recognition of the similarity of the island’s forests to those of the adjacent mainland received more detailed study by Loucks who in his classification and mapping (Figure 6), recognized affinities between both the island's coniferous and hardwood forests, and the equivalent forests on the mainland.

The recognition of the island's forest-types — All six researchers were aware of the presence of different forest-types on the island All of them recognised the prevalence of “hardwood forest", or "tolerant hardwood forest", on the dry or 'upland’ soils of the island, five of them listing beech, sugar maple and yellow birch as its principal species“). At the same time they also noted the presence of areas dominated by coniferous forest, especially black spruce, tamarack and white—cedar.11

Most of them accounted for this heterogeneity in island forests in terms of the effects of topography and elevation on soil drainage patterns, even if some of them only did so by making the general (and not wholly accurate) comment that the upland areas were occupied by tolerant hardwoods and the lower lands by various types of coniferous forest”. Erskine, and to a less extent Loucks, commented on the effects of soil texture on drainage patterns, especially the association of clay and sandy soils with particular forest-types.

Several of the studies drew attention to a noticeable geographical pattern in the distribution of the forest-types, observing especially that the coniferous forests tended to occur in the west and east, and along the north shore.13 However, the author showing the greatest understanding of the differentiation of forest-types on the island was

‘0 Halliday, Rowe, Clark, Erskine and Loucks. Also, several of

Stilgenbauer’s ‘forest belts‘ comprised ‘Deciduous Hardwoods‘. 1‘ See especially Erskine and Loucks for the presence of black spruce forests on the bogs and swamps (with tamarack also as an element). Black spruce forest is also implied by Halliday, Clark and Rowe, while the presence of white-cedar in the swamps is mentioned by Erskine, Stilgenbauer, Clark and Halliday.

12 Halliday, Rowe and Clark.

‘3 See especially Erskine and Loucks, but it is also implied by Rowe, Clark and Stilgenbauer. Such geographical differentiation had of course long been noted by some of the early describers of the island’s forests, and especially by the naturalist Francis Bain in his 18905 textbook (see Sobey 2006, pp. 16-17).

Erskine who also made frequent reference to differences in their geographical distribution on the island.

The mapping of the island's forest-types — Two of the authors (Stilgenbauer and Loucks) went on to portray this geographical heterogeneity in the form of maps.14 It should be noted that neither of their maps aimed to show what was actually present at the time of mapping (in fact much of the island’s original forest would then have been cleared), but rather the old—growth climax forest that is likely to have occurred had forest clearance never taken place.15 Nor did they aim to display a high level of detail or accuracy in their maps.16 Loucks, working at the level of Maritime forests, sub- divided the island into three ’forest districts’ (Figure 5): one comprised the hardwood areas of the central and eastern parts of the province, while the other two were coniferous districts, one in the west along the shores of Northumberland Strait, the other running along the whole length of the north shore. Thirty years before, Stilgenbauer had drawn a map (Figure 1) that showed both a greater variation within island forests than in Loucks' map (he divided the island into six forest 'belts’) as well as a greater fineness in his forest boundaries. However, the descriptions of his 'belts' are brief and qualitative, and the criteria he used to delimit them are not stated.

The forest before European settlement — All six of the studies were carried out at a time when only about a third of the island’s land surface retained a forest cover”, with that forest, heavily disturbed and cut—over, being quite different in tree species

1‘ Halliday and Rowe also produced maps that included Prince

Edward Island (Figures 2 and 3), but being on a broad scale of the whole of Canada, they show no internal differentiation at all in the forests of the island.

‘5 This is the “original—natural” forest in the sense of Peterken (1996), though the two maps may also be indicative of the “future— natural” forest (also sensu Peterken), namely the type of forest that would ultimately occur over the whole island if all areas were allowed to revert naturally to climax forest.

‘6 In fact, it was not until the provincial Forestry Branch began to collect comprehensive data on the island‘s forests, both from ground plots and through the use of aerial photography, that detailed maps on forest and tree distribution have been able to be produced (see, e.g., Sobey and Glen 1999, 2004).

17 At the time of the first aerial photographic survey of the island in 1935, only 32% of the island's land surface was under forest (Glen 1997), though thereafter this proportion was to increase — by 1960 it had reached 36% (Anon. 2003, p. 2).