Sampling Methods for Estimating Change in Forest Resources

Abstract

Changes in forest resources have been estimated in a variety of ways. This paper focuses on extensive forest surveys rather than on sentinel‐site investigations. The sampling design and plot design used are key to precise estimates of change. Alternative sampling designs include temporary surveys, Continuous Forest Inventory, and Sampling with Partial Replacement. Each can be used in conjunction with stratified sampling or double sampling for stratification. Plot designs can involve variable‐radius or Bitterlich sampling for trees, and fixed‐area plots for most attributes. In extensive surveys, it is efficient to group plots into clusters. Plots must be sampled at a frequency that is commensurate with the rate of change, degree of interest, and funding available. Often, plots are less than a hectare in size and spaced widely across the population. Continuous Forest Inventory, with or without stratification, is efficient for estimating current values, net change, and components of change. Much work remains in scaling to understand landscape‐level interactions and to identify stressors and indicators of forest health and sustainability.

Plot size for modelling the spatial structure of Sudanian woodland trees

“…Plot size is a key factor determining the accuracy of statistical estimations of parameters (Scott 1998). Our results showed a significant effect of plot size on univariate and bivariate spatial structure parameters.…”

“…Kangas (2006) concluded that the coefficient of variation increases with decreasing plot size, which is consistent with our results. Similarly, Scott (1998) found that as plot size increases, within-plot variance increases and between-plot variance decreases, resulting in a smaller variance estimate across all plots. The bivariate spatial analyses for each sample in our study suggest a random spatial distribution for Isoberlinia spp.…”

A New Method for Selection of Umbrella Species for Conservation Planning

“…The disadvantage of this method is that it requires extensive biotic surveys, which in turn may require substantial investment of resources (Longino andColwell 1997, Mac Nally 1997). Identification of umbrella species could reduce the sampling effort necessary to prioritize areas for protection or other land uses, because it usually is faster and cheaper to sample a subset of a taxonomic group than to survey all members of the group (Scott 1998). Umbrella species have limited application in landscapes where practitioners are interested in shortcuts to assess ecosystem integrity for purposes of land management.…”

REDD+ in West Africa: Politics of Design and Implementation in Ghana and Nigeria

“…A fundamental part of this in both Nigeria and Ghana is expert work at rendering carbon visible through maps, figures, charts. Often, these activities render carbon visible by rendering other things invisible, drawing on similar historic, simplifying forestry logic that produced bio-diverse landscapes as timber [98,99]. While these representations are informing mass education and mobilisation in Ghana, in Nigeria carbon representations are being uneasily territorialised through superimposition of significantly simplified images of carbon forests on actual, dynamic forest landscapes.…”