Our purpose is to preserve New Zealand's whitewater resources and enhance opportunities to enjoy them safely.

Buller WCO Martin Unwin Evidence

IN THE MATTER of the Water and Soil Conservation Act 1967
IN THE MATTER of an application for a Water Conservation Order pursuant to section 20A of the Act
IN THE MATTER of an inquiry into the Draft National Water Conservation (Buller River) Order 1989

Brief of Evidence of Martin John Unwin


My name is Martin John Unwin. I am 43 years old, and have lived in Christchurch for 21 years. I have a MSc (in Physics) from the University of Canterbury, and am currently employed as a scientist with the National Institute of Water and Atmospheric Research (NIWA). In presenting this evidence on behalf of the NZCA, I am speaking as a private individual with an interest in canoeing, rather than as an employee of NIWA.

Prior to the formation of NIWA, I worked as a scientist with MAF Fisheries from 1976 to 1992. During this time, I was closely involved with analysis and write-up of the National Angling Survey (NAS), conducted by MAF during 1979/80. This was a nation-wide postal survey designed to measure the relative importance of angling rivers throughout New Zealand, and hence to identify river fisheries of national, regional, and local importance. In the course of this work, I co-authored over 20 reports based on the survey results.

I began white water kayaking in 1979, in association with the University of Canterbury Canoe Club. I have taken part in kayaking trips throughout the South Island, including several extended trips in the Murchison area paddling the Buller River and its tributaries. I have paddled all but one of the recognised whitewater rapids on the Buller, and also the Gowan, Mangles, Matakitaki, and Maruia Rivers.


The 1991 NZCA River Use Survey

The 1991 River Use Survey (RUS) was initiated by the NZCA in recognition of the need for up-to-date information on river usage by NZCA members. The objectives of the RUS were (1) to measure the relative importance of New Zealand rivers to NZCA members; (2) to measure the usage of New Zealand rivers by NZCA members; and (3) to develop a database for future advocacy of New Zealand rivers.

The survey was implemented via a questionnaire distributed to affiliated member clubs throughout the country, and hence to individual NZCA members. The key item of this questionnaire was a table listing 200 river sections, chosen after consultation with each club and grouped by geographical region, with a matrix of data columns against each river for respondents to complete as appropriate. The full questionnaire took the form of an 11-page booklet, including a covering letter explaining the purpose of the survey, a list of club contacts, and a detailed set of instructions and guidelines. A copy of the questionnaire is included with this evidence as Appendix I Not online.

Many rivers, including the Buller River, provide differing canoeing opportunities in different reaches. Where appropriate, these reaches were listed separately in the RUS questionnaire. For the Buller River four sections were listed, as detailed in Table 1.

Table 1: Key river sections on the Buller
River Section Key Rapids/Features (grades dependent on flow)
Lake Rotoiti to Owen River Rapids from lake outlet to Teetotal Hut (II/III); Granity Creek rapid (IV); 2 Mile Island slalom course.
Owen River to Murchison No major rapids; overall Grade 1/II. Camping areas at Owen River, Mangles confluence, Murchison.
Murchison to Lyell Creek O'Sullivans Rapid (III/IV); Whale Creek Rapid (III/IV); Jet Boat Rapid (III/IV); Ariki Falls (IV); Earthquake Slip (III/IV).
Below Lyell Creek Lower gorge; little white water, Grade I/II.

Respondents were asked to identify any river sections which they had paddled, and to provide the following information for these rivers: the total number of canoeing trips made to the river; the importance of the section to them as a canoeist, on a scale of 1 (lowest) to 5 (highest); and the travel time, scenic beauty, wilderness feeling, and degree of canoeing challenge, also rated on a 1-5 scale. To maximise consistency between respondents, guidelines detailing the characteristics associated with each 1 to 5 grade were included with the survey instructions, so as to establish a clearly delineated hierarchy of responses for each attribute.

Survey Results

A total of 604 replies were received from whitewater paddlers. This represents approximately 50% of the affiliated NZCA club members who participate in whitewater paddling. Collectively, these respondents provided 9,788 assessments of individual rivers, representing a total of over 35,000 paddling trips.

For the purposes of this evidence, I have identified 103 river sections, on 70 rivers, comprising all sections paddled by at least 25 respondents. Each of these rivers is listed in Appendix 2 Not online, together with the following statistics and their rank (from 1 to 103): the mean grade for "overall importance"; the total number of respondents who had paddled that river; the total number of trips made on that river; the average number of trips per paddler; the average value for the four attributes used to characterise each river (travel time, scenic beauty, wilderness feeling, degree of canoeing challenge); and a "geographic diversity" index which measures how widely respondents travelled to reach each river.

Measures of Importance

Appendix 2 Not online provides five statistics which directly relate to the concept of importance or relative value. These are the overall importance grade; total number of paddlers; total number of trips; the canoeing challenge grade; and the geographic diversity index. For each of these measures, the top ten ranking rivers are listed in Table 2.

Table 2: The 10 highest ranking rivers for each of five measures of relative importance, based on the 1991 NZCA River Use Survey.
Measure of relative value
Rank Overall importance Number of paddlers Number of trips Canoeing challenge Geographic diversity
1 Motu Waikato
Karamea Buller
(Owen - Murchison)
2 Karamea Otaki
(Nga awapurua)
(Other section(s)) *
(Aniwhenua - Matahina)
3 Landsborough Buller
(Murchison - Lyell)
Landsborough Matakitaki
4 Mohaka
(Te Hoe - Willowflat)
(Reids farm)
(Murchison - Lyell)
5 Kawarau
(Roaring Meg)
(Owen - Murchison)
(Reids Farm)
(Above Deep Creek)
6 Buller
(Murchison - Lyell)
(Rotoiti - Owen)
(Utiku - Mangaweka)
Wairoa Kawarau
7 Kawarau
(Potu - Boulder Reach
Motu Buller (Murchison - Lyell)
8 Shotover
(Below Deep Creek)
(Utiku - Mangaweka)
(Potu - Boulder reach)
(Below Deep Creek)
9 Kawarau
(Other section(s)) *
Matakitaki Buller
(Rotoiti -Owen)
(Te hoe - Willowflat)
(Wheao - Murupara)
10 Rangitikei
(Matawhero - Pukeokahu)
Waimakariri Rangitaiki
(Wheao - Murupara)
(Murchison - Lyell)
(Rotoiti - Owen)
Note 1: Rivers or sections of rivers within the Buller catchment are indicated by light shading.
Note 2: Reaches of the Kawarau River other than Dogleg and Roaring Meg were not listed explicitly in the survey questionnaire, but would include rapids such as Sargoods Weir (now submerged under Lake Dunstan), Citroen, Retrospect, and Nevis Bluff.

Overall Importance

The average importance grade for each river is a measure of how it is regarded by respondents who have paddled it. For the 10 highest ranked rivers, the mean importance grade ranged from 4.30 to 4.68. The Buller River, from Murchison to Lyell, had a mean importance grade of 4.41, the 6th most highly ranked river in New Zealand. Of the respondents who had paddled this section of the Buller, 53% gave a grade of 5 ("Extremely important; this river offers a unique canoeing experience"), and 35% gave a grade of 4 ("Important; this river provides an experience exceeded by few other rivers").

Number of Paddlers and Number of Trips

The number of respondents who paddled each river, and the total number of trips, provide two measures of the level of usage of each river. In general, these two statistics are highly correlated, and yield similar rankings; six rivers in the top ten for number of paddlers were also in the top ten for number of trips.

All three mainstem sections of the Buller River above Lyell were ranked in the top ten for number of paddlers. The section from Murchison to Lyell was ranked 3rd; from Owen River to Murchison was ranked 5th; and from Rotoiti to Owen River was ranked 6th. The Matakitaki River was ranked 9th. Two sections of the Buller River were ranked in the top ten for number of trips: the Murchison to Lyell section was ranked 4th, and the Rotoiti to Owen River section was ranked 9th.

To compare usage of the Buller River with other multi-section rivers listed in the questionnaire, usage estimates for the eight most visited rivers were derived from Appendix 2 Not online, taking into account respondents who had paddled more than one section of each river. These estimates are listed in Table 3. The Buller River received the highest number of trips (3052), slightly ahead of the Waikato River (2788) and the Tongariro River (2459). These three rivers also attracted similar numbers of paddlers, ranging from 272 (Tongariro) to 264 (Buller). The Buller River was paddled by just under half (43.7%) of the NZCA whitewater canoeists who responded to the survey.

Table 3: Total usage (number of paddlers and number of trips) for eight multi-section rivers, based on data from the 1991 NZCA River Use Survey.
(number of sections)
Total trips
(10% trimmed)
Number of paddlers % of respondents
Buller (4)3052 *26443.7
Waikato (3)278826844.4
Tongariro (4)245927245.0
Manawatu (2)194414423.8
Otaki (2)164519732.6
Rangitikei (5)161921134.9
Hurunui (3)161317328.6
Kawarau (3)142613923.0
* Includes respondents who paddled the Buller River, but did not specify which sections

Canoeing Challenge

The average value of the "canoeing challenge" grades for each river provides a fourth measure of its value to NZCA members. In general, the average grade for "canoeing challenge" was directly related to the average grade for "overall importance". Rivers considered to present the greatest challenge also tended to be highly ranked for "overall importance".

The Murchison to Lyell section of the Buller River was ranked 10th for canoeing challenge, with an average grade of 3.90. Although ranked somewhat below the most extreme whitewater rivers (such as the Karamea, Landsborough, and Kawarau), this section of the Buller was regarded as challenging by most canoeists, receiving a grade of 4 (challenging; requires full use of my canoeing skills without actually extending them) or 5 (very challenging; the limit of what I feel to be within my canoeing ability) by 72% of the respondents.

Geographic Diversity

A fifth measure of the importance of each river is the extent to which it attracts canoeists from a wide geographical area. Other things being equal, the NZCA would tend to place a higher value on a river which is visited by canoeists from throughout New Zealand than on one which is visited mainly by local canoeists. For example, the Manawatu and Otaki gorges were very heavily used, but were paddled mainly by Wellington canoeists, suggesting that their importance is local or regional rather than national.

To quantify the geographic diversity of the paddlers visiting each river, respondents were grouped into one of seven regions, based on their home address. The number of respondents in each district was then used to calculate an index which measures how widely they were distributed throughout New Zealand (Appendix 2)Not online. For a river paddled by canoeists from only one of the seven districts this index would be 0, whereas for a river paddled by equal numbers of canoeists from all seven districts this index would be 1.

The Buller River attracted paddlers from a broader geographical area than any other river in New Zealand. In terms of the distribution index as defined in the preceding paragraph, the section of the Buller from Owen River to Murchison was the highest ranking of the 103 rivers listed in Appendix 2 Not online. Three other river sections within the Buller River catchment were also ranked in the top ten: these were the Matakitaki River (3rd), the Buller from Murchison to Lyell (7th), and the Buller from Rotoiti to Owen River (10th).

Other Attributes of the Buller River System

A total of 11 river sections within the Buller catchment were included in the RUS. As well as the four sections of the Buller River mainstem listed in Table 1, these included the Gowan, Owen, Mangles, Matakitaki, Matiri, Maruia, and Inangahua Rivers. Results for all these rivers are summarised in Table 4.

Table 4: 1991 River Use Survey results for the Buller River and its major tributaries.
Number of paddlers Number of trips
(10% trimmed)
Overall importance Travel time Scenic beauty Wilderness feeling Canoeing challenge Geographic diversity
River Name Total Rank Total Rank Mean Rank Mean Rank Mean Rank Mean Rank Mean Rank Mean Rank Index Rank
Buller (Rotoiti - Owen R) 176 6 703 9 4.45 25 3.91 27 3.80 80 3.54 37 2.80 55 3.29 28 0.78 10
Buller (Owen R - Murchison) 180 5 577 14 3.56 34 3.79 35 3.80 79 3.35 47= 2.67 63 3.07 43 0.84 1
Buller (Murchison - Lyell) 190 3 1062 4 6.21 11 4.41 6 3.98 91 3.67 30 3.04 41 3.90 10 0.79 7
Buller (Lower Gorge) 47 73= 88 80= 2.10 69 3.50 51= 3.89 86 3.55 36 2.93 44 3.02 46= 0.65 43
Gowan 65 54= 97 77 1.64 90 2.98 84 3.77 78 3.02 65 2.65 65 2.66 61 0.76 14
Owen 23 -* 28 - 1.33 - 3.10 - 3.96 - 2.96 - 2.55 - 2.81 - 0.72 -
Mangles 132 18= 264 37 2.22 66 3.18 71 4.05 93 3.29 51 2.71 58 2.64 62 0.74 19
Matakitaki 159 9 435 22 3.04 47 3.90 28 3.84 84 3.35 47= 2.93 45 3.44 24 0.80 3
Matiri 85 41= 138 62 1.79 83 3.55 50 3.93 87 3.45 44 2.98 43 3.20 36 0.70 31
Maruia 104 34 167 56 1.78 84 3.65 42 4.09 95 3.92 21 3.65 21 2.93 51 0.73 24
Inangahua 2 -* 2 - 1.00 - 3.00 - 4.50 - 2.00 - 1.00 - 2.00 - 0.00 -
* The Owen and Inanganhua Rivers were paddled by less than 25 respondents, and were therefore not included in the group of 103 rivers used to rank each attribute

A total of 4,183 canoeing trips were recorded within the Buller system, of which 73% took place on the mainstem and a further 10.4% on the Matakitaki River. The next three most heavily used tributaries, the Mangles, Maruia, and Matiri Rivers, collectively accounted for 569 trips, or 13.6% of the total.

The section of the Buller from Murchison to Lyell was by far the most highly ranked river in terms of overall importance and canoeing challenge. The remaining sections of the Buller mainstem, and the major tributaries, were most commonly assigned "importance" and "challenge" grades of 3 or 4, indicating that while they were considered to be moderately important and challenging, similar canoeing opportunities could generally be found elsewhere.

Grades for "travel time" were uniformly high (3.77 - 4.09) throughout the Buller catchment, with over 75% of the respondents indicating that canoeing the Buller River normally involved at least a full weekend trip. These ratings are consistent with the broad geographic distribution of Buller River paddlers, and with the popularity of the Murchison area as a summer holiday canoeing venue.

Rivers in the Buller catchment were regarded as scenically "attractive" or "very attractive" by 80% of the respondents, but apparently lacked the large-scale grandeur required to elicit a grade of 5 ("inspiring"). With the exception of the Maruia River (ranked 21st out of 103), most sections were ranked between 30 to 50. Similar comments apply to the grades for "wilderness feeling", indicating that while respondents felt isolated from human activity while on the river, roads could usually be reached without undue difficulty.

An important feature of the Buller catchment is the range of high quality canoeing opportunities it provides within a compact geographical area. Although the RUS did not specifically investigate whether proximity to other canoeable rivers significantly increased their value, the results provide considerable indirect evidence to suggest that this is the case. The level of canoeing activity within the Buller system, and also on the Kawarau and Shotover Rivers, is considerably higher than might be expected given their distance from major population centres. By contrast, paddling runs such as the Rangitata gorge and the Taieri gorge, both of which lie close to major population centres but are in areas where neighbouring rivers provide only limited canoeing opportunities, were relatively lightly used despite being very highly regarded.


Based on the RUS, four sections of the Buller River are of exceptional significance to NZCA members. These are the three sections of the mainstem from Lake Rotoiti to Owen River, from Owen River to Murchison, and from Murchison to Lyell; and the Matakitaki River.

Of these rivers, the Buller River from Murchison to Lyell was clearly the most highly regarded, and is arguably the most important whitewater canoeing river in New Zealand. In terms of the five measures of relative value listed in Table 2, this section of the Buller was the only river to achieve a top ten rank for all five measures, and the only river to be ranked in the top ten both for measures of usage (number of paddlers/trips) and for measures relating to the quality of the canoeing experience (overall importance, canoeing challenge).

The mainstem of the Buller River, from Lake Rotoiti to Lyell, was the most heavily used river identified by the survey. The Waikato and Tongariro Rivers supported a comparable level of usage, with both rivers including at least one highly regarded section, but neither was ranked as highly as the Buller in terms of overall importance and/or popularity with canoeists from throughout New Zealand.

The Matakitaki River was the most popular Buller tributary in terms of the number of paddlers attracted to it, being visited by more than 25% of the survey respondents. It was by far the most highly rated tributary in terms of overall importance and canoeing challenge, for which it was highly rated (although not in the top ten) in a national context.

In a national context, the Buller River is unique. In terms of the measures of importance provided by the RUS, the only other comparable river is the Kawarau. Both rivers were popular with canoeists from throughout New Zealand, and were regarded as highly important. However, while the Kawarau River (and its tributary the Shotover River) provide some of the most technically demanding whitewater in New Zealand, with few if any easy sections, it lacks the variety of rapids and range of river environments that make the Buller River accessible to canoeists of all skill levels. The Kawarau is a river for the experienced paddler; the Buller provides something for everyone. Moreover, although the RUS did not specifically examine whether proximity to other canoeable rivers contributed to their value to NZCA members, there is strong evidence that the diversity of canoing opportunities provided by the Buller tributaries contributes significantly to its importance.

On the basis of data collected via the 1991 NZCA River Usage Survey, the mainstem of the Buller River, from Lake Rotoiti to Lyell, and the Matakitaki River, are clearly of exceptional importance to NZCA members. On behalf of NZCA, I therefore support the application for a National Water Conservation Order on these rivers.

This site is a beta version.