Oversights in Hindsight: Reflections from a Rambling Biogeographer

As an instructor in GIS applications at a tertiary institution I can think of no better way to use my editorial privilege than to discuss some of the obvious pitfalls awaiting enthusiasts embarking on their first research project. One of the most rewarding tasks in my research career was gathering and analysing data from Nullica State Forest, a eucalypt-dominated forest in New South Wales, Australia. The analysis within a GIS environment was important because of the issues relating to logging and the concern of environmentalists and the general public. The fieldwork was an idyllic journey in itself. If one started a bush walk on a ridge it would be common to find largely dry sclerophyll communities dominated by silver top ash reaching up to 30 or 35 metres in height gloved by thick, dark brown furrowed bark on lower trunks with smooth creamy coloured upper limbs. Descending from the ridge down slope, often via precipitous inclines, you would probably pass through a variety of stringybarks such as the distinctive yellow stringybark, taller than silver top ash with a canopy of glossy green leaves. In the valleys the understorey becomes more dense and the variety of species of eucalypts becomes richer, including peppermints and sometimes the majestic grey Maiden's gum, reaching up to 50 metres. It was possible to find sections of this forest that could be true wilderness. The depths of our tall timber country invokes respect for its monumental grandeur and variability and a desire to understand its geography. This guest editorial allows me the opportunity to reflect on this study and the use of GIS (and like technologies or tools) to help us in understanding such a complex entity as a eucalypt forest. One of my initial aims was to develop an understanding of where certain species of eucalypts were most likely to occur across the spatial expanse of the forest in terms of several sets of environmental variables. This involved gathering field data selected using a stratified (based on a GIS geology coverage) random sampling procedure. During a number of field trips, seeing the forest in the blazing heat of summer and experiencing a fall of snow in winter, 309 plots were collected and nine transects of varying length were walked. These were combined with data available from a previous study to create a set of roughly 400 sampled quadrats of eucalypt data of individual species counts. Does this constitute an impressive field data set? It may seem so to the average GIS analyst. However, given that there are at least 18 species of eucalypt growing in Nullica, this was barely adequate. These data were entered into a GIS and Transactions in GIS, 2001, 5(3): 189±192