Find research datasets worth reusing

Age of Nonstructural Carbohydrates in Four Tree Species at Harvard Forest 2015

We estimated the mean age of sugars within and between different organs of four temperate tree species using the radiocarbon (carbon-14) bomb spike approach. Radial patterns of carbon-14 in the stemwood and coarse roots showed that sugars tended to became older when moving towards the pith.

Gene Expression and Tree Growth in the CTFS-ForestGEO Plot at Harvard Forest 2017-2019

Major goals in ecosystem ecology have been to scale from leaves to canopies and to determine whether individual-level, intra-species and inter-specific variation is critical for models projecting ecosystem processes now and in the future. The project is important in that it examines these issues in detail considering genotypes and levels of gene expression all the way up to canopy level CO2 flux. Ecological genomics and transcriptomics are nascent fields that have been primarily restricted to model species in natural and (mostly) controlled environments. To date, we have very few studies of non-model organisms in nature and/or studies of functional genomics through space and time. The research is producing extraordinarily rich datasets regarding the gene expression of trees across populations, through space in each population, across the growing season and across years and linking this information to growth and gas exchange. It will, therefore, provide tremendous insights into how much variation exists in nature thereby guiding sampling designs in future ecological 'omics projects. More importantly, it will provide unusually detailed phenotypic information for important non-model species that have large impacts on the CO2 flux of eastern US forests.

openCC0May 2024View details →

Overstory Mapped Tree Plots at Harvard Forest since 1990

These plots were established and mapped in 1990 for an experiment designed to study the effects of selective overstory tree mortality. The planned manipulation was to kill and leave standing one species in each of four plots, to simulate mortality by a species-specific pathogen. This manipulation was never done, for logistical reasons and because the appearance of the hemlock woolly adelgid provided a more pressing "natural" experiment to study; however, the plots are maintained and have been used in other studies. There are four 50m x 50m plots, located in a mixed hardwood forest (red oak and maple species are major components), north of the experimental hurricane. Tree data from the control plot of the experimental hurricane study could be added to this set for some analyses, since all the plots are in the same general area and forest type, and similar types of measurements were made on all of these trees. Tree diameter and condition were re-surveyed in these plots in Autumns 2003, 2013 and 2023, and saplings growing into the "tree" size class of greater than or equal to 5cm diameter were measured, tagged and mapped.

Nonstructural Carbon, Phenology and Wood Formation in Three Tree Species at Harvard Forest 2017-2019

This data set comprises various observations and measurements across the 2017 to 2019 growing season for seven red maple (Acer rubrum), eight red oak (Quercus rubra), and six white pine (Pinus strobus) in the Prospect Hill Tract of Harvard Forest. The observations include spring and fall leaf phenology and basic allometry, such as diameter at breast height and height. For the leaf phenology, we followed the protocol from John O’Keefe (HF003). Measurements include wood growth data from weekly microcores and a three time characterisation of growing season nonstructural carbon concentrations (soluble sugars and starch) for stems and leaves. Additionally, stem CO2 efflux was measured once a month for the 2018 growing season and weekly for the 2019 growing season.

PhenoCam Images and Canopy Phenology at the Harvard Forest Witness Tree since 2014

The PhenoCam Network uses imagery from digital cameras to track vegetation phenology and seasonal changes in vegetation activity in diverse ecosystems across North America and around the world. Imagery is uploaded to the PhenoCam server at the University of New Hampshire, where it is made publicly available in near-real time, every 30 minutes from sunrise to sunset, 365 days a year. The data are processed using simple image analysis tools to yield a measure of canopy greenness, from which phenological metrics are extracted, characterizing the start and end of the growing season. These transition dates have been shown to align well with on-the-ground observations of tree phenology at Harvard Forest (HF003). Long-term PhenoCam data can be used to track the impact of climate variability and change on the rhythm of the seasons. This dataset contains one mid-day image for each camera. Please see the PhenoCam Network website (https://phenocam.nau.edu/webcam/) for more information and additional images.

openCC0Jan 2024View details →

Visual Counts of Tree Reproduction near NEON plots at Harvard Forest since 2020

Mast seeding, a resource pulse that has cascading effects in the environment, is a measure that can provide insight into forest dynamics. When masting data is collected in sequential years it can provide information on how tree populations are responding to climatic and environmental variables, and can also be used to relate to other indices, such as seed-eating animal species. The objective of this study is to quantify the yearly seed production of mast seeding tree species at Harvard Forest which are located near National Ecological Observatory Network (NEON) plots. This project is part of larger NEON-enabled project examining mast seeding on a continental scale at 25 NEON sites in the United States, which uses mast seeding records in conjunction with NEON collected data products like mammal box-trapping, tick drags, and bird point counts. Data collected on mast seeding can be linked to these other indices at regional and continental scales.

Juvenile Tree Responses to Soil Warming at Harvard Forest since 1995

The main goal of this study is to assess eastern tree species’ growth, survivorship, and phenological responses to soil warming in order to forecast future changes in forest succession and carbon dynamics. From 2003 through 2010, we determined that shade-tolerant, normally slower-growing species benefitted most from warming. We are continuing to study species-species responses to investigate the duration of and mechanisms behind species- and functional group responses to climate warming. Monitoring long-term demographic and physiological responses of juvenile trees with and without soil warming will allow us to model future eastern tree species successional shifts under warmer climate conditions.

Hydraulic Pathways in Leaves of Temperate Trees at Harvard Forest 2002

The transport of water, sugar and nutrients in trees is restricted to specific vascular pathways, and thus organs may be relatively isolated from one another (=sectored). Strongly sectored leaf-to-leaf pathways have been shown for the transport of sugar and signal molecules within a shoot, but not previously for water transport. The hydraulic sectoriality of leaf-to-leaf pathways was determined for current year shoots of six temperate deciduous tree species (three ring-porous: Castanea dentata, Fraxinus americana and Quercus rubra, and three diffuse-porous: Acer saccharum, Betula papyrifera and Liriodendron tulipifera). Hydraulic sectoriality was determined using dye staining and a hydraulic method. In the dye method, leaf blades were removed, and dye was forced into the most proximal petiole. For each petiole we counted the vascular traces shared with the proximal petiole. For other shoots, measurements were made of the leaf-area specific hydraulic conductivity for leaf-to-leaf pathways (kLL). In five of six species patterns of sectoriality reflected phyllotaxy; both the sharing of vascular bundles between leaves and kLL were higher for orthostichous than non-orthostichous leaf pairs. Species-differences in leaf-to-leaf sectoriality were determined as the proportional differences between non-orthostichous vs. orthostichous leaf pairs in their staining of shared vascular bundles and in their kLL; for the six species these two indices of sectoriality were strongly correlated (R2 = 0.94; P less than 0.001). Species varied 8-fold in their kLL-based sectoriality, and ring-porous species were more sectored than diffuse-porous species. Differential leaf-to-leaf sectoriality has implications for species-specific coordination of leaf gas exchange and water relations within a branch, especially during fluctuations in irradiance, water and nutrient availability.

openCC (other)Dec 2024View details →

Tree-ring measurements from permanent plot in old-growth hemlock-hardwood forest, Huron Mts., MI

This package includes tree growth-ring widths for increment cores collected from a long-term 'macroplot' established in old-growth hemlock-northern hardwoods forest at the Huron Mts. of northern MI. Tree demographic monitoring data for the entire ca. 3.0 ha macroplot are available in the EDI package edi.1416.1. In 1994 and 1995, increment cores were taken for all 'core-able' trees greater than ~ 10 cm diameter for a subsection of the macroplot about 1 ha in area, along with some additional Tsuga canadensis trees beyond that 1 ha section. Cores are NOT cross-dated. See Methods for more details. This data-package may be cross-referenced to the demographic data in edi.1416.1 using stem numbers.

openCC (other)Apr 2025View details →

Tree-ring measurements from permanent study plot in old-growth hemlock-hardwood forest, Dukes RNA, Hiawatha NF, Marquette Co., MI

This package includes tree growth-ring widths for increment cores collected from a long-term 'macroplot' established in old-growth hemlock-northern hardwoods forest at the Dukes Research Natural Area/Dukes Experimental Forest in the Hiawatha National Forest in Marquette Co., MI. Tree demographic monitoring data for the entire ca. 3.0 ha macroplot, from 1992 to 2019, are available in the EDI package edi.1526.1. In 1993, 1994 and 1995, increment cores were taken for all 'core-able' trees greater than ~ 10 cm diameter for a subsection of the macroplot about 1 ha in area. Trees that were obviously badly rotten and hollow or steeply leaning were not cored. Cores are not cross-dated. See Methods for more details. This data-package may be cross-referenced to the demographic data in edi.1526.1 using stem numbers.

openCC (other)May 2025View details →

Dataset on tree diversity metrics and aboveground carbon storage in Southeastern U.S. oak-pine forests, 2009–2019

This dataset contains measurements of tree structural and taxonomic diversity, stand attributes, and aboveground carbon storage from mixed oak-pine forests in Florida, Georgia, and Alabama, located in the southeastern United States. Data were collected from 946 mixed oak-pine, and 7224 longleaf-slash pine and oak-pine Forest Inventory and Analysis (FIA) plots respectively spanning the years 2009 to 2019. Variables include aboveground carbon, aboveground biomass, tree density, basal area, stand age, and diversity metrics such as Shannon indices for tree species and diameter-based structural classes. Functional diversity metrics—including functional dominance and functional divergence—are also included. These data were used to support a published study examining the interactive effects of diversity metrics on carbon storage using structural equation modeling. The geographic coverage represents humid subtropical forest regions of the southeastern U.S.

openCC (other)Jul 2025View details →

CO2 and CH4 fluxes from living and standing dead trees in Howland Research Forest, Maine USA, 2024

Methane (CH4) is the second-largest contributor to human-induced climate change, with significant uncertainties in its terrestrial sources and sinks. Tree stems, both living and dead, play crucial roles in forest ecosystem CH4 and carbon dioxide (CO2) flux dynamics, yet much remains unknown regarding the environmental drivers of fluxes. We measured CH4 and CO2 fluxes from 51 living trees (Picea rubens, Tsuga canadensis, Acer rubrum) along an upland-to-wetland gradient at Howland Research Forest, a net annual sink of CH4, in Maine USA. We also measured CH4 and CO2 fluxes from six standing dead red spruce stems (snags). We measured fluxes every two weeks throughout the growing season (April to November 2024) and at three heights (for a subset of red spruce stems) to capture a range of environmental conditions.

openCC (other)Aug 2025View details →

Data for: Simulated postfire tree regeneration suggests reorganization of Greater Yellowstone forests during the 21st century

Tree regeneration underpins forest resilience, but how postfire tree regeneration will change with future climate and fire regimes is difficult to anticipate. Areas of sparse and failed postfire tree regeneration have been documented in western US forests, but how future recovery pathways will unfold is uncertain. We conducted a simulation study in the Greater Yellowstone Ecosystem (GYE; United States) using a process-based model, iLand, to ask how rates, composition, and spatial patterns of postfire tree regeneration vary with 21st-century climate. Subalpine forest and fire dynamics were simulated through 2100 under four climate scenarios, 2 × 2 factorial of aridity (wet and dry) and temperature (warm and hot), in five GYE landscapes. We tallied postfire tree seedling density by species in simulated fires (> 400 ha) at five years postfire. This data set contains three data sets to reproduce analyses for changes rates of regeneration, proportion of burned cells with regeneration failure, and postfire reorganization pathways. We include the data and R scripts used for these three analyses in the publication associated with these data.

openCC (other)Feb 2024View details →

Forest tree, woody debris, root ingrowth, soil respiration and characterization data from long-term research plots for LTREB at the University of Michigan Biological Station

The NSF-funded project "LTREB: Drivers of temperate forest carbon storage from canopy closure through successional time" (2014-2024) supports research to meet the following goals: 1) elucidate mechanisms responsible for changes in C storage over decades to centuries; 2) link processes leading to persistence and resilience of forest C storage following disturbance; 3) quantify the effects of potential drivers such as forest structure, N availability, climate change, and atmospheric deposition on decadal and longer-term trajectories of C storage. Field activities for this research are conducted at the University of Michigan Biological Station (UMBS) on a pair of chronosequences and several old reference forests. Synthesis activities utilize data collected from these field sites in support of the LTREB project, as well as data synthesized from other sources (e.g., long-term UMBS plot data, AmeriFlux data, FIA data) all intended to address the core questions of the LTREB project. This dataset has been compiled and expanded over a series of versions, with new data types and observations appended periodically. Presently, the dataset includes observations from tree inventory censuses, woody debris sampling, fine root ingrowth cores, soil respiration measurements, and two sets of soil collections aimed at quantifying a range of physical, chemical, and biological properties of soil.

Tree Seed Dispersal in Hemlock Removal Experiment at Harvard Forest 2005

Throughout the northeast, the hemlock woolly adelgid (Adelges tsugae) threatens eastern hemlock (Tsuga canadensis) through direct mortality resulting from infestation followed by defoliation and indirect mortality in the form of pre-emptive logging. The efficacy of regeneration of vegetation following hemlock decline depends upon advance regeneration of seedlings and saplings, seed dispersal, and recruitment. In this study, we investigated (1) whether the basic parameters of height of release and wind velocity affected seed dispersal distance and (2) tested the fit of a basic ballistic model of seed dispersal to empirical data in areas both with and without canopies. We collected empirical data from seed dropping and seed rain experiments at Harvard Forest. Height and wind velocity only affected seed dispersal distance in open areas. Predicted values of dispersal distance generated by the basic ballistic model did not provide a good fit to observed dispersal data. Poor fits of the ballistic model to the data were due to the model’s inability to account for rare, long distance dispersal events. More complex models with additional parameters are necessary to model non-localized seed dispersal.

Sapwood Temperatures in Hemlock Trees at Harvard Forest HEM Tower 1997-1998

Sapwood temperatures in four large hemlocks, 130-200 years old and about 40 to 90 cm basal diameter and 25 to 29 m tall, were measured in order to estimate sapwood respiration in these trees as part of a physiological model for hemlock forest carbon exchange. As expected, sapwood temperature varied most in small branches and least in the main bole of the tree near the ground. An additional finding from these measurements is that sapwood begins to freeze at -1 to -2 degrees C, as shown by a stabilization and in some cases an increase of about 1 degree C in sapwood temperature when air temperature cooled continuously to below this temperature range.

Tree Growth and Coarse Woody Debris in Regenerating Forests at Harvard Forest since 2008

This project is a field-based study to measure sequestration of atmospheric carbon dioxide in a regenerating New England forest. This study established long-term biometric plots suitable for measuring changes in carbon storage through time in three forest stands: an early-20th-century conifer plantation, a naturally regenerating former conifer plantation harvested in the 1990s, and a conifer plantation scheduled for harvest next winter. The first three years of this project determined the initial carbon budget of these forest stands, measured carbon fluxes into and out of these stands, and laid the groundwork for future investigations. Subsequent years will investigate larger-scale questions, such as how successional patterns affect carbon sequestration, and how these patterns change with stand age. The work also addresses how forestry practices influence carbon sequestration, and provide guidance for how forest management could enhance terrestrial carbon uptake in the future.

Nonstructural Carbohydrates in Forest Trees at Harvard Forest 2007-2010

Forest trees accumulate and store surplus nonstructural carbohydrates (NSC) as resources to be used to support future growth. This can be viewed as a conservative investment strategy, providing reserves that the tree can utilize in times of stress. NSC reserves are thus important in the context of forest responses to climate change, and forest ecosystem carbon cycling. At quarterly intervals over a three year period, we have monitored stemwood nonstructural carbohydrate (NSC) reserves of the dominant tree species of New England at three sites along a forest composition gradient: an oak-dominated transition hardwood forest (Harvard Forest), a maple-beech-birch northern hardwood forest (Bartlett Experimental Forest), and a spruce-fir boreal transition forest (Howland Forest). Adding to recent evidence indicating that in mature forest trees the nonstructural carbohydrate pool is surprisingly large, our results suggest that this pool is also highly dynamic, and unexpectedly old. In most species, seasonal dynamics in starch (2-4x higher in the growing season, lower in the dormant season) mirrored those of sugars. This data set contains 3 y of quarterly NSC concentration measurements (stemwood starch and sugars - sucrose, glucose, fructose, raffinose, stachyose - in mg per g OD wood tissue) for red maple, eastern hemlock, and red oak at the Harvard Forest.

Inorganic Nitrogen Pools and Tree Composition in Hemlock Removal Experiment at Harvard Forest 2007-2009

This study assessed how the change in leaf litter input and environmental conditions, from that of hemlock stands to earlier successional deciduous stands, will affect inorganic nitrogen availability at the interface of the mineral soil and LFH horizons. In early October 2007 we exchanged 0.5 x 0.5 m "loaves" of LFH horizons from hemlock and deciduous stands between forest types, while keeping control plots in the original forest stands. Inorganic nitrogen outputs from the litter loaves were monitored for 20 months using ion-exchange resins, removing and replacing the resins three times at 5 - 8 month intervals. Results indicated that there were differences between the control hemlock and deciduous loaves - NO3 and NH4 availability were higher in deciduous stands. However, the transfer of litter loaves from one stand type to another did not reveal changes in nutrient availability due to a shift in habitat. Rather, the methodology of moving litter loaves to a new site caused greater changes in nutrient availability, particularly over the summer months.

Forest Inventory for Tree Demography and Carbohydrate Reserves at Harvard Forest 2009-2011

The objective of this study is to establish a network of forest plots spanning the eastern U.S. focused on understanding how regional scale climate patterns affect patterns of tree demography (e.g. growth, mortality, dieback). Within each research site we also aim to understand the landscape-scale variability in demographic rates and how these rates are affected by edaphic variables by aligning plots along primary environmental gradients (elevation, soils, hydrology, fire return interval). The demographic data obtained will also be compared to regional and landscape-scale patterns in carbon reserves in adults and sapling by sampling root and stem concentrations of nonstructural carbohydrates (TNC) and relate these to demographic patterns, life-history traits, and plant C:N ratios. Besides addressing important ecological questions directly, this study is designed to improve the representations of vegetation dynamics and carbohydrate reserves in regional and landscape-scale forest ecosystem models--two of the least data-constrained processes in such models--by parameterizing and validating the modules for these processes in the Ecosystem Demography model (ED v2.1). This data set contains two years of census data for eight mapped plots distributed across Harvard Forest along the aforementioned primary environmental gradients.