IN THE MATTER of the Resource Management Act 1991

AND

IN THE MATTER of the hearing of applications by Timberlands West Coast Limited to Buller and Tasman District Councils for land use consents for sustainable beech forest management (including harvesting and associated activities)

EVIDENCE OF Richard Mark Allibone

ON BEHALF OF THE APPLICANT

(Relating to freshwater fisheries )

  1. Introduction

    My name is Richard Mark Allibone. I am a fisheries scientist employed by the National Institute of Water and Atmospheric Research, in Hamilton. I hold the degrees of BSc. (majoring in zoology and geology), MSc. (majoring in zoology) and PhD., all from the University of Otago.

    I have been undertaking research on New Zealand's native fishes since 1988. Initially my research was conducted as a student and research assistant at the University of Otago. This was followed by employment with the Department of Conservation as a freshwater fish specialist in the Nelson/Marlborough and Otago Conservancies. I have been on the NIWA staff since 1996, initially in a post doctoral position and more recently as a fisheries scientist. My initial research focused on the structuring of populations of migratory and non-migratory fishes. In the course of this work and subsequent research a number of new native fish species have been discovered. These investigations are continuing (NIWA's freshwater biodiversity programme), and at present include studies of the non-migratory native fishes present in the Buller and Grey River catchments. Past and present investigations also include work on indigenous forest streams and rivers on Stewart Island, the West Coast, the Catlins District (Otago), Mt Taranaki and in the Waikato. A major component of much of this research has been to determine habitat use by native fish, native fish community structure and critical physical habitats for sustaining native fish. The Stewart Island research also includes the investigation of the native fish fauna of a river in area of clear cut forest abstraction until the 1930s. Current investigations also include research into the sustainability of the whitebait fishery, and assessment of the impacts of exotic fishes on native fishes.

    I have published ten refereed papers of which nine report on issues of relevance to this application. I have also authored 25 non-refeered reports, oral presentations and conference posters, of which the majority also pertain to issues of native fisheries management in New Zealand. I am a member of three professional societies; the New Zealand Limnological Society; the Australian Society for Fish Biology and the American Fisheries Society.

    With this background in New Zealand native fish and fisheries Timberlands West Coast Ltd have requested me to review the original fisheries report produced by Ryan Environmental in 1997; to address fisheries issues raised by the Royal Society submission on the beech production forest sustainable management plan; and to make recommendations for any alterations to the management plan for improved fisheries protection.

2. Freshwater fish fauna and Significant habitat

Ryan (1997) identifies ten native freshwater fish species as likely to be present in the streams in Timberlands West Coast sustainable beech management areas. The freshwater fish data is all from electric fishing operations and there is a lack of fisheries investigations of streams within the Timberlands estate. The reliance on electric fishing data is flaw in the Ryan (1997) report and the lack of data from Timberlands beech forest estate a significant omission in the assessment of fish present in the beech forest areas. Freshwater fisheries research in the 1990s has included debate regarding the ability of electric fishing to efficiently sample deeper water, and for some galaxiids species, kokopus in particular, to be collected using this method. There is a growing recognition that fish trapping and spotlighting can provide better estimates of the presence and abundance of all three kokopu species and better assess the biomass of large eels. Therefore I consider that the electric fishing data is likely to have underestimated the abundance of the three endangered kokopu species.

Furthermore, the lack of bully species in Ryan's (1997) assessment is puzzling and it is likely that at least three bully species are present in or adjacent to the Timberlands estate. Gobiomorphus huttoni (redfin bully) and Gobiomophus hubbsi (bluegill bully) are migratory species that are common in the Grey and Buller River catchments. These species are likely to be found in riffle areas of the major rivers and smaller streams for over 100 km upstream from the river mouths. A further species Gobiomorphus breviceps (upland bully), a non-migratory species, is also present in the Grey and Buller River catchments and likely to be widespread especially in the upper reaches and tributary rivers and streams. However, none of these bully species are considered threatened; they are common and widespread throughout much of New Zealand. Their absence from the original assessment is not a cause for any revision to the proposed management regime.

Geotria australis (Lamprey) is also not noted as likely to be present by Ryan (1997). This species has been regularly recorded in the Grey River catchment, although only once in the Buller River catchment.

Wetland fish species do not appear to have be considered by Ryan (1997) or in submissions. Neochanna apoda, the brown mudfish (a category B threatened species), is a wetland inhabitant of the West Coast. It is not known if this fish exists within the sustainable management forests. Populations of this species are known to occur in the lower reaches of both the Grey and Buller River catchments (New Zealand Freshwater Fish Database). This fish is often found in ephemeral pools and streams and also occurs in wetlands, generally where no other fish species are present. Areas found to contain populations of this fish are recommended for high priority riparian protection.

Biodiversity research investigating non-migratory galaxiids has also not been assessed by Ryan (1997). In a recent revisions of galaxiid systematics, Galaxias vulgaris has been split into six species in the lower South Island (McDowall & Wallis 1996, McDowall 1997, McDowall & Chadderton 1999). Populations of G. vulgaris in the Buller River catchment have not been assessed and species status could be considered as undetermined. Allibone and Wallis (1993) present results that show that G. vulgaris in the Rappahannock River (a Buller River tributary, near the Maruia working circle) are genetically distinct from other populations of this species. Similarly on going research (NIWA biodiversity programme) is showing that populations of Galaxias divergens in the Grey and Buller Rivers are genetically distinct from other populations. Anecdotal evidence indicates that non-migratory galaxiids in the Buller catchment (New Zealand Freshwater Fish Database) (and nationally) appear to be in decline irrespective of land use (Allibone pers. obs.). Even populations recorded from the Lewis Pass National Park appear to have declined or become extirpated in the last 30 years despite very limited human interference. Therefore, due to the genetic distinctiveness of non-migratory galaxiid populations and their decline, it is recommended that streams containing these species in forest areas be given medium or high priority status as a precautionary approach.

Ryan (1997) correctly states that four of the fish species present are classified as threatened by the Department of Conservation (Tisdall 1994). However, this classification was carried out in 1994 and Ryan (1997) notes that the classification can be somewhat subjective and species could be classified as less threatened. At present the Department of Conservation's threatened species classifications are undergoing revision and it is likely that the status of the kokopu species and koaro will be revised, possibly downwards. This is in part due to the recognition that sampling techniques have been inappropriate at times and also in part due to discoveries of new populations of Galaxias postvectis and G. argenteus. However, there is also the possibility that non-migratory Galaxias species will classified as more highly threatened.

Ryan (1997) discusses in detail water quality issues associated with the beech forest logging. Given the proposed harvest regime the review in my opinion is correct in the conclusion that there is little likelihood of any detrimental effects to fisheries.

3. current Sustainable beech management proposals

The management scheme proposed presents a very low risk to the freshwater fish species in the Grey and Buller Rivers.

It is my considered opinion that the logging scheme, with removal rates of trees at 15 trees per hectare every 15 years presents little risk, if any to threatened fish even in low priority riparian margin zone streams. I have encountered native fish species such as G. argenteus and G. fasciatus in highly modified streams with little if any woody vegetation in the riparian margin. The low rates of tree removal do not appear likely to give rise to a significant decline in the overhead shade leading to a decline in fish abundance. Many other species of native fish, the non-migratory galaxiids and bully species, frequently occupy streams and rivers with little or no shading vegetation. Removal of trees from the riparian margins of such stream is highly unlikely to have any detrimental effects on the fish populations. Physical damage to the riparian margin of streams during felling operations is also of little concern, if the logging operation fells trees away from the stream when ever possible.

Sediment inputs into streams are a major concern with any forestry operation. The management plan reduces these concerns to a minimum. The removal of logs from harvesting sites to loading areas by helicopter will prevent a significant damage to the forest floor and remove a major source of sediment release. Roading then has the highest potential to increase sediment loading of streams, especially around culverts and fords. This issue has been recognised and roading construction and maintenance plans are in place to reduce this potential sediment source. As indicated below (section 5) there are also opportunities to carry out habitat enhancements during the road construction phases for the benefit of native fishes.

4. The royal Society Submission

This section addresses issues in the in the Royal Society submission section 3.4.21

Comments have indicated that the low priority streams are cause for concern because logging can occur right up to the stream edge. I do not consider this to be issue with the management plan. The rate of tree removal at 15 trees per hectare every 15 years is sufficiently low as not to present a risk of decline in native fish population sizes. Species such as the G. fasciatus are known to be small stream dwellers, however they can also be found on occasion in larger streams. The reliance of this species on terrestrial food source has been mentioned. However, the tree removal rates seem unlikely to reduce terrestrial invertebrate food densities and result in the starvation or loss of fish from an area.

"The Ryan report accepts a criticism from DOC that surveys were not undertaken to assess the presence/absence or abundance of fish species within affected areas, by remarking "A valid point. They weren't which is why the present report was commissioned". However the Ryan report itself was based entirely on historic information available to DOC when it made that criticism (mostly from the NIWA NZ Freshwater Fish Database), so the point that was agreed to be valid by Ryan, remains valid. It remains true that no surveys of native fish have been undertaken in the Timberlands forest management area. All that is available is Ryan's incomplete and rather flawed analysis of some-one else's data."

The lack of fisheries data for the beech forest areas could appear to be problem. However, the collection of this data is required for the classification of streams for riparian management purposes. The Timberlands management plan (Annex 3 page 216) states that high priority riparian zones will be placed on streams that contain important native fish populations or are fish spawning grounds. This classification can not be carried out until the fish faunas are known and therefore I expect that the data will be collected. Once this data has been collected streams can be classified according to Ryan's (1997) recommendations. Then the streams with high fisheries values will be given the appropriate riparian protection. Therefore the present lack of fisheries data does not prevent the correct riparian margins begin determined, it just means that at present classsifications can not be carried out.

"àThis may be particularly important where there is an appreciable hyporheic zone (organisms living within the sediments beneath the stream bed which can provide a significant proportion of fish food)à"

" Neither the Ryan report nor the Timberlands documents give any mention of the role of terrestrial (forest sourced) invertebrates in the diets of many native fish, and evidence is accumulating that this is a very important role indeedà"

These statements on the diet of fish presented by the Royal Society submission are inconsistent and appear to be presenting worse case scenarios. There is concern that sediments and vegetation may accumulate in pool habitats leading to anoxic conditions and hence a decline in hyporheic invertebrates and loss of fish food. The submission then notes that terrestrial food sources are very important for native fishes. It has not been established that terrestrial food sources are very important for native fishes, except perhaps for banded kokopu in small shaded streams where instream production is limited. It is more accurate to say many of the native fishes are opportunistic in their food take and will consume any soft bodied invertebrates available. If one food source has declined in abundance then prey switching will occur and alternate food sources will be exploited. The diet statements in the Royal Society submission also fail to take into account habitat use of the different fish species. For example G. postvectis, the category A threatened species, is a resident of moderate to high gradient streams. Its diet does include more benthic invertebrates than for instance G. fasciatus, but in the high gradient streams the flushing regimes will allow rapid transport and removal of sediments and vegetation to occur. Conversely G. fasciatus that occurs in low gradient streams feeds less on benthic items and is adept at utilising streams with a high proportion of fine sediment in the substrate. It is even found in streams with naturally occurring locally anoxic conditions (Allibone per. obs.); the fish simply do not maintain long residence times in these areas.

For the Gobiomorphus species not considered by Ryan (1997) diet is dominated by a broad range of aquatic organisms and any decline in terrestrial invertebrates will have no impact on their feeding.

Finally for food to be a concern this has to be limiting the fish population. The disturbed nature of many New Zealand waterways means native fish populations rarely if ever attain the maximum population sizes possible with the available food resources. Huryn (1998) in the only systematic study of native fish production showed, for the non-migratory fish Galaxias eldoni that the population utilised approximately 30% of the available food resources and was recruit limited not food limited.

"The Timberlands report encouragingly talks about protecting the spawning habitats of fish by its management practices. While this may seem impressive and reassuring, the spawning sites of many species are as yet undiscovered, making it impossible to provide focussed protection of the appropriate habitatsà"

"àsuch small streams are crucial habitats for some native fish, especially banded kokopu, possibly also for lamprey spawning "

The submission says that spawning habitats for many species are unknown. Lamprey spawning sites are one such unknown, hence the suggestion that small streams are important lamprey spawning sites is misleading. Ryan (1997) lists ten fish species as being present or likely to be present in the Timberland beech forest areas. Of these ten, the two Anguilla species spawn at sea and the spawning habitat of G. maculatus is in tidal areas of rivers outside of the impact areas of the logging operations. Spawning habitat for G. fasciatus has been described, (Hopkins 1979, Mitchell & Penlington 1982, Mitchell 1991) and G. brevipinnis spawning habitat has been described from streams in Australia (O'Connor & Koehn 1998) and a G. brevipinnis spawning site was found last year in a Mt Taranaki stream (Allibone unpub. data). G. vulgaris has had its spawning habitat well described and repeatedly located (e.g. Cadwallader 1976, Allibone & Townsend 1997). On going investigations are targeting the spawning behaviour and habitats of G. argenteus and G. postvectis, already spawning timing is known. Timing of spawning is also known for the final two species G. divergens (Hopkins 1971) and Cheimarrichthys fosteri (McDowall 1990). Similarly, for the bully species not listed by Ryan (1997) the spawning behaviour and habitats of G. huttoni and G. breviceps are well known (McDowall 1990) and spawning timing understood for G. hubbsi. Finally Eldon (1978) describes the spawning habitat and behaviour of N. apoda. Therefore while not all spawning habitats of native fishes are known, there is a good understanding of the spawning biology of many of the species likely to be present in the Timberlands forest areas. Furthermore on-going research is targeting the threatened species for which data is absent and management programmes can be modified if necessary.

"Timberlands' statement (3.2.2.2) that "The vulnerability of fish species in timber production forests almost universally relates to matters of water quality, sedimentation and temperature" is manifestly untrue. There are also issues of: Shade as a form of cover; bank stability and instream woody debris as forms of instream cover; and terrestrial invertebrates as food, none of which get any or much consideration. Also there are changes in flow patterns, including greater flow fluctuations leading to higher peak flows and lower minimal flows following forest removal"

The key words in this statement are the last three "following forest removal". The sustainable beech management plan will not be leading to forest removal. The intention is to mimic the natural forest gaps. Shade as a form of cover is not important for many of the native fish species (e.g. the bullies, G. vulgaris and G. divergens), most of the species listed as likely to be present also exist in the open braided rivers elsewhere in New Zealand. G. argenteus, G. postvectis and G. fasciatus are fishes common in forested streams. However, riparian margins of 20 m can be expected to provide suitable shading. For example, the largest G. postvectis population known in New Zealand resides in a pasture stream with a riparian buffer strip, and not a completely forested area. Furthermore, the management of slash material can be used to provide instream cover improving G. argenteus habitat. Bank stability also does not seem likely to be compromised as the remaining stumps and root wads will stabilise the banks until they decay, by which time new seedlings should have established.

"The absence of any mention of trout, and their habitats is quite astonishing. Trout may be introduced, and predators that affect indigenous fishes, but they are also a resource that is highly valued by anglers and deserve some consideration. "

Section 6.8 of the Timberlands West Coast Ltd resource consent application notes that Fish and Game West Coast has identified important recreational fisheries and these will be incorporated into the riparian margin management. Ryan (1997) also notes that Fish and Game West Coast was consulted during the formulation of riparian management guidelines. Important trout fisheries and spawning habitats will be in the larger rivers and streams that automatically receive high priority status for riparian management. Therefore protection is provided and this meets the needs of the relevant fisheries managers. Small streams (up to 3m wide) which will have lower priority riparian zones are rarely if ever fished so provide no recreational return and any impact that occurred on such salmonid populations would not reduce recreational fishing opportunities.

5. Adjustment to management practises.

The management plan often refers to ephemeral streams, especially with regard to low priority riparian management zones. Many submissions raise concerns over whether the low priority zones are appropriate. To some extent this concern may be lessened by the defining of what Timberlands West Coast is referring to as an ephemeral stream. Ephemeral streams should be defined as streams that do not maintain a permanent flow and are dry for some periods of the year. These streams are unlikely to contain a significant fish fauna apart from possibly N. apoda due to their temporary or irregular presence. These will also not form important spawning habitats for fish species, except again for N. apoda. Such streams are distinct from small permanently flowing streams that do have the potential to contain significant native fish faunas.

The road culvert construction protocols that include the requirement for fish passage are appropriate in the Grey and Inangahua working circles. However, in the Maruia working circle culvert construction and maintenance should include an assessment of fish present before determining the fish passage requirements. Introduced fish species (generally brown trout) have been shown to have deleterious impacts on non-migratory galaxiids. Therefore prior to providing or improving fish passage at existing culverts the fish species present above and below the culvert should be determined. If introduced fishes do not occur upstream of the culvert no fish passage should be provided, although passage could still be provided for climbing native species such as anguillids. Similarly when constructing new culverts there is the possibility of blocking introduced fish passage to then restore upstream areas for non-migratory native fishes. It is expected that opportunities of this nature will be limited, but if possible then the road construction can provide positive conservation outcomes for the declining non-migratory galaxiids. Examples of this nature are known in the pine plantation forestry areas of the North Bank of the Wairau River (Marlborough) and Waipori River (Otago).

The sustainable management programme will remove windthrow trees as part of the harvesting regime. It is recommended that any windthrows that lie in water course are left in place. These naturally occurring accumulations of woody material in streams provide important habitats for stream invertebrates and fish. Fisheries management also often includes the provision of woody debris in streams to increase habitat complexity and provide cover for fish species. It is possible that slash material could be left in streams to provide habitat for G. argenteus and Anguilla species. Recent experimental (Bonnett pers. com.) and survey evidence (Bonnett pers com., Allibone pers. obs.) show G. argenteus has a strong preference for instream cover and overhead shade is of less significance. The placement of slash, without leaf material, in streams would provide such cover and more than compensate for any loss of overhead shading due to the tree removal. This could be carried out initially in one of the streams in the fisheries monitoring programme. Areas with slash material that was left in the stream could be monitored to determine if G. argenteus and Anguilla species colonise these areas of instream cover.

The management plan calls for monitoring of freshwater fishes to determine if the sustainable logging programme is having impacts upon the fish populations. One set of paired catchments is not sufficient for the monitoring. This is because the distribution of fishes in the Buller and Grey catchments are such that no set of paired catchments will have all vulnerable fish species. It is recommended that monitoring sites are selected in one of either the Grey or Inangahua working circles targeting migratory species and a second monitoring area is set up in the Maruia working circle for non-migratory species. It should also be noted that a monitoring programme has a limited power to detect changes in the abundance of the native fish and directly link this to impacts of the harvesting programme. This is due to a lack of historic data on the fish populations in the harvesting areas and a limited understanding of the links among recruitment process, life history and responses of these parameters to environmental changes.

6. Conclusion

It is my opinion that once knowledge of the fisheries present in the beech forests is attained the sustainable beech forest management plan of Timberlands West Coast Ltd presents has a very low risk of impact to the native fisheries.

7. References

Allibone, R. M.; G. P. Wallis 1993: Genetic variation and diadromy in some native New Zealand galaxiids (Teleostei: Galaxiidae). Biological Journal of the Linnean Society 50: 19-33.

Allibone, R. M.; C. R. Townsend 1997: Reproductive biology, species status, and taxonomic relationships of four recently discovered galaxiid fishes in a New Zealand river. Journal of Fish Biology 51: 1247-1261.

Cadwallader, P. L. 1976: Breeding biology of a non-diadromous galaxiid, Galaxias vulgaris Stokell, in a New Zealand river. Journal of Fish Biology 8: 157-177.

Eldon, G. A. 1978: The life history of Neochanna apoda Günther (Pisces: Galaxiidae). Chirstchurch, Ministry of Agriculture and Fisheries, Fisheries research Bulletin 19: 44p.

Hopkins, C. L. 1971: Life history of Galaxias divergens (Salmonoidea : Galaxiidae). New Zealand Journal of Marine and Freshwater Research 5: 41-57.

Hopkins, C. L. 1979: Reproduction in Galaxias fasciatus Gray (Salmoniformes : Galaxiidae). New Zealand Journal of Marine and Freshwater Research 13: 225-230.

Huryn, A. D. 1998: Ecosystem-level evidence for top-down and bottom-up control of production in a grassland stream system. Oecologia 115: 173-183.

McDowall, R. M. 1990: New Zealand freshwater fishes: a natural history and guide. Auckland, Heinemann Reed 553p.

McDowall, R. M. 1997: Two further new species of Galaxias (Teleostei: Galaxiidae) from the the Taieri River, southern New Zealand. Journal of the Royal Society of New Zealand 27: 199-217.

McDowall, R. M.; W. L. Chadderton 1999: Galaxias gollumoides (Teleostei: Galaxiidae), a new fish species from Stewart Island, with notes on other non-migratory freshwater fishes present on the island. Journal of the Royal Society of New Zealand 29: 77-88.

McDowall, R. M.; G. P. Wallis 1996: Description and redescription of Galaxias species (Teleostei: Galaxiidae) from Otago and Southland. Journal of the Royal Society of New Zealand 26: 401-427.

Mitchell, C. P.; B. P. Penlington 1982: Spawning of Galaxias fasciatus Gray (Salmoniformes: Galaxiidae). New Zealand Journal of Marine and Freshwater Research 16: 131-133.

Mitchell, C. P. 1991: Deposition of Galaxias fasciatus eggs with Galaxias maculatus eggs at a tidal site. New Zealand Journal of Marine and Freshwater Research 25: 201-205.

O'Connor, W. G.; J. D. Koehn 1998: Spawning of the broad-finned Galaxias Galaxias brevipinnis Günther (Pisces: Galaxiidae) in coastal streams of southeastern Australia. Ecology of Freshwater Fish 7: 95-100.

Ryan P. 1997: The potential impacts of the sustainable management of West Coast beech production forests on native fish. Ryan Environmental Report 3./96. Prepared for Timberlands West Coast Ltd. 29p.