DECLINE OF KAHAWAI - REPORT

Decline of the Kahawai Fishery

Report by Mark Feldman

late 1990's

The following report was prepared for one of the Ministry meetings that occurred in the late 1990’s. It is a collection of the scientific data available at that time that highlighted the declining catches and smaller size of kahawai available to the recreational sector.

Much of this data had not been integrated properly into the knowledge base and still isn’t. This information, when viewed as a whole, shows how much the kahawai fishery has declined.

The Status of the Kahawai Stocks from the Point of View of the Recreational Sector

Introduction
The Big Game Fishing Council, the Recreational Fishing Council and the vast majority of 400,000 marine recreational fishers believe the kahawai resource available to them has declined significantly over the past decade. We believe this decline in our resource is the direct result of excessive purse-seine fishing and an expanding commercial set net fishery.

Part II, Section 10a, of the new Fisheries Act declares, “ Decisions should be based on the best available information.” We believe the information we are presenting here is the best available for kahawai and is far more reliable than the simple model developed at the last Working Group meeting.

Instead of focusing on a speculative model it is important for the Working Group to keep in mind that we don’t know the MSY for kahawai, but we do know that the access of recreational fishers to the resource is poor, and that commercial fishing operations have taken a huge toll on the resource; about 45 million kahawai caught in the past decade.

There are two main aspects of the last Working Group report that we believe need to be changed. One is the assumption that recreational catches were lower in the 1980’s than they are now. We don’t believe that is correct. It’s much more likely that recreational and customary catches were two or more times today’s levels. We also believe that the emphasis of the Working Group needs to shift away from our focus on our simple model towards the available evidence that indicates a serious loss of access to kahawai for the recreational and customary fishers.

When evaluating the following arguments it is important to consider them all as a group. It is easy to nit-pick most of the individual arguments to death by demands for more accuracy, but when you consider all the arguments together it is simple to understand why we believe the recreational sector has been disenfranchised over the past decade.

The following arguments are divided into two aspects

1. The present recreational kahawai catch is very poor.
2. The recreational CPUE was much better a decade ago.

The Present Recreational Catch
Over the past few years a variety of boat ramp and diary studies have been completed. These studies have concluded that the average fisher catches 0.4 kahawai per person per fishing trip. We don’t believe any reasonable person would disagree that this is a very poor catch rate.

For those of you who are not recreational fishers it might be helpful to point out that the CPUE for snapper in the North Region is three times higher than the kahawai CPUE and it’s well established the snapper fishery is below the Bmsy in the North.

A Comparison of Recreational CPUE in the 1980’s and 1990’s.
The following five arguments, when considered as a group, strongly suggest that the recreational CPUE has declined dramatically between 1983 and the present. It is our belief that the average recreational CPUE in the early 1980’s was 1-2 kahawai per person per fishing trip. This is 2-5 times today’s CPUE.

Listed below are summaries of our five main points. Details on each of the five arguments can be found on the pages indicated at the end of each summary.

1. Changing pattern of tag returns in the Bay of Plenty between 1983 and 1991.
   A comparison of the tag returns from the kahawai tagging programs in 1983 and 1991 was made. The comparison was based on the tag returns from purse-seine operations and recreational rod and line fishermen; it was limited to the Bay of Plenty.
   In 1983 the rod and line fishers returned 72% of the tags. By 1991 that proportion had dropped to 27%. During the same period the proportion of kahawai tags returned by the purse-seine operators rose from 28% to 73%.
   We believe this reflects the decline in kahawai available to the recreational sector because of excessive commercial fishing activities.
   A more detailed explanation follows on page 4.
2. Fishing magazines surveys of recreational kahawai fishers.
   In 1997 both major fishing magazines cooperated to distribute a kahawai survey to their members. A total of 2002 replies were received. Of those 2002 replies 47% felt that kahawai stocks had declined significantly over the previous five years, 32% felt the stocks had declined a little, 11% felt they were the same, 8% felt they had increased a little and 2% felt they had increased significantly. Most of these fishermen had over 20 years experience. Experience, age and home location did not influence the answers to questions about the status of the kahawai stocks. Similar results were obtained during surveys in 1989, 1990 and 1992.
   A more detailed explanation follows on page 7.
3. The fork length of kahawai has declined in the Bay of Plenty.
   We believe there has been a decline of 5.7cm in the fork length of purse-seine caught kahawai in the Bay of Plenty between 1983 and 1992. This issue has been debated at the Pelagic Working Group before. However, there were two significant errors made at the Working Group and two items of new information, including a wide ranging evaluation of kahawai fork length between 1983 and 1991. All the data presented indicates that the fork length of kahawai available to the recreational sector has declined over the past decade.These are described in detail on page 9.
4. The recreational CPUE had declined dramatically in the Bay of Plenty.
   From January – April 1982 a recreational survey at the Motu River mouth in the Bay of Plenty revealed a CPUE of 2.55 kahawai per hour for visitors to the area and 4.17 fish per hour for the local residents.
   From March – April 1991 a MAF survey of the area from Opotiki to Te Kaha (includes the Motu) showed a CPUE of 0.1 kahawai per hour. This comparison is not ideal because the survey area in 1991 included areas outside the Motu and a different time period. However, the surveyor in 1991 has stated that catch rates at the river mouth were approximately the same as other areas along the beach, and peak catches in the area generally occur in March (a month included in both surveys).
   This issue has been discussed at the Working Group before but there is new evidence made available from a more detailed breakdown of the 1991 MAF survey and an interview with the surveyor in 1991.
5. Declining purse-seine catch in KAH3
   For years now the purse-seine vessels have been unable to catch their limit in KAH3. With two boats operating in KAH3 the purse-seiners were able to catch up to 5000 tonnes per year in the late 1980’s. From 1991-92, 92-93, 93-94 and 94-95 these same two boats were unable to catch their limit at any time. This suggests their CPUE is 1/2 to 1/3 of what it used to be in the late 1980’s.
   If the purse-seiners, guided by airplanes, cannot land kahawai in KAH3, it’s easy to understand why recreational fishers feel they can no longer catch kahawai either. Given that it is much easier to catch kahawai with an airplane, its reasonable to assume the recreational CPUE in KAH3 is less than 1/3 of what it used to be in the 1980’s.

Conclusion
It is widely accepted by stock assessment scientists that decreasing fork length, declining CPUE for commercial vessels and poor recreational catch rates are the early signs of a declining fishery. We believe that all these factors now apply to kahawai.

The new Fisheries Act has made it clear that our priority is to be “cautious when information is uncertain, unreliable or inadequate”. We believe that the last report from the Working Group failed to do this.

The emphasis of the Working Group needs to shift from relying on a poorly developed model, towards focussing on the well established fact that there are no longer enough kahawai to supply the needs of recreational and customary fishers in NZ.

To achieve this goal the report of the Working Group needs to be changed as follows:

1. Under ‘Recreational Fisheries’ emphasis needs to be placed on the poor recreational CPUE in the key area of KAH1. We also need to point out there has been a decline in the size of kahawai available to the recreational fisher as well as the number.
2. Under ‘Biology’ lower values of “M” need to be considered and perhaps other changes made as well. Clearly there’s a problem with our model since it does not correspond to the experience of fishers on the water.
3. Under ‘Stock Assessment’ the assumptions about past recreational catches need to be increased to accommodate the information we’ve presented here. We suggest the probable recreational catch in the 1980’s was probably 4000-5000 tonnes.
   In addition, the emphasis of the entire section needs to shift away from our poorly developed model towards the evidence we’ve presented here because this evidence represents the “best available information” required under the new Fisheries Act.
4. Under ‘Status of the Stocks’ the basic message needs to change from a positive one to an acknowledgement that recreational and customary fishers have inadequate access to the kahawai stocks because of excessive commercial fishing over the past decade.

TOP

A Comparison of the Rates of Recovery by Purse-seiners and Rod and Line Fishers of Tagged Kahawai from the 1983 and 1991 Tagging Studies in the Bay of Plenty

Summary
A comparison of the tag returns from the kahawai tagging programs in 1983 and 1991 was made. The comparison was based on the tag returns from purse-seine operations and recreational rod and line fishermen; it was limited to the Bay of Plenty.

In 1983 the rod and line fishers returned 72% of the tags. By 1991 that proportion had dropped to 27%. During the same period the proportion of kahawai tags returned by the purse-seiner operators rose from 28% to 73%.

We believe this indicates the recreational kahawai catch was much higher in the 1980’s than it is now, and the number of kahawai available to the recreational sector has declined significantly. A change needs to be made to the Working Group document to reflect the probability that the recreational catch and CPUE were higher in the 1980’s than they are now.

Background
The boat ramp and diary studies done by the Ministry during the 1990’s have revealed an average catch of one half of a kahawai per person per trip in the North Region; a poor catch by any measure.

Unfortunately there is very little data available to quantify the recreational catch in the mid 1980’s, before the sharp increase in purse-seine catches that were associated with the introduction of the QMS. Up until now only one study was available; a CPUE study at the Motu River mouth done in 1982. This study was never repeated in exactly the same manner, but a comparable study of the Bay of Plenty in 1991 suggested a severe decline.

In 1997, the results of the 1991 kahawai tagging study became available. For a number of reasons outlined by Elizabeth Bradford in her 1997 paper called, “Estimation of Kahawai Recreational Catch from Tagging Returns ….” It was not possible to use either the 1983 or the 1991 tagging studies to calculate a kahawai biomass or quantify the recreational catch. In both of these studies all the tags recovered were not returned, the amateur and commercial effort distributions were not the same and the tagged kahawai were probably not evenly distributed in the population.

Since both the 1983 and 1991 tagging studies suffer from the same flaws in the same way it is reasonable to assume the effects smooth out when the studies are compared. For instance, its well known that tags from individually handled fish are more likely to be returned than tags from purse-seine caught fish. That is not a problem if you only want to compare the studies, because tags recovered by purse-seiners had the same chance of being returned in both 1983 and 1991. Likewise, tags recovered by hook and line fishers had the same change of being returned in 1983 and 1991.

In a similar fashion, the differences between the amateur and commercial effort distributions were likely to be the same in 1983 and 1991, and the irregularities of the mix of tagged kahawai with the general population were probably equivalent too.

In 1983 most of the kahawai tagging operations were performed from June to September. In 1991 the tagging was completed in July. During the 1980’s- 1990’s the total recreational fishing effort was probably stable with population increases offset by a lower percentage of people fishing. Weather conditions, which influence the total recreational catch, were actually better after the 1991 tagging effort than after the 1983 effort, so weather was not the cause of the decline in the proportion of kahawai caught by the recreational sector. The total purse-seine catches in KAH1 during the periods 1983-87 and 1991-95 were almost identical so that is also not a factor in the relationship. The percentage of tag returns in the 1983 study was 9.7% (138/1427) and the percentage for 1991 was 11.9% (551/4622); certainly comparable return rates.

Results
During both the 1983 and 1991 tagging studies large numbers of kahawai were tagged in the Bay of Plenty. Records of tag returns were kept over the following years. These tag returns were divided between the purse-seine fishery and the recreational rod and line fishery. By using these tag return figures it is possible to calculate a relationship between the proportion of tag returns by the purse-seiners versus rod and line fishers in 1983 and 1991.

In 1983 72% of all tags returned in the Bay of Plenty were from rod and line fishers. In 1991 that proportion declined to 27% of all tag returns. The reverse was true for the purse-seine fishery; in 1983 28% of tag returns were from the purse-seiners. By 1991 that proportion had increased to 73% of returns.

Data and Sources

A Comparison of Total Tag Returns Between the 1983 and 1991 Kahawai Tagging Studies in the Bay of Plenty

Summary
In 1997 both major fishing magazines cooperated to distribute a kahawai survey to their members. A total of 2002 replies were received. Of those 2002 replies 47% felt that kahawai stocks had declined significantly over the previous five years, 32% felt the stocks had declined a little, 11% felt they were the same, 8% felt they had increased a little and 2% felt they had increased significantly. Most of these fishermen had over 20 years experience. Experience, age and home location did not influence the answers to questions about the status of the kahawai stocks.

Background
Recreational fishing surveys have been performed in 1989, 1990 and 1992. The surveys had between 400-1000 replies each. All of the surveys showed that over 80% of recreational fishermen believed the kahawai resource was in decline. The average respondent was over 35 years of age with 15 years of fishing experience; so these survey results reflect a considerable amount of expertise over time.

Survey Results: Key survey results are as follows:

How many years have you been fishing?

How many days a year do you go fishing?

On what percentage of your trips do you target kahawai at least some of the time?

Think back five years. In the intervening time do you think the kahawai stocks have…

A copy of the survey follows.

TOP

1997 Kahawai Survey

Preliminary results: 1709 replies
Question Five
Think back five years. In the intervening time do you think the kahawai stocks have…

Question Two:
How many years have you been fishing?

TOP

The Decline of Fork Length of Kahawai in the Bay of Plenty Purse-seine Fishery from 1983 to 1992

Summary
We believe there has been a decline of 5.7cm in the fork length of purse-seine caught kahawai in the Bay of Plenty between 1983 and 1992. This issue has been debated at the Pelagic Working Group before. However, there were two significant errors made at the Working Group:

1. We were misinformed about the nature of the 1983 catch samples. We were told the 1983 fish selected for measurement were the larger fish available. This is not correct. They were selected at random.
2. We were not presented with the results of a boot-strap analysis that concluded there was a 95% chance the 1991-92 catches did, indeed, have a smaller fork length than in 1983.

In addition, some new evidence has come to light about the fork length of line caught fish in 1983 which shows a general decline in the fork length of line caught fish throughout the North Island when compared to 1991 data.

Available Information
As part of a 1983 tagging study of kahawai the fork length of 332 kahawai were measured in the Bay of Plenty. Of these 332 fish, 32 were line caught. Their median length was 49.8cm (see page 12 for details).

The other 300 fish were purse-seine caught; three landings were sampled with 100 fish sampled in each landing. The first two landings measured represented a single school each. The last landing may have involved more than one school (Gavin James provided this information). The median length of the three landings of purse-seine caught fish were 52.7, 49.3 and 51.8cm. These fish were selected at random. The average of these means is 51.3cm.

In 1991-92 a shed study was done on kahawai. In 1991 five landings were sampled with 6778 fish measured. These landings also represented purse-seine targeted schools. The mean size was 46.1cm.

In 1992 seven landings were sampled. These landings also represented purse-seine targeted schools. A total of 12,431 fish were measured. The mean size in 1992 was 45.25cm.

The average of the 1991-92 means was 45.6cm.

Just recently an analysis was done comparing line caught kahawai in 1983 to line caught fish in 1991 in various locations around the North Island. In every location there was a decline in the fork length of the fish. This data is available on page 15.

The majority of line caught fish in 1983 were caught trolling. Most fish in 1991 were caught on bait. We were concerned about a possible difference in size of fish caught, based on whether a lure or bait was used. An analysis of available 1991 data indicates bait fishing usually catches the biggest kahawai so that cannot be a factor in the decrease in fork length we found. That data is available on page 16.

TOP

Mean Fork Lengths of Kahawai
(target purse-seine only); 1983 vs. 1991-92

Discussion
Several points need to be made about the available data.

1. It would be difficult to imagine how the combined commercial and recreational fisheries in the Bay of Plenty could remove around 40,000 tonnes of kahawai (about 25 million fish) in a decade without a decline in fork length.
2. One of the arguments against accepting the 1983 data as representative is that kahawai school by size; therefore just sampling three or four schools is not adequate. It is true that kahawai school by size; never-the-less, those sizes vary widely. Even a casual glance at the graphs presented on pages 12 and 13 shows a 20cm range in the size of fish in each school. That’s a big difference; it increases the chances our 1983 sample is truly representative of purse-seine caught fish at the time because the concept of kahawai schooling “by size” is really not so accurate.
3. In 1994 Brian Jones did a bootstrap simulation of the data presented in Table 2 from his 1994 stock assessment paper (page 11 of this report). He selected any three of the 1991-92 means at random 600 times. The results were that 95% of the time the 1991-92 catches had a smaller fork length than in 1983.
4. The 32 line caught kahawai in the 1983 study had a mean size of 49.8cm, a figure close to the size of the purse-seine caught fish in 1983 and much larger than the sizes from the boat ramp surveys (42.1cm in 1991 and 44.1cm in 1994) this decade. This data is available on page 12.
5. The data presented on page 15 compares the fork length of line caught kahawai between 1983 and 1991. At all the locations with comparable data available the fork length of kahawai have declined. This further supports our argument that the fork length of kahawai available to recreational fishers has declined over the past decade.
   
   Table 2: Purse seine landings of kahawai sampled in the Bay of Plenty in 1983, 1991 and 1992 (n = number in sample, Mean = mean fork length, s.d = standard deviation).

(From: Kahawai information presented at the 1994 Stock Assessment by J.B. Jones)

TOP

Biological Sampling

Data on length frequencies, sex composition, age frequencies (otolith readings), and stomach contents were collected from commercial landings and during tagging studies from damaged fish.

Age and length frequencies
Little information on kahawai spawning and nursery areas is available, but it appears that most sheltered bays and estuaries in the North Island are used as nurseries, especially those off the east coast, north of the Bay of Plenty. Apart from the sheltered estuarine waters in Tasman Bay and near Farewell Spit, juveniles have not been found in substantial numbers in South Island waters (NZ Ministry of Agriculture and Fisheries unpublished data).

Age and fork length measurements were taken from kahawai caught in several areas (Table 7). Although the catching methods varied, the lengths of the fish en each area did not vary with the method used (Tables 8a-d). Fish caught by purse seine were assumed to represent local fish because the mesh size of the nets was small enough to retain juvenile, as well as mature, kahawai. Relatively more small fish were caught by line than by purse seine, but this was probably because lining was usually used to catch fish in sheltered nearshore waters where smaller kahawai are often found. Within each area fish size did not vary substantially between schools, though fish in one of the two schools sampled from east Tasman Bay on 21 April 1983 were reported as being larger than usual for the area.

Although almost 20% of the sample taken by setnet from the Waitaki River in 1984 comprised small fish of about 40cm, the sample taken by line in 1983 from this area had no fish of this size. It is unlikely that this absence of small fish resulted from the fishing method used, because 40cm fish were caught by lining in other areas that year; it is more probable that there were no small fish in the sampling area in 1983. Excluding this sample, the length frequencies of fish caught by the two methods were similar.

Whole otoliths were read by the method described by Eggleston (1975). For otoliths which required burning to read, rings were clearer when the otolith was sectioned and polished before burning (Paul 1976, James 1984). The age – length relationships were similar to those reported by Eggleston (1975). An age – length frequency plot for all samples combined is given in Table 9. Although mean length and age increased with latitude on the east coast South Island, this trend was not apparent in other areas (see Table 7). Movement of tagged fish between the North and South Islands..

Table 7: Age and length data for kahawai from areas sampled

* Standard deviation
+ Minimum to maximum age

* Standard deviation
+ Minimum to maximum age

Table 8a: Length frequencies for Bay of Plenty samples by date and method of capture

TOP

Appendix 1

Appendix 2