Harvard Forest Research

Land use/land cover change influenced by ownership.

Principal Investigator: David Kittredge
UMASS - University of Massachusetts: Jan 01 2007 - Jan 01 2009:

Abstract:

Most of the forest in the PIE-HFR-HBR region is owned by private families and individuals. Land cover change is fairly obvious to detect with remotely sensed data identifying the loss of forest or its conversion to developed use. Land use of forest can be far more subtle and difficult to detect, yet can be an indicator of woods pre-disposed to subsequent, eventual land cover change. Forest land use may shift subtly from “managed” in some way (in the agricultural sense), to “un-managed” but still wooded (i.e., no longer “working woodland”, so to speak), to developed or parcelized into such small ownerships as to render it basically trees in the overstory and lawns and driveways in the understory.

In landscapes dominated by private, non-industrial ownership, the transition or tension zone in the urban-suburban-rural continuum can conceivably be described by parcel size thresholds, whereby once wooded ownerships become too small to effectively support land management (Kittredge et al. 1996), they essentially become wooded backyards, buffers from adjacent development, and habitat to generalist fauna that tolerates proximity to human disturbance. Kittredge et al. (in prep.) show, for example that average private ownership declines significantly along the east-west population gradient in Massachusetts (Figure 1). Similarly, McDonald et al. (2006) determined that commercial timber harvest in Massachusetts is most sensitive to road density and real estate values, and after certain thresholds are reached, it ceases to occur as a land use activity (Figure 2). This phenomenon is not unique to Massachusetts. Wear et al. (1999) estimated that commercial timber harvest in a Virginia landscape ceased when population density reached 150 /mile2.

A conceptual model to explain private woodland owner behavior (Figure 3) posits that various physical and landscape context variables influence the decisions made by owners about the future of the land, in terms of harvest and sale or parcelization. These variables are characteristics of an ownership itself (both biophysically in terms of size and timber condition and socially in terms of owner traits - e.g., socioeconomic conditions) as well as the landscape context - physically and socially.

We propose to sample at two spatial scales across the urban: rural gradients in MA and NH to identify social and biophysical thresholds beyond which private ownerships “tip” into a non-managed condition (i.e., change their forestland use), and likewise to identify similar thresholds that define when forest land subsequently changes land cover to a developed condition. Using mail surveys and interviews, we will contact individual woodland owners and assess social, ecological, and physical factors about their land and decision-making. This is critical, as the individual ownership is the key decision-making juncture in our conceptual model. We will also sample social and biophysical factors at the larger scale of the township or county to assess contextual landscape factors that may influence landowner decision-making (e.g., real estate values, population density, land use). Figures 4 and 5, for example, show the relationship between landscape conditions such as population density and forest land use, and the proportion of a township in parcels less than 20 acres. Parcelized landscapes are more common with higher population densities and have less overall forest cover. We will use various forms of multivariate analysis to develop descriptive segments of the landowner population based on their attitudes and behaviors. We will employ spatial regression and trend analysis to investigate how the landscape-level variables interact with landowner segments and estimated behaviors. We hypothesize that this is a staggered or lagged set of conditions and seek to identify the critical variables that define these dynamic transitions from rural, wooded, and managed forest to suburban, wooded, and unmanaged, to urban and non-wooded (or urban forest).
: