The digitization of vast herbarium collections has made millions of plant specimen images freely available online, which can now be used to generate phenotypic datasets of unprecedented scope. Here, we assess the potential of computer vision tools to automate the extraction of predicted leaf mass per area (LMApred) from digitized herbarium specimens. We use an automated pipeline to extract leaf area and petiole width from 22 680 leaves, representing a phylogenetic informed sample of 1580 species of woody angiosperms. LMApred is estimated using a proxy equation that models the scaling relationship between petiole width and leaf mass. We assess potential sources of error in LMApred estimates and evaluate whether documented LMA-climate patterns are recovered using this dataset and phylogenetic comparative methods. Our LMApred dataset responds mainly to temperature and solar radiation and presents a positive correlation with latitude. The proxy equation, not the automated pipeline, is responsible for most of the error in LMApred estimates. Our pipeline underscores the power of combining herbarium digitization with new techniques for automated trait scoring. The increased size of datasets generated using this tool allows investigation of potential LMA-climate relationships with a geographically balanced sample while also utilizing comprehensive phylogenetic information.

Climate and environment strongly influence the size, shape, and toothiness (physiognomy) of plants' leaves. These relationships, particularly in woody non-monocotyledonous angiosperms, have been used to develop leaf-based proxies for paleoclimate and paleoecology that have been applied to reconstruct ancient terrestrial ecosystems for the last ~120 million years of Earth's history. Additionally, given that these relationships have been documented in living plants, they are important for understanding aspects of plant evolution and how plants respond to climatic and environmental changes. To conduct these types of analyses on modern and fossil plants, leaf physiognomy must be measured accurately using a reproducible methodology. This protocol describes a computer-based method for measuring and analyzing a variety of leaf physiognomic variables in modern and fossil leaves. This method allows for the measurement of leaf physiognomic traits, in particular variables related to leaf serrations, leaf area, leaf dissection, and linearity that are used in the digital leaf physiognomy proxy for reconstructing paleoclimate, as well as petiole width and leaf area, which are used for reconstructing leaf mass per area, a paleoecological proxy. Because this digital leaf trait measurement method can be applied to fossil and living plants, it is not limited to applications related to reconstructing paleoclimate and paleoecology. It can also be used to explore leaf traits that may be informative for understanding the function of leaf morphology, leaf development, phylogenetic relationships of leaf traits, and plant evolution.

temperature (MAT), mean annual precipitation (MAP), and leaf dry mass per area (Ma), along with other useful quantitative leaf traits.Methods in this package described in Lowe et al. (in review).

Paleoclimate reconstructions can provide a window into the environmental conditions in Earth history when atmospheric carbon dioxide concentrations were higher than today. In the eastern USA, paleoclimate reconstructions are sparse, because terrestrial sedimentary deposits are rare. Despite this, the eastern USA has the largest population and population density in North America, and understanding the effects of current and future climate change is of vital importance. Here, we provide terrestrial paleoclimate reconstructions of the eastern USA from Miocene fossil floras. Additionally, we compare proxy paleoclimate reconstructions from the warmest period in the Miocene, the Miocene Climatic Optimum (MCO), to those of an MCO Earth System Model. Reconstructed Miocene temperatures and precipitation north of 35°N are higher than modern. In contrast, south of 35°N, temperatures and precipitation are similar to today, suggesting a poleward amplification effect in eastern North America. Reconstructed Miocene rainfall seasonality was predominantly higher than modern, regardless of latitude, indicating greater variability in intra-annual moisture transport. Reconstructed climates are almost uniformly in the temperate seasonal forest biome, but heterogeneity of specific forest types is evident. Reconstructed Miocene terrestrial temperatures from the eastern USA are lower than modeled temperatures and coeval Atlantic sea surface temperatures. However, reconstructed rainfall is consistent with modeled rainfall. Our results show that during the Miocene, climate was most different from modern in the northeastern states, and may suggest a drastic reduction in the meridional temperature gradient along the North American east coast compared to today.

The Early Miocene of Rusinga Island (Lake Victoria, Kenya) is best known for its vertebrate fossil assemblage—particularly of early hominoids and catarrhines—but the multiple stratigraphic intervals with well-preserved fossil leaves have received much less attention. The Hiwegi Formation has three fossil leaf-rich intervals: Kiahera Hill, R5, and R3. Here, we made new fossil collections from Kiahera Hill and R3 and compared these floras to previous work from R5 as well as modern African floras. The Kiahera Hill flora was most similar to a modern tropical rainforest or tropical seasonal forest and was a warm and wet, closed forest. This was followed by a relatively dry and open environment at R5, and R3, which was most similar to a modern tropical seasonal forest, was a warm and wet spatially heterogenous forest. Floral composition of these floras differed dramatically but Kiahera Hill and R3 were more similar to each other than either flora was to R5. The Kiahera Hill flora had few monocots or herbaceous taxa and was dominated by large leaves and had a higher species richness and greater evenness than the R3 flora. Our work, coupled with previous studies, suggests that R3 had a landscape of both closed forest and more open areas with seasonal ponding. The absence of morphotypes from the R5 flora that were present in the Kiahera Hill and R3 floras provides evidence for local expatriation during the R5 time interval. These results demonstrate that there was a considerable change in both climate and vegetation over an ~500 kyr interval of the Kiahera Hill, R5, and R3 floras. Thus, this work suggests that the Hiwegi Formation on Rusinga Island samples multiple environments in the Early Miocene and provides important context for the evolution and habitat preference of early apes.

PREMISE:The size and shape (physiognomy) of woody, dicotyledonous angiosperm leaves are correlated with climate. These relationships are the basis for multiple paleoclimate proxies. Here we test whether Vitis exhibits phenotypic plasticity and whether physiognomy varies along the vine.METHODS:We used Digital Leaf Physiognomy (DiLP) to measure leaf characters of four Vitis species from the USDA Germplasm Repository (Geneva, New York) from the 2012-2013 and 2014-2015 leaf-growing seasons, which had different environmental conditions.RESULTS:Leaf shape changed allometrically through developmental stages; early stages were more linear than later stages. There were significant differences in physiognomy in the same developmental stage between the growing seasons, and species had significant differences in mean physiognomy between growing seasons. Phenotypic plasticity was defined as changes between growing seasons after controlling for developmental stage or after averaging all developmental stages. Vitis amurensis and V. riparia had the greatest phenotypic plasticity. North American species exhibited significant differences in tooth area:blade area. Intermediate developmental stages were most likely to exhibit phenotypic plasticity, and only V. amurensis exhibited phenotypic plasticity in later developmental stages.CONCLUSIONS:Leaves have variable phenotypic plasticity along the vine. Environmental signal was strongest in intermediate developmental stages. This is significant for leaf physiognomic-paleoclimate proxies because these leaves are likely the most common in leaf litter and reflect leaves primarily included in paleoclimate reconstructions. Early season and early developmental stages have the potential to be confounding factors but are unlikely to exert significant influence because of differential preservation potential.

PREMISE OF THE STUDY The size and shape (physiognomy) of woody, dicotyledonous angiosperm leaves are correlated with climate and these relationships have been used to develop. proxies. These proxies assume that leaf morphology plastically responds to meteorological conditions and that leaf traits change isometrically through development.METHODS We used Digital Leaf Physiognomy (DiLP) to measure leaf characters of multiple Vitis species from the USDA Germplasm Repository in Geneva, NY from the 2012-2013 and 2014-2015 growing seasons. These growing seasons had different temperature and precipitation.KEY RESULTS We found three primary results: (1) there were predictable significant differences in leaf characters in leaves of different developmental stages along the vine, (2) there were significant differences in leaf characters in leaves of the same developmental stage between the growing seasons, and (3) there were significant differences in leaf characters between growing seasons.CONCLUSIONS We found that Vitis leaf shape had the strongest relationship with growing season meteorological conditions in taxa growing in their native range. In addition, leaves have variable phenotypic plasticity along the vine. We interpret that the meteorological signal was strongest in those leaves that have completed allometric expansion. This is significant for leaf physiognomic-paleoclimate proxies because these leaves are most likely to be preserved in leaf litter and reflect the type of leaves included in paleoclimate reconstructions. We found that leaf development does have the potential to be a confounding factor, but it is unlikely to exert a significant influence on analysis due to differential preservation potential.

Plants are strongly influenced by their surrounding environment, which makes them reliable indicators of climate and ecology. The relationship between climate, ecology, plant traits and the geographic distribution of plants based on their climatic tolerances have been used to develop plant-based proxies for reconstructing paleoclimate and paleoecology. These proxies are some of the most accurate and precise methods for reconstructing the climate and ecology of ancient terrestrial ecosystems and have been applied from the Cretaceous to the Quaternary. Despite their utility, the relationships between plant traits and climate that underlie these methods are confounded by other factors such as leaf life-span and phylogenetic history. Work focused on better understanding these confounding factors, incorporating the influence of phylogeny and leaf economic spectrum traits into proxies, expanding modern leaf trait-climate and ecology calibration datasets to additional biogeographic areas and climate regimes, and developing automated computer algorithms for measuring leaf traits are important growing research areas that will help considerably improve plant-based paleoclimate and paleoecological proxies.

The geometries and topologies of leaves, flowers, roots, shoots, and their arrangements have fascinated plant biologists and mathematicians alike. As such, plant morphology is inherently mathematical in that it describes plant form and architecture with geometrical and topological techniques. Gaining an understanding of how to modify plant morphology, through molecular biology and breeding, aided by a mathematical perspective, is critical to improving agriculture, and the monitoring of ecosystems is vital to modeling a future with fewer natural resources. In this white paper, we begin with an overview in quantifying the form of plants and mathematical models of patterning in plants. We then explore the fundamental challenges that remain unanswered concerning plant morphology, from the barriers preventing the prediction of phenotype from genotype to modeling the movement of leaves in air streams. We end with a discussion concerning the education of plant morphology synthesizing biological and mathematical approaches and ways to facilitate research advances through outreach, cross-disciplinary training, and open science. Unleashing the potential of geometric and topological approaches in the plant sciences promises to transform our understanding of both plants and mathematics.