New Zealand: Forest Restoration at Aratiatia, North Island


The construction of a hydroelectric dam along New Zealand’s Waikato River in 1964 resulted in the degradation of more than 108 hectares of native forest. The site’s proximity to a major tourist destination, Aratiatia Rapids, prompted the government to conduct a revegetation effort aimed at improving the aesthetic qualities of the area and, thus, ensuring the continued vitality of the tourism industry there. Forty-six native species of trees and shrubs were initially sown–all fast-growing and short-lived–and sites where planting failed were subsequently planted again. In 1999, 35 years after the plantation of the oldest regenerated stands, a detailed analysis of the forest’s recovery was undertaken in order to evaluate the long-term success of the project; assess how closely planted forest simulates naturally-regenerated stands of similar age; evaluate the effect of canopy composition on subsequent vegetation development; and understand dispersal modes and other life-history attributes of successfully regenerating species. Findings from this study indicate that the planting effort succeeded in creating native forest that mimics young secondary forest on relatively inhospitable new surfaces. If not for the attempt at restoration, the degraded sites would now support only weedy adventive shrubland.

Quick Facts

Project Location:
Aratiatia, Tauhara Forest, Waikato, New Zealand, -38.631229, 176.20509300000003

Geographic Region:
Australia & New Zealand

Country or Territory:
New Zealand

Temperate Forest

Temperate Forest - Mixed

Area being restored:
108 hectares

Organization Type:
Governmental Body


Project Stage:

Start Date:

End Date:

Primary Causes of Degradation


Degradation Description

Extensive hydroelectric development of the Waikato River, New Zealand’s longest and second largest, was undertaken by the former New Zealand Electricity Department in the decades following the Second World War. It involved local destruction of natural landscapes, and at Aratiatia, the last dam built on the river, losses of major aesthetic and tourist values in the vicinity of a nationally significant tourist attraction, Aratiatia Rapids.

Reference Ecosystem Description

Historical records (Kirk 1872) of pre-European vegetation pattern in the district, inference (Nicholls 1986), and extant evidence (Podocarpus totara (totara) logs buried beneath tephra) suggest that broadleaved forest with scattered emergent conifers constituted the primary forest. Conifers, mostly P. totara and Prumnopitys taxifolia (matai), were locally dominant on flatter sites with deeper tephra deposits and subject to cold air ponding. The most characteristic mesophytic broadleaved trees and shrubs were probably Knightia excelsa (rewarewa), Weinmannia racemosa (kamahi), Elaeocarpus dentatus (hinau), Nestegis cunninghamii (black maire), Melicytus ramiflorus (mahoe), Pseudopanax arboreus (five finger), Fuchsia excorticata (tree fuchsia), Schefflera digitata (pate), and Coprosma grandifolia (McKelvey 1963).

Project Goals

The original replanting project, the first large-scale attempt at ecological restoration in New Zealand, was undertaken to restore woody vegetation to unaesthetic patches of bare ground resulting from the construction of a hydroelectric dam.


The project does not have a monitoring plan.


Because of its proximity to Aratiatia Rapids, a popular tourist destination, this project is of particular interest to stakeholders in the tourism industry. However, members of surrounding communities, whose livelihoods can be expected to benefit (even if indirectly) from increased tourism, are also stakeholders in the project.

Description of Project Activities:
Rehabilitation of grossly disturbed areas was undertaken at the completion of dam construction in 1964. It involved extensive redistribution of largely unsorted regolith. Subsequent artificial contouring to create rolling slopes on gentler topography and alternating benches and faces on steeper sites mimicked the landforms present earlier. A two-stage planting scheme using only native species followed (F.S.R. Jackson and J. Greenwood 1995, personal communication), aimed at simulating natural successions in the locality. (1) Fast-growing, short-lived shrubs--Hebe stricta var. stricta (koromiko), Coprosma robusta (karamu), C. lucida, Leptospermum scoparium (manuka), and Brachyglottis repanda (rangiora)--were planted as nurse vegetation on bare substrates, with subsequent spot herbiciding of adventive grasses when required. (2) Fast-growing, short-lived small trees such as P. tenuifolium, K. ericoides, Sophora tetraptera (kowhai), and Cordyline australis (cabbage tree) along with longer-lived medium-sized (e.g. Griselinia littoralis [broad-leaf]) and tall (e.g. P. totara) trees were interplanted about 5 years later. Although at least 46 native species were planted, including at least eight not indigenous to the district, only a handful of species was widely planted. Local seed sources were used for species naturally present. Planting stock was initially bare-rooted, later potted, and later still placed in root-trainers. Where initial planting failed, sites were replanted, sometimes several times.

Ecological Outcomes Achieved

Eliminate existing threats to the ecosystem:
Construction of the hydroelectric dam was completed in 1964, and reforestation of the denuded areas began in the years immediately thereafter. Synthetic forest now covers some 108 hectares (B.R. Clarkson, unpublished data) along 5 km of the Waikato River. In 1999, a detailed analysis of the restored sites was undertaken. Sampling was confined to new surfaces devoid of vegetation at the time of the dam's completion. Few grossly disturbed areas were left unplanted over the ensuing 18 years, making it difficult to find comparable samples of unplanted ground, which might or might not have been subject to planting attempts. Fifty temporary plots, varying in size between 36 and 64 square meters, and large enough to include 15-20 canopy individuals, were randomly placed in homogenous patches of vegetation, 36 in planted areas and 14 in adjacent unplanted or unsuccessfully planted areas. Three canopy communities were identified by TWINSPAN classification of basal area of woody species. (1) Adventive Cytisus scoparius (broom) shrubland (10-23 years) up to 4 m tall, almost entirely on unplanted sites with C. scoparius the only woody species consistently present, i.e. occurring in at least half the plots (12 plots). (2) Pittosporum tenuifolium-Kunzea ericoides short forest (15-33 years) up to 9 m tall with a scattered Hebe stricta understory (21 plots), almost all on planted sites; natural Podocarpus totara and Coprosma robusta seedlings were consistently present. (3) Pittosporum tenuifolium-Sophora tetraptera short forest (14-31 years) up to 10 m tall, solely on planted sites (17 plots), with natural Melicytus ramiflorus, Aristotelia serrata (wineberry), Cordyline australis, P. totara, and C. robusta seedlings characteristically present. Four ground layer communities were identified from the TWINSPAN classification of cover scores. Two were dominated by adventive grasses: (1) Dactylis glomerata (cocksfoot)-Holcus lanatus (Yorkshire fog)-Festuca rubra (Chewing's fescue) grassland with Anthoxanthum odoratum (sweet vernal), Agrostis capillaris (browntop), and Lotus pedunculatus (birdsfoot trefoil) also consistently present occurred mostly on unplanted sites, and (2) A. odoratum-A. capillaris grassland with L. pedunculatus and D. glomerata consistently present mostly (71%) on planted sites. The remaining two communities, wholly on planted sites, were dominated by litter: (3) Mosses were widespread in one, while (4) Muehlenbeckia australis (pohuehue), Pteridium esculentum, and adventive Rubus fruticosus agg. were characteristically present in the other. Eighty-three vascular species were recorded on planted sites, two-thirds of them native; trees and shrubs (53%) and ferns and fern-allies (29%) dominated the native component. Of 46 planted native species, 23 were recorded in planted plots and others were noted outside them. In addition, 30 unplanted native species were founding plots, comprising five trees and shrubs (all with bird-dispersed fruit), four lianas (two with bird-dispersed fruit, two with wind-dispersed fruit), sixteen ferns and fern-allies, and five grasses and herbs. Forty-one species were recorded on unplanted sites, nearly three-quarters of them adventive; grasses and herbs (59%) dominated the adventive component. All species are common, widespread plants. Although the dozen or so commonly planted species at Aratiatia are now reproductively mature, ephemeral seedlings of only four, C. robusta, P. tenuifolium, P. totara, and A. serrata, occurred in at least one-quarter of planted plots. Of these, only C. robusta was consistently present, occurring in 77% of plots. Ephemeral seedlings of other commonly planted species--C. australis, Pseudopanax arboreus, S. tetraptera, H. stricta (8%), K. ericoides, and Pittosporum eugenioides (lemonwood) (3%)--were more local. Despite limited planting, ephemeral seedlings of Myrsine australis (mapou), M. ramiflorus, and Griselinia littoralis were widespread. Established seedlings were less widespread, with only one commonly planted species, S. tetraptera, occurring in at least one-quarter of planted plots. Established seedlings of other commonly planted species, P. totara, C. australis, and A. serrata, were somewhat more local, but again, seldom-planted M. ramiflorus, G. littoralis, and M. australis were unexpectedly widespread. Major differences in regeneration patterns were evident between the two planted communities. Established seedlings of seven canopy species were reasonably widespread in Pittosporum-Sophora forest, but only two (S. tetraptera and P. totara) in Pittosporum-Kunzea forest. Indigenous vascular species richness was much higher in Pittosporum-Sophora forest (51 species total) than in Pittosporum-Kunzea forest (28 species), with the cumulative species/area curves following quite different paths. Despite the proximity of seed sources in planted stands, seedlings of only five planted tree and shrub species--P. tenuifolium, P. totara, A. serrata, G.littoralis, and C. robusta--were recorded on unplanted sites, none of them commonly, and only as ephemeral seedlings.

Factors limiting recovery of the ecosystem:
The lifespan of pioneering (in this case planted) species is likely to be an important determinant of the duration of primary successional pathways (Druce 1957; Wassiliev 1982). With the oldest stands now aged 33 years, planted forest at Aratiatia may be expected to begin senescing and collapsing in some 20 years' time. Early stages of natural secondary succession on intact substrates in the locality are now dominated variously by Pteridium esculentum and adventive Rubus frutiscosus and Ulex europaeus (B.R. Clarkson, unpublished data). Later stages--short forest dominated by P. tenuifolium, K. ericoides, and Pseudopanax arboreus--may pre-date the advent of weedy adventive shrubs in the district. P. esculentum stands elsewhere can persist for 30 years or more (Wassiliev 1982), U. europaeus even longer (Druce 1957). P. tenuifolium invades moderately dense P. esculentum stands in dry regions, and P. arboreus invades P. esculentum (Wardle 1991) and K. ericoides stands (Esler 1967) in wetter ones. K. ericoides also establishes directly on bare substrates such as dune sands (Smale 1994) and mudstone (Smale et al. 1997) and in short open pasture (Grant 1967). Divergences between these natural secondary successions on intact substrates and primary successions on unplanted new surfaces at Aratiatia may result from the poorer physical or chemical properties of the latter. For example, the scarcity of P. esculentum on unplanted sites probably reflects its dislike of compacted soils, and the absence of U. europaeus its need for a moderate level of soil fertility (Wardle 1991). Planted stands now mimic young secondary forests in the district which developed before the advent of weedy adventive shrubs, having bypassed earlier successional communities dominated by P. esculentum. Amongst conifers, P. totara seedlings originating from plantings are already widespread, but only a few planted Prumnopitys taxifolia are present. Natural seedlings of some broadleaved trees and shrubs (e.g. M. ramiflorus, P. arboreus) are reasonably widespread, while others (e.g. Fuchsia excorticate, Coprosma grandifolia) are more local. Still others (e.g. Knightia excelsa, Weinmannia racemosa, and Schefflera digitata) are absent from planted stands. Although some of them were planted and others still occur naturally in the locality (B.R. Clarkson, unpublished data), seed sources may be too small and distant for effective dispersal into planted stands. Major differences in floristic richness and regeneration patterns, and thus successional pathways, are evident between the two planted forest communities, as elsewhere in New Zealand (Stewart & Woods 1997). Established seedlings of seven canopy species are reasonably widespread in species-rich Pittosporum-Sophora forest and may foreshadow the development of a conifer/mixed broadleaved forest. In species-poor Pittosporum-Kunzea forest, however, seedlings of only two canopy species--S. tetraptera and Podocarpus totara--are at all common. Podocarpus totara is among the most light-demanding of New Zealand podocarps (Ogden & Stewart 1995), commonly establishing directly in degraded pasture from seed sources nearby. Enright and Ogden (1995) suggest that "southern conifers (such as P. totara) are prime candidates for habitat restoration, despite the inevitably slow rate at which they form mature forests, because success is virtually guaranteed at minimal cost." Despite the lack of site amelioration, P. totara at Aratiatia has grown as well as in plantations on more amenable sites with intact regolith (Pardy et al. 1992). By contrast, growth rates of the most commonly planted species at Aratiatia, Pittosporum tenuifolium and K. ericoides, are much slower than those of similar-aged trees planted on more amenable sites (Pardy et al. 1992). The greater success of P. totara in terms of growth rate and regeneration suggests that fast-growing, short-lived shrubs and small trees could have been bypassed and P. totara planted directly on new surfaces to form long-lived (normal lifespan approx. 600 years: Ebbett 1992, unpublished thesis), large-stature, aesthetically pleasing conifer stands. Almost all the commonly regenerating species in synthetic forest at Aratiatia are bird-distributed, as in other documented restoration plantings in New Zealand (Reay & Norton 1999). The greater abundance of seedlings in Pittosporum-Sophora than Pittosporum-Kunzea forest may reflect the presence of mature S. tetraptera, a species which attracts nectar-feeding birds such as the native honeyeater Prosthemadera novaseelandiae (tui), a major disperser of small-fruited native trees and shrubs (Beveridge 1964). The shallow litter common under Pittosporum-Sophora forest may also provide a more hospitable substrate for seedling establishment than the grass swards typical of Pittosporum-Kunzea forest. Mean aspect and slope, factors that might be expected to affect soil moisture regimes and therefore seedling establishment, scarcely differ between the sites occupied by the two communities.

Socio-Economic & Community Outcomes Achieved

Key Lessons Learned

After three decades, gross differences are now evident between the plant communities on planted and unplanted or unsuccessfully planted sites at Aratiatia, and between the successional pathways likely to occur in them. In aesthetic terms, the large-scale planting of fast-growing, short-lived native shrubs and trees has been outstandingly successful in restoring woody vegetation to large areas of unsightly bare ground next to a major tourist attraction.

Although undertaken essentially for aesthetic reasons, this restoration also appears to have been successful, so far, in ecological terms. It has rapidly established native forest that mimics young secondary forest on relatively inhospitable new surfaces, which would otherwise now support only weedy adventive shrubland. Some planted early successional species have begun regenerating. With limited seed sources nearby, many later successional trees and shrubs of the region are still absent, and it is too early to predict the longer-term future of synthetic forest, particularly floristically-poor Pittosporum-Kunzea forest with its grassy gound layer.

Long-Term Management

Only continued monitoring will show whether further management (beyond the landscape reconstruction and planting already carried out) is necessary, or whether natural processes are operating at a level sufficient to ensure that development of the vegetation will continue along relatively natural pathways.

Sources and Amounts of Funding

Funding for restoration activities was secured through the former Department of Lands and Survey. The monitoring study that produced the results described herein was funded by the Crown, through the Foundation for Research, Science and Technology.

Other Resources

Smale, M. C., P. T. Whaley, and P. N. Smale. 2001. Ecological Restoration of Native Forest at Aratiatia, North Island, New Zealand. Restoration Ecology 9: 28-37.

Primary Contact

Organizational Contact