Data from: Effects of deer browsing on eastern hemlock growth and survival
Data files
Feb 12, 2024 version files 732.81 KB
Abstract
Studies during the past six decades convincingly demonstrate that persistent browsing by native white-tailed deer (Odocoileus virginianus) eliminates the ability of eastern hemlock (Tsuga canadensis), white cedar (Thuja occidentalis), yellow birch (Betula alleghaniensis), and several other long-lived tree species to regenerate in forests of the upper Midwest and northeastern United States. The data shared here were used for three publications to date (see the Related Works section, below): The first is Alverson et al. 2019, which critically examined the relative contributions of deer versus snowshoe hare in reducing the growth and survival of eastern hemlock. The second is Alverson et al. 2024, a more detailed examination of growth and mortality rates of eastern hemlock, and the relationship of those two variables. The third is Waller et al. 2024, which explores the Regeneration Ratio, a new metric that integrates growth ("transit time" through the zone of vulnerability to browsing by deer and other herbivores) and mortality (half-life), and can be used to assess regeneration potential in populations of other tree species. (We are interested in collaborations to further test this metric for additional temperate and tropical species.)
This is a rich data set from extensive fieldwork in northern Wisconsin and the western Upper Peninsula of Michigan, USA, from 1991 to 2011. It will be useful to others who wish to check our results, or examine specific subsets of data, e.g., the growth and mortality of specific size classes of hemlocks, behaviour of cohorts during a specific decade relative to regional changes in deer or hare abundance, or the characteristics of cohorts growing in specific ownerships. Although we examined the relationship of diameter and height, we did not consider allometry in detail in these publications. We encourage others to explore the allometry of these indigenous, juvenile hemlocks and would be happy to share the limited additional data that we have, upon request.
README: Effects of deer browsing on eastern hemlock growth and survival
https://doi.org/10.5061/dryad.fbg79cp2n
This is a rich data set from extensive fieldwork in northern Wisconsin and the western Upper Peninsula of Michigan, USA, from 1991-2011. It will be useful to others who wish to check our results, or examine more specific subsets of data, e.g., growth and mortality of specific size classes of hemlocks, cohorts during a specific decade relative to regional changes in deer or hare abundance, or cohorts growing in specific ownerships. In particular, although we examined allometric relationships between diameter and height, we included little about this in our publications. We encourage others to explore this, and we have a limited amount of additional data on the ages of some of the indigenous hemlock juveniles (seedlings and saplings) that we are willing to share.
Description of the data and file structure
The dataset comprises 936 rows and 117 columns. Each row corresponds to a unique hemlock individual that was tagged and followed throughout the study (or until death). The sets of columns describe (1) seedling and site identity, and whether the seedling was transplanted or protected by a deer-proof "exclosure" cage; (2) site location; (3) diameter growth; (4) height growth; (5) browse and pellet-count data for hare and white-tailed deer at the study sites; and, (6) mortality data relevant to Cox proportional hazard analysis.
Empty cells signify that we have no data for a particular hemlock individual for the specified census period, as a result of one of these three conditions:
(1) The individual died since the previous census, so it has no living tissue on which to measure height or diameter. An examination of the mortality data in the rightmost columns of the dataset clearly indicate when individuals died.;
(2) The study site where this individual grew was not visited during that census period. This mostly applies to sites on Menominee tribal lands, which were not visited after 1997 and sites in the Ottawa National Forest, which were not visited after 2007. In all cases, the mortality data in the rightmost columns of the dataset will indicate the year that that individual was "censored" (cf. Cox proportional hazards analysis), i.e., was lost to the study with an unknown fate;
(3) In a very few cases, we simply missed a measurement on that individual in the field during a given census.
Description of columns in Alverson.HemlockForDryad.18dec2023.xslx (and .csv):
SEENO - Unique number assigned to each seedling; each was stamped onto a stainless steel tag, loosely attached with a stainless wire at base of stem at soil surface.
SITE - Study site number.
OWN - Ownership of site as of 1991. CNF, Chequamegon National Forest; LDF, Lac du Flambeau Reservation (First Nation lands); MEN, Menominee Reservation (First Nation lands); NNF, Nicolet National Forest (all in Wisconsin); ONF, Ottawa National Forest (Michigan). The Chequamegon and Nicolet National Forests later merged into the Chequamegon-Nicolet National Forest.
ORIGN - Source of hemlock seedlings. All individuals here are “WI,” which means they were seedlings indigenous to Wisconsin or the Upper Peninsula of Michigan.
TRANS - Transplant status. “N” indicates an individual growing in situ; “T” indicates that it was transplanted from a nearby location in northern Wisconsin or the western Upper Peninsula of Michigan.
EXCLO - Exclosure status. “C” indicates an individual growing inside a deer-proof cage that was permeable to hare, rabbits, and other small mammals; “U” indicates that it was not protected by an exclosure cage.
DegN - Degrees of north latitude. All locality data were recorded with a handheld GPS in 2007, with the exception of the Menominee (MEN) data, which were estimated in 2023 using paper maps from our field work and GoogleEarth.
MinN - Minutes of north latitude.
DegW - Degrees of west longitude.
MinW - Minutes of west longitude.
Acc - Accuracy (meters) of locality information. These were derived from a handheld GPS except in the case of the Menominee (MEN) data, which were assigned very conservative values based on our review of paper maps of those sites in 2023.
D91S - Diameter (mm) of individual seedling at base of stem in spring 1991.
D92S, D93S, … D11S - Diameter (mm) at base of stem in spring 1992, 1993, … 2011.
D.y0 - Diameter (mm) at base of stem the first spring that the individual was measured.
D.y1, D.y2, … D.y20 - Diameter (mm) at base of stem in spring 1 year, 2 years, … 20 years after the individual was first measured.
H91S - Height of individual (cm) from ground level to highest living foliage in spring 1991.
H92S, H93S, … H11S - Height (cm) from ground level to highest living foliage in spring 1992, 1993, … 2011.
H.y0 - Height (cm) from ground level to highest living foliage the first spring that the individual was measured.
H.y1, H.y2, … H.y20 - Height (cm) from ground level to the highest living foliage in spring 1 year, 2 years, … 20 years after the individual was first measured.
B92S, B93S, … B11S - “Recent” browse. If an individual appeared to have been browsed since the prior census, it was scored as “1” and if not, as “0”. An imperfect metric, in part because the census periods were not evenly spaced and we didn’t critically assess of how long prior to a census the browse occurred and could still appear as “new.” Note that we didn’t distinguish between deer and hare browse, although typically the difference was quite obvious.
0-BrowseCensuses - The number of censuses (spring visits) during which “recent” browse on that individual was NOT detected.
1-BrowseCensuses - The number of censuses during which “recent” browse on that individual WAS detected.
#BrowseCensuses - The number of censuses that individual hemlock was observed and scored for browse.
AnyBrowse - Whether any browse was detected on that individual during the entire period of its observation; 1 = yes, 0 = no.
N91 - Number of new deer pellet groups at site during the survey in spring 1991. Specifically, these are the number of new deer-pellet groups in eight circular 0.01-acre sampling points (totaling 323.9m2/site). Careful preliminary trials were made to avoid counting pellets deposited in previous years: see Alverson and Waller 1997 for details of the trials and sampling protocol.
N92, N93, … N11 - Number of new deer pellet groups at site in spring 1992, 1993, … 2011.
SumDeer - Total number of deer pellet groups at the site across all spring pellet surveys.
#DeerCensuses - Number of spring censuses for deer pellets at the site.
MaxDeer/Census - The maximum number of deer pellet groups at the site in any census.
HPEL91 - Number of individual hare pellets at site during spring survey in 1991. Specifically, these are the number of new hare pellets in eight circular 0.01-acre sampling points (totaling 323.9m2/site) at the site. See Alverson and Waller 1997 for the detailed protocol.
HPEL92, HPEL93, …. HPEL07 - Number of individual hare pellets at site during spring surveys in 1992, 1993, … 2007.
SumHPEL - Total number of hare pellets at the site across all spring pellet surveys.
#HPELCensuses - Number of spring censuses for hare pellets at the site.
MaxHPEL/Census - The maximum number of hare pellets at the site in any census.
HAR91 - An alternative metric for hare presence at the site, specifically the number of circular, 0.01-acre sampling points in which hare pellets were detected during the spring survey in 1991.
HAR92, HAR92, …HAR07 - Number of sampling points in which hare pellets were detected during the spring surveys in 1992, 1993, … 2007.
SumHAR - Total number of sampling points with hare points at the site, across all spring pellet surveys.
#HARCensuses - Number of spring censuses for hare pellets at the site.
MaxHAR/Census - The maximum number of sampling points within site that had hare pellets in any given census.
AliveYR - The first calendar year that the living seedling was observed and measured. (This and the remaining columns are useful for Cox proportional hazards analyses.)
DIED - Whether the individual was observed dead (1) or not (0) during the study. Note that a score of “0” can mean either that the individual lived to the end of the study or that we ceased following it when it was still alive and don’t know its fate at the end of the study.
DeadYR - The first census year in which the individual was observed dead.
LastYR - The last census year in which the individuals was observed alive.
TimeAlive - The number of years the individual was alive during the study.
Censor - Indicates whether we knew the fate of the individual throughout the study (Censor = 0); or if we were unable to ascertain its fate after the last time we observed it alive, i.e., it was “lost” to the study (Censor = 1).
Sharing/Access information
The authors can be contacted to clarify, or to share additional data (including the dataset used in earlier publications, which also included non-native hemlocks that were used as "phytometers" to assess browsing pressure early in the twenty-year field experiment.)
Methods
Site selection, study plot location, assignment of individual hemlocks to plots and exclosure treatment, and protocols for growth and mortality data, and for pellet counts, are fully described in W.S. Alverson and D.M. Waller, 1997. Deer populations and the failure of hemlock regeneration in northern forests. Pp. 280-297 in W.J. McShea et al., eds. The science of overabundance: Deer ecology and population management. Smithsonian Institution Press, Washington, DC.
A description of the analyses, and a summary of protocols, are provided in W.S. Alverson, D.A. Riege, and D.M. Waller. 2024. Losing the slow race: How deer hinder growth, survival, and regeneration of juvenile hemlocks (Tsuga canadensis). Forest Ecology and Management. [https://doi.org/10.1016/j.foreco.2023.121466]