Background
Although much of New Zealand's hill and high country had a good woody species cover centuries ago, most has now succumbed to man's use of fire, axe and grazing animals. Today, much of the South Island's high country is treeless; a situation considered "natural" by many of the landowners as well as visitors to the area.

However, most of the high country remains a natural forest environment and this encourages not only good growth of some introduced species such as Lodgepole (Pinus contorta), Corsican (P. nigra) and Ponderosa pines (P. Ponderosa) and Douglas fir (Pseudotsuga menziesii), but also natural regeneration or spread.

Such is the success of woody species on unimproved land, especially now that grazing pressure (from rabbits as well as sheep) is declining and the use of fires is restricted, that land users could be forced to choose between forests and improved pastures. This may well become one of the principal planning issues of the tussock grasslands and mountainlands over the next few decades.

Extent of spread
The total area of spread is difficult to gauge, as densities vary from one tree/ha to tens of thousands/ha. In the North Island the best known incidence of spread occurs on the Central Plateau, where in 1975 some 30,000 ha was said to have an infrequent to dense covering of Lodgepole pine. At the cost of many millions of dollars the majority of unwanted trees have since been removed.

In the South Island high country the area containing more than 100 trees/ha amounts to about 6000 ha. If hill country is included the area involved increases at least two-fold. The largest South Island area of spread is on the Amuri Range near Hanmer where some 6000 ha have been colonised by Corsican pine.

Major species
Conifers are the most common introduced species to spread, growing on over 90% of the area currently dominated by regeneration. The most common species in the South Island high country can be ranked in decreasing dominance as Corsican pine, Douglas fir, European larch (Larix decidua), lodgepole pine, and radiata pine (P. radiata). Less extensive areas of Scots pine (P. sylvestris), maritime pine (P. pinaster) and Ponderosa pine are also found. Sycamore (Acer pseudoplatanus) is the most common broadleaf species involved.

Some species seed earlier and are more vigorous spreaders than others. Lodgepole pine is generally regarded as having the greatest regeneration ability amongst the conifers but the majority are liable to spread.

Biology of spread

Seed production
Once mature, most conifers will produce cones every year with heavy seed production every 3-5 years. For example, in a good year, a 12-year old lodgepole pine is capable of producing 15,000 seeds.

The onset of coning in conifers is influenced by the severity of the environment. Trees produce cones earlier and often in greater quantity in drier areas than in wetter areas. Observations suggest that local progeny will spread earlier than did their parents.

Seed dispersal and germination
Most dispersal is by wind, usually to the south-east of the source, away from the prevailing north-west winds. Autumn and winter are the main periods of seed dispersal.

The majority of spread is fringe spread and is encountered within a few hundred meters downwind of the source. However, it is not uncommon to find wildings some kilometres downwind of the parent trees. This long distance spread usually originates from "take-off" sites - ridges or hilltops or from sites on or near slopes exposed to the prevailing winds, and often results in the establishment of lone outlier trees, which if left alone will eventually parent a new pocket of wildings. For unknown reasons, the appearance of wildings is often several years after the first seed is produced.

The vast majority of seed germinates within a year of dispersal but the occasional seedling can emerge after 4-5 years. Germination is slower in drier sites.

Sequence of spread
Many environmental factors affect seed production, dispersal and seedling establishment. Therefore the sequence of spread is intermittent and areas of regeneration do not contain representatives from all seed years. Large spread events resulting from decades of natural regeneration generally occur as masses of even-aged seedlings of distinct age structure. In many cases, age classes are separated by 20- to 30- year intervals.

Management factors influencing spread
1. Choice of species. Some species such as Lodgepole, Corsican and Scots pine are more spread prone than others such as radiata and ponderosa pine.

2. Plantation design. Most wildings come from seed produced at the margin of plantations. A two-row edge of less spread-prone species will reduce spread potential.

3. Siting. The risk of uncontrollable spread is minimised if trees are never planted on take-off sites or immediately adjacent to undeveloped tussock or scrub country.

4. Surrounding land management. Generally, trees will not spread onto land where there is either a well established dense vegetation cover such as intact forest or improved pasture, or there is periodic, intensive grazing (mob stocking).

5. Grazing. Historically, grazing has been the main (often unrecognised) agent of wilding prevention/control. However, on extensively managed land which is not grazed uniformly, total wilding control is unlikely. It can only be guaranteed where oversowing, topdressing, subdivision and mob stocking are practised or where heavy stocking is applied at least annually, preferably in the autumn.

Conifers are extremely difficult to eliminate completely once they are more than one ear old and have developed a significant amount of woody tissue. Grazing can more readily control the spread of species such as radiata, lodgepole and Ponderosa pine than that of Scots pine, Douglas fir and Corsican pine (the least palatable conifer).

Palatability to sheep

Spread prevention/control
Characteristically, tree spread is very visible and predictable. Therefore the prevention or control of unwanted spread should be easier than for the likes of gorse or broom.

If management pays heed to the above-mentioned factors influencing spread, problems with future plantings need not arise.

Containment
Where large areas of unwanted spread already exist, management should consider containment areas where wildings are tolerated (as too expensive to remove) surrounded by control zones where no trees are allowed to reach the age of seed production. Wildings rarely occur before parents reach the age of 8 with the occasional exception of Lodgepole pine, so managers have some years in which to interrupt the spread sequence if so desired. In areas where wilding spread is not wanted, managers should check for wildings every 5-8 years (depending on the species involved), and remove existing or potential wilding sources, particularly outlier trees and those on take-off sites.

HIGH	Lodgepole pine (8) Scots pine (12) Douglas fir (12) Corsican pine (13) European larch (12) Ponderosa pine (13) Muricata pine (12) Radiata pine (10)
LOW

Spreading vigour and age of significant viable seed production.

Removal techniques
Most wildings are removed by hand pulling (<50 cm tall) or by cutting at ground level (leaving NO green foliage) with hand tools or machines such as chainsaws and scrub-cutters. These have proven to be the most reliable removal techniques. However, chemicals are also available for either foliage spraying or for application to the cut stump.

Guidelines for the control or management of wilding trees have been prepared and are available from the Ministry of Forestry.

Cost. This is often high, but ranges from a few dollars to many hundreds per hectare, depending on age, stocking and type of terrain. It is doubtful whether we currently have the financial resources to manage the existing areas of unwanted spread. Therefore managers should concentrate on preventive measures, particularly as there is likely to be considerable increase in high country areas planted in the near future.

Stocking/ha 0-10 10-100 100-1000 1000-5000

$/ha 1-10 10-70 70-250 250-700

Wilding removal costs

The "do nothing" option
Even though this is the cheapest management option, it is difficult to support. The result is "forestry by default" which proponents argue gives us "something for nothing". Self-sown forests can create a valuable resource but the majority do not. Too often the result is a forest of a non-desirable species, occupying a non-preferred site, with an uneven age structure and producing wood of inferior quality.

If a reversion to woody species is seen to be unavoidable, management should try to work with it rather than do nothing. By seeding or planting a desirable woody species (native or introduced) the 'natural' plant succession can be induced to work for, rather than against, management.

Management for timber production
While most wilding spread is not suited to timber production, where species is suited to the site and the density of trees is sufficiently high and uniform to induce good form, the opportunity can exist for management for timber production. Management inputs are most likely to involve either thinning or additional gap planting to achieve desired stand densities. The only wilding species which are likely to provide a production forestry opportunity are Douglas fir, European larch, Ponderosa pine and Corsican pine. For the last two species it is very important that the tree spread is of good genetic quality (provenance) in terms of form and vigour. The tending regimes described in the individual species can be used to manage wilding stands.

Conclusion
Introduced trees will forever be a part of our hill and high country scene. Although most plantings are not parenting wildings, the potential for natural tree spread is large and increasing. To avoid wildings unnecessarily threatening important high country values such as landscapes, rare ecosystems and traditional land uses, present and future plantings of introduced trees must be properly planned and managed.