenge of evaluating the ecological sensitivity to temperature of TMRF, applying a multidimensional approach in protected areas over a 440–2,950 m asl altitudinal gradient in Costa Rica, synthesizing results of a long-term research programme (2012-present). We evaluate the sensitivity to the current spatial temperature gradient of eleven ecosystem properties in three categories: forest composition and diversity, thermal characteristics of forest stands and forest structure and dynamics.MethodsData are from 29 to 32 plots of 50 m x 50 m (0.25 ha) distributed over the gradient, in which all trees, palms and tree ferns ≥ 10 dbh are identified to species and measured for recruitment, growth and mortality. An experimental study of leaf litter decomposition rates was carried out in twelve plots. Current and future (SSP 585, 2070) values of mean annual temperatures MAT were obtained from online climate surfaces. Thermal characteristics of forest stands were determined using MATs of species occurrences in GBIF and include a new index, the Community Thermal Capital Index (CTCI), calculated as CTI-MAT.ResultsWe classified degrees of sensitivity to temperature as very weak, weak, moderate or substantial. All eleven ecosystem properties are substantially sensitive, so changes in their values are expected under rising temperatures. Species density, the community temperature index CTI, tree recruitment and mortality rates and leaf litter decomposition rates are positively related to temperature, while the community weighted mean thermal niche breadth, the CTCI, net basal area increments, stand basal area and carbon in aboveground biomass are negatively related. Results point to zones of vulnerability in the protected areas.DiscussionIn montane forests, positive values of the CTCI–climate credit– robust basal area growth and very low mortality and leaf litter decomposition rates suggest healthy ecosystems and no risk of mountaintop extinction. Lowland forests may be vulnerable to degradation and biotic attrition, showing current basal area loss, high mortality and climate debts. National and local actors are participating in a process of adoption of the sensitivity analysis and recommendations regarding zones of vulnerability.

Geoclimatic events driving South American aridization have generated biota differentiation due to barriers and new environment formation. New environments allow species climatic niche evolution, or the geographical expansion of an existing one. Understanding the role these processes play may clarify the evolution of South American biota. Gomphrena L. ranges across almost all the continent's arid environments. We tested whether South American drylands are biogeographically connected through the Gran Chaco but, due to different aridity levels, lineage diversification could have also been associated with the evolution of climatic niches and morphological or physiological traits. With available data, we generated a dated phylogeny, estimated ancestral ranges, performed diversification analyses, reconstructed ancestral states of two characters, and examined if niches have changed between lineages. Results showed that Gomphrena diversified throughout the easternmost South American drylands ~15.4 Ma, and subsequently three independent clades colonized the western arid regions during the last Andean pulse, and after the marine transgressions (~4.8–0.4 Ma) via the Gran Chaco. The colonization implied an increase in the diversification rate of annuals over perennials and the progressive east–west differentiation of the occupied climatic niche. This diversification was influenced by C4 photosynthesis, which could have acted as a niche opener to conquer new environments after the Paranaean Sea withdrew. Spatiotemporal patterns found in Gomphrena suggest that geographical expansion and evolution of climatic niches played a common but decoupled role in promoting diversification. These results show that the Gran Chaco may have acted as a historical connection linking South American drylands.

Predicting the likelihood that non-native species will be introduced into new areas remains one of conservation’s greatest challenges and, consequently, it is necessary to adopt adequate management measures to mitigate the effects of future biological invasions. At present, not much information is available on the areas in which non-native aquatic plant species could establish themselves in the Iberian Peninsula. Species distribution models were used to predict the potential invasion risk of (1) non-native aquatic plant species already established in the peninsula (32 species) and (2) those with the potential to invade the peninsula (40 species). The results revealed that the Iberian Peninsula contains a number of areas capable of hosting non-native aquatic plant species. Areas under anthropogenic pressure are at the greatest risk of invasion, and the variable most related to invasion risk is temperature. The results of this work were used to create the Invasion Risk Atlas for Alien Aquatic Plants in the Iberian Peninsula, a novel online resource that provides information about the potential distribution of non-native aquatic plant species. The atlas and this article are intended to serve as reference tools for the development of public policies, management regimes, and control strategies aimed at the prevention, mitigation, and eradication of non-native aquatic plant species.

Brazilian moist forests and savannas are some of the most species-rich biomes in the Neotropics. In the transition zones between these regions, ecotones often accumulate even higher taxonomic diversity. However, whether these ecotonal communities consist of overlapping species widespread from the neighbouring biomes or a specific set of locally adapted species still needs to be clarified. Regional differences in species richness may be influenced by factors such as species' environmental tolerances, life forms, or species’ range sizes. To investigate the species richness found in ecotones, we used the ‘milk-weed’ family (Apocynaceae), which comprises both widespread and narrowly distributed trees, lianas, and shrubs, as a model to evaluate if (i) their observed richness in ecotones is promoted by widespread species or by locally adapted species; (ii) trees, lianas, and shrubs show different richness patterns in savannas, ecotones, and forests; and (iii) species found in ecotones have broader environmental tolerances than other species in the family. We used a taxonomically curated georeferenced dataset to compare the range sizes of 643 species of Apocynaceae from 73 genera listed for Brazil, comprising 298 species with a liana life form and 345 trees, herbs, or shrubs. We recorded 335 predominantly forest species, 56 savanna species, and 152 ecotone species, for which we quantified species richness, areas of occurrence, precipitation, and temperature ranges and tested for differences in range sizes and environmental tolerances between habits and ecoregions. Our results indicate that (i) Apocynaceae species occurring in ecotones have wider geographical ranges than species not occurring in ecotones; (ii) lianas showed higher area-weighted richness in ecotones than other life forms; and (iii) species found in ecotones had broader environmental tolerances than species restricted to moist forests or savannas. These results indicate that the species richness found in ecotones between savannas and moist forests in Brazil is not necessarily a consequence of higher endemism and local adaptation but may also be a result of overlapping ranges of widespread species typically associated with neighbouring biomes. Together, our findings add to our understanding of ecotones and biomes as continuous, gradual biogeographical transitions instead of sharply defined ecological units.

Premise: Cacti are characteristic elements of the Neotropical flora and of major interest for biogeographic, evolutionary, and ecological studies. Here we test global biogeographic boundaries for Neotropical Cactaceae using specimen‐based occurrences coupled with data from visual observations, as a means to tackle the known collection biases in the family.MethodsSpecies richness and record density were assessed for preserved specimens and human observations and a bioregional scheme tailored to Cactaceae was produced using the interactive web application Infomap Bioregions based on data from 261,272 point records cleaned through automated and manual steps.Key ResultsWe find that areas in Mexico and southwestern USA, Eastern Brazil and along the Andean region have the greatest density of records and the highest species richness. Human observations complement information from preserved specimens substantially, especially along the Andes. We propose 24 cacti bioregions, among which the most species‐rich are: northern Mexico/southwestern USA, central Mexico, southern central Mexico, Central America, Mexican Pacific coast, central and southern Andes, northwestern Mexico/extreme southwestern USA, southwestern Bolivia, northeastern Brazil, Mexico/Baja California.ConclusionsThe bioregionalization proposed shows biogeographical boundaries specific to cacti, and can thereby aid further evolutionary, biogeographic, and ecological studies by providing a validated framework for further analyses. This classification builds upon, and is distinctive from, other expert‐derived regionalization schemes for other taxa. Our results showcase how observation data, including citizen‐science records, can complement traditional specimen‐based data for biogeographic research, particularly for taxa with specific specimen collection and preservation challenges and those that are threatened or internationally protected.This article is protected by copyright. All rights reserved.

The popularization of citizen-science platforms has increased the amount of data available in a fine spatial and temporal resolution, which can be used to fill distribution knowledge gaps through probabilistic maps. In this study, we gathered expert-based information and used species distribution models to produce two independent maps of the greater rhea ( Rhea americana , Rheiformes, Rheidae) distribution in the state of Rio Grande do Sul, Brazil. We integrated municipality level detection/non-detection data from five citizen-science datasets into a Bayesian site occupancy model, accounting for false negatives, sampling effort, habitat covariates, and spatial autocorrelation. We addressed whether habitat (grassland and crop field cover, number of rural properties) and spatial autocorrelation explains the realized occurrence of the species and compared model-based and expert-based occurrence maps. The mean estimated percentage of occupied municipalities was 48% (239 out of 497 municipalities), whereas experts declared 21% of the municipalities (103) as occupied by the species. While both mapping approaches showed greater rhea presence in most municipalities of the Pampa biome, they disagreed in the majority of the municipalities in the Atlantic Forest, where more fieldwork must be undertaken. The greater rhea distribution was exclusively explained by the spatial autocorrelation component, suggesting that the species expanded its distribution towards the north of the state, reaching the Atlantic Forest, following deforestation and agriculture expansion.

Endemic to the Neotropics, Monnina is the second largest genus of Polygalaceae, yet little is known about its phylogenetic history, biogeography, and morphological character evolution. To address these knowledge gaps, we conducted Bayesian and maximum likelihood (ML) analyses of nuclear ITS and plastid trnL–F regions to test the monophyly of Monnina s.l. We used this phylogenetic framework to (i) infer divergence time estimates of lineages within the genus and reconstruct their historical biogeography; (ii) reconstruct the evolution of morphological characters of putative ecological and evolutionary importance in Monnina; and (iii) test for correlations between our phylogenetic hypothesis and environmental data. Our results reveal that Monnina is monophyletic with an indehiscent, 1–2-seeded fruit as a synapomorphy for the genus. We identify six clades within Monnina based on our combined phylogenetic results: Clades A, B, and D are primarily distributed in southern and eastern South America, Clades C and E are primarily Central Andean, and Clade F is chiefly distributed in the Northern Andes and Central America. The ancestor of the Monnina stem lineage dispersed from Australia/Africa to South America during the late Eocene to early Oligocene. The divergences of major lineages within the genus began in the early Miocene. We inferred the most recent common ancestor of Monnina to be an herbaceous plant with one-seeded samaroid fruits. The origins of fleshy fruits and shrubby habits are phylogenetically correlated within Monnina, and their concerted convergent evolution may have promoted increased net diversification rates in the two most species-rich subclades of the genus.

Aim The aim of this study was to further advance our understanding of the species-rich, and ecologically important angiosperm family Combretaceae to provide new insights into their evolutionary history. We assessed phylogenetic relationships in the family using target capture data and produced a dated phylogenetic tree to assess fruit dispersal modes and patterns of distribution. Location Tropical and subtropical regions. Time Period Cretaceous to present. Major Taxa Studied Family Combretaceae is a member of the rosid clade and comprises 10 genera and more than 500 species, predominantly assigned to genera Combretum and Terminalia, and occurring on all continents and in a wide range of ecosystems. Methods We use a target capture approach and the Angiosperms353 universal probes to reconstruct a robust dated phylogenetic tree for the family. This phylogenetic framework, combined with seed dispersal traits, biome data and biogeographic ranges, allows the reconstruction of the biogeographical history of the group. Results Ancestral range reconstructions suggest a Gondwanan origin (Africa/South America), with several intercontinental dispersals within the family and few transitions between biomes. Relative abundance of fruit dispersal types differed by both continent and biome. However, intercontinental colonizations were only significantly enhanced by water dispersal (drift fruit), and there was no evidence that seed dispersal modes influenced biome shifts. Main Conclusions Our analysis reveals a paradox as drift fruit greatly enhanced dispersal distances at intercontinental scale but did not affect the strong biome conservatism observed.

Understanding the interactions between abiotic (environmental and anthropogenic) factors and species diversity and distribution patterns is fundamental to improving the ecological representativeness of biodiversity management tools such as protected areas (PAs). However, significant knowledge gaps remain about how species’ ecological and evolutionary opportunities are associated with abiotic factors, especially in biodiversity-rich but economically ill-equipped countries such as Kenya. Here, we explored the interactions of terrestrial mammal diversity patterns and abiotic factors across species groups and ecoregions in Kenya. We coupled data on terrestrial mammal occurrences, phylogeny, functional traits, and environmental predictors in Kenya to derive multiple diversity indices, encompassing species richness and phylogenetic and functional richness, and mean pairwise and nearest taxon distances. We explored the interactions of these indices with several abiotic factors using multivariate regression analyses while adjusting for spatial autocorrelation. The results showed weak correlations between species richness versus the phylogenetic and functional diversity indices. The best-fit models explained variable proportions of diversity indices between species groups and ecoregions and consistently retained annual temperature and precipitation averages and seasonality and human footprint as the strongest predictors. Compared to the species-poor xeric northern and eastern Kenya regions, the predictors had weak associations with diversity variances in the species-rich mesic western and central Kenya regions, similar to focal species groups compared to ordinal classifications and the combined species pool. These findings illustrate that climate and human footprint interplay determine multiple facets of terrestrial mammal diversity patterns in Kenya. Accordingly, curbing human activities degrading long-term climatic regimes is vital to ensuring the ecological integrity of terrestrial mammal communities and should be integrated into biodiversity management frameworks. For a holistic representation of critical conservation areas, biodiversity managements should also prioritize terrestrial mammal phylogenetic and functional attributes besides species richness.