You are here

Harvard Forest Data Archive


Litterfall in Hemlock Removal Experiment at Harvard Forest since 2005

Related Publications



  • Lead: Audrey Barker Plotkin
  • Investigators: Aaron Ellison, David Foster, David Orwig
  • Contact: Information Manager
  • Start date: 2005
  • End date: 2020
  • Status: ongoing
  • Location: Simes Tract (Harvard Forest)
  • Latitude: +42.47 to +42.48 degrees
  • Longitude: -72.22 to -72.21 degrees
  • Elevation: 200 to 240 meter
  • Datum: WGS84
  • Taxa: Tsuga canadensis (eastern hemlock)
  • Release date: 2023
  • Language: English
  • EML file: knb-lter-hfr.161.20
  • DOI: digital object identifier
  • EDI: data package
  • DataONE: data package
  • Related links:
  • Study type: long-term measurement
  • Research topic: physiological ecology, population dynamics and species interactions; large experiments and permanent plot studies
  • LTER core area: primary production, disturbance patterns, human-environment interactions
  • Keywords: hemlock, hemlock woolly adelgid, litterfall, timber harvest, vegetation dynamics
  • Abstract:

    Hemlock decline in New England is caused by direct and indirect effects of invasion of the hemlock woolly adelgid. Direct damage from the insect is causing gradual mortality of hemlock, and widespread harvesting of hemlock in advance of mortality, in contrast, causes immediate mortality and removal of biomass from the site. Although both processes affect thousands of acres of forest annually we have only a limited understanding of their effects on forest ecosystem function and productivity and the nature of the subsequent forest community. We anticipate that harvesting will yield different consequences than gradual mortality from the insect. Therefore we designed an experiment to simulate these contrasting impacts, by logging or girdling hemlock stands. Results from the experimental treatments will be compared to the changes observed in forests that are being infested by the adelgid, and can also be included in integrated analyses of a suite of large experiments that form a core component of the Harvard Forest LTER program.

  • Methods:

    Study Design

    The study uses large plots and a Before After Control Impact (BACI) design. Plots were established and most response parameters were monitored for 1-2 growing seasons prior to the manipulations. There are two replicates of four treatments: control, hardwood control, commercial logging and girdling. Eventually the control will become infested by hemlock woolly adelgid; the effects of HWA will overlay all treatments. Intensive study plots are 30 x 30 m within 90 x 90 m treatment units - that is, the center (or "core area") is surrounded by approximately one tree height of treatment area (overstory tree heights range from ~25-35m) . Intensive measurements focus on the central plots, but the buffers and outlying areas provide additional area for other studies and manipulations. Treatments include:

    1. Control plots are hemlock dominated and received no manipulation other than sampling.

    2. Hardwood Control plots simulate a likely future forest condition after hemlock woolly adelgid kills all hemlock in a stand and deciduous trees grow into the site. They received no manipulation other than sampling.

    3. Commercial Logging plots simulate an intensive commercial logging operation. Hemlock and other commercially valuable trees were removed, including larger hardwoods and pine for saw logs as well as smaller stems that a logger might take in order to improve future stand quality, facilitate skidding and general operation, or initiate a new cohort of sprouts. Two-thirds to three-fourths of the stand basal area was cut in these two plots, using hand-felling by chainsaw and skidding with a rubber-tired skidder. The intent was to generate an extreme response that is within the range of those seen on real commercial cuts. This treatment took place in Winter 2004-05.

    4. Girdling plots simulate some of the characteristics of hemlock woolly adelgid. All sizes of hemlock were girdled, resulting in gradual mortality and standing dead hemlock. No other species were girdled, and there is no further site disturbance. Important characteristics of hemlock woolly adelgid infestation missing include the very lengthy period of decline during which the plant is undergoing physiological stress and metabolic imbalance that may induce biogeochemical and microbial changes on the site. This treatment took place in late May 2005.

    Litterfall Collection Design

    Within each of these plots, five litterfall baskets were placed at random coordinates throughout the 90m x 90m plot. Baskets were placed in the field at the beginning of September 2005 (after treatments were implemented). Each basket is considered a sample unit. The baskets are Sterilite plastic bins with vent holes, measuring 40.6cm long by 33cm wide by 25.4cm deep. The collection area is 0.111m2. Several holes were drilled in the bottom of each basket for drainage and to hold anchor screws. Each basket was lined with fine (no-see-um) mesh that was clipped to the edges and suspended slightly above the bottom of the basket to keep the litter drier. We cut two sets of mesh liners so that litter could be quickly collected in the field by bringing in the entire mesh liner if necessary (e.g. snow or ice present). Data collected from 2 baskets in Plot 7 is not available from April 2010-September 2014. The baskets were moved to new locations within Plot 7 in August 2014, and collections from December 2014 can be used.

    Sample Frequency

    Quarterly (late March, mid-June, mid-October, early December). This sample frequency should be sufficient to determine litterfall biomass and give a sense of litterfall phenology, but would not be adequate for chemical analysis or detailed phenology.

    Sample Processing

    If very wet, samples are air-dried prior to sorting. Samples are stored in dry conditions. Leaf litter is sorted to major species groups, whereas twigs/bark/reproductive parts are pooled into one category.

    Hemlock includes: Tsuga canadensis. White pine includes: Pinus strobus. Oaks includes: Quercus rubra (mostly), Quercus alba, Quercus velutina. Birches includes: Betula lenta (mostly), Betula alleghaniensis, Betula papyrifera. Maples includes: Acer rubrum (mostly), Acer saccharum (plot 8). Other deciduous includes: Fraxinus americana, Fagus grandifolia, Prunus serotina, other species. Miscellaneous detritus includes: Leaf parts that are too small to identify, dust, insects removed. Twigs, bark, cones, etc. includes: All non-leaf plant parts, all species together. Total includes: Sum of all above categories.

    In June 2017, a spongy moth outbreak prompted us to separate out spongy moth larval droppings (aka ‘frass’) as its own category.

    Since hemlock is the species of most interest in this study, special efforts are made to efficiently sort the tiny hemlock needles. After the large pieces are sorted out, the material is sorted by a combination of hand-sorting and sieving through 2mm screen and 1mm screen.

    After sorting, samples are oven-dried at 70 degrees Celsius for 48 hours prior to weighing. Samples are weighed within 15 minutes of removal from the oven. Samples are weighed to the 0.01 g.

  • Organization: Harvard Forest. 324 North Main Street, Petersham, MA 01366, USA. Phone (978) 724-3302. Fax (978) 724-3595.

  • Project: The Harvard Forest Long-Term Ecological Research (LTER) program examines ecological dynamics in the New England region resulting from natural disturbances, environmental change, and human impacts. (ROR).

  • Funding: National Science Foundation LTER grants: DEB-8811764, DEB-9411975, DEB-0080592, DEB-0620443, DEB-1237491, DEB-1832210.

  • Use: This dataset is released to the public under Creative Commons CC0 1.0 (No Rights Reserved). Please keep the dataset creators informed of any plans to use the dataset. Consultation with the original investigators is strongly encouraged. Publications and data products that make use of the dataset should include proper acknowledgement.

  • License: Creative Commons Zero v1.0 Universal (CC0-1.0)

  • Citation: Barker Plotkin A. 2023. Litterfall in Hemlock Removal Experiment at Harvard Forest since 2005. Harvard Forest Data Archive: HF161 (v.20). Environmental Data Initiative:

Detailed Metadata

hf161-01: litterfall

  1. plot: plot number. Range 1-8.
  2. basket: basket number. Range 1-5 for each plot. Plot 7 excludes baskets 1 and 5 from April 2010 – September 2014. Analyses should account for having only 3 baskets for these collections.
  3. date: date of collection
  4. seq: chronological order of each collection
  5. hemlock: Tsuga canadensis, mass of needles (unit: gram / missing value: NA)
  6. whpine: Pinus strobus, mass of needles (unit: gram / missing value: NA)
  7. oaks: Quercus rubra, Quercus alba and Quercus velutina, mass of leaves (unit: gram / missing value: NA)
  8. birches: Betula lenta, Betula alleghaniensis and Betula papyrifera, mass of leaves (unit: gram / missing value: NA)
  9. maples: Acer rubrum, Acer saccharum and Acer pensylvanicum, mass of leaves (unit: gram / missing value: NA)
  10. othdecid: other deciduous species, mass of leaves (unit: gram / missing value: NA)
  11. misc: unidentifiable fragments, mass (unit: gram / missing value: NA)
  12. twbacone: twigs, bark, cones, seeds – woody & reproductive structures, mass (unit: gram / missing value: NA)
  13. frass: caterpillar droppings, most commonly from spongy moth (unit: gram / missing value: NA)
  14. total: sum of all categories, i.e., total mass collected per basket per date, mass (unit: gram / missing value: NA)
  15. notes: notes