| Home » Fisheries Research and Development » 13th Salmonid Habitat Enhancement Workshop »
Restoration of a self-sustaining Atlantic salmon population to the river Lagan, Belfast, Northern Ireland
Robert Rosell
Department of Agriculture for Northern Ireland, Aquatic Systems Group, Newforge Lane,
Belfast BT9 5 PX
Abstract
Atlantic salmon became extinct in the River Lagan, which enters the Irish Sea through the port of Belfast, Northern Ireland, between 1750 and 1800, the extinction coinciding with a period of major population growth, industrialisation and finally the construction of a navigable waterway based on the river.
From 1950 to 1990, water quality in the river improved as a result of improved sewage treatment, the Lagan Navigation was abandoned and fell into disuse, and many industrial effluents were diverted to sewer. These changes led to a feasibility study on re-introducing the salmon to the river being considered.
The feasibility study was carried out in 1988. It led to the conclusions that there was considerable freshwater habitat suitable for holding salmon fry available in tributaries and in the upper river, and that water quality was adequate to allow salmon to exist in the tributaries and to migrate through the main river. Problems were identified in the form of several impassable weirs, vestiges of the former navigation and linen mills, and in a general lack of clean spawning gravels to accompany nursery habitats.
In 1991, the first of a series of plantings of unfed fry and smolts sourced from the river Bush hatchery took place. The first adult salmon returned to the Lagan in 1993. Since 1993, experimental stocking has continued, with annual planting of fry, leaving some areas unstocked each year to look for progeny of natural spawning activity. Adult runs have risen to hundreds each year with an increasing Lagan bred wild component. One tributary has not been stocked since 1997 and has had two generations of wild spawned fish. The best year for returns to date was 2000 when an estimated 800 fish ascended the Lagan, almost reaching the provisional spawning target of 1000 fish. An estimated 200 of these 800 were wild-bred fish.
A programme of fish pass construction has given salmon access to the whole river system, though some weirs remain which delay upstream migration. Future work is aimed at improving access, restoring habitat available for natural spawning and addressing the remaining, largely agricultural water quality issues in the river catchment. Liaison is in place with engineering staff proposing to reinstate the navigation for recreational purposes, to ensure that the mistakes of the 1700s are not repeated, salmonid habitat is preserved and access past weirs is ensured. If these issues can successfully be addressed, there is every chance that a self-sustaining population of salmon can be restored to the Lagan within the next decade.
Introduction – The river Lagan, extinction of salmon and improvements leading to current restoration attempts.
The River Lagan (Fig. 1) drains a catchment of some 576 Km2 in the south-east of Northern Ireland and enters the Irish Sea through the port of Belfast. The catchment consists mainly of enriched agricultural grassland in the upper parts, with a lower section draining urban Belfast and the town of Lisburn. There is one significant tributary, the Ravernet river, and there are several minor tributaries. Water quality is generally fair though there are localised problems and occasional pollution incidents, mainly due to effluents from farms. Rainfall in the catchment averages around 1000mm per year and river flows range from 1 to 40 cubic metres per second.
Fig. 1. Location map
Belfast and the Lagan valley developed as a major industrial centre from the 1700s, accompanied by population growth, untreated sewage effluents, and discharges from industry direct to rivers and streams. In the late 1800s a canal was built from Lough Neagh to Belfast, using some of the river as a navigable waterway and diverting water from other areas to supply separate canal sections. Many linen mills were powered by water abstracted from the river by means of weirs and mill dams. Contemporary reports indicate that that there were conflicts between the linen industry and the Canal for supply of water implying that upstream migration would have been severely impeded even if weirs were passable to ascending salmon. While there are few surviving contemporary records prior to current work, the latest record of a salmon population in the river dates from 1744, prior to construction of the navigation. By 1830 accounts of the fish of the river Lagan made no mention of salmon, nor of there ever having been any.
During the 20th century the canal and linen mills gradually fell into decline, but population growth continued, and an estuarine barrage was built on the river. Water quality in the urban reaches of the river upstream of this barrage fell to nuisance conditions, with saline stratification the norm and very low dissolved oxygen levels. Most sewers of the late 19th and early 20th century were combined systems, carrying both rain and foul water. As a result, overflow of contaminated water to the river and its tributaries was a frequent occurrence. A fish survey in the early 1970s found no fish at all in the urban reach of river through Belfast.
Brown trout and several other species remained present in the upper reaches of the river throughout the worst of the downstream urban problems, indicating that the freshwater habitats for salmonids were largely intact. In latter half of the 20th century came improved sewage treatment, separation of foul and storm water in new developments, the decline of some polluting industries or the treatment of their effluents, and the abandonment of the Belfast to Lough Neagh Navigation. The 1980s saw some recreational angling for non-migratory fish developing in the Belfast reaches of the river, and there were very occasional reports of migratory salmonids (salmon or sea trout) being seen in the river.
This paper describes the process of experimental re-introduction of Atlantic salmon to the river Lagan from a state of total extinction to the beginnings of a breeding population. The project has been fish-centred from the outset, using planting of swim-up fry as the main stocking method, and using the experimentally stocked fish to test literature and survey based assumptions of habitat suitability, passability of weirs, and water quality requirements. The absence of salmon for 200 years before the experiments enabled initial trials to be conducted in the total absence of wild fish.
Restoration Phase 1: Feasibility study 1988-1989
The generally improving state of the river prompted government fisheries scientists to initiate a feasibility study on salmon re-introduction in 1988 and 1989. This feasibility study contained the following elements:
1 An assessment of passability or otherwise of all weirs on the river to ascending salmon, using the guidelines for fish pass construction then in use by UK government agencies.
2 Examination of water quality records for the river and its impounded estuary and assessment of these against the documented requirements of salmonids as described in the literature and laid down in European legislation on the water quality required to support fish life.
3 Semi–quantitative assessment of the potential available spawning and juvenile salmon habitats in the river to quantify the carrying capacity for salmon juveniles in the river system. Most of the river system was walked by two staff, assessing habitat quality visually against a simple semi-quantitative system The Scoring system used for these assessments was abstracted and simplified from the available literature and is given in table 1.
Table 1. Habitat scoring system
SALMON HABITAT GRADING SYSTEM USED IN 1988/9 LAGAN SURVEYS |
|
Nursery areas |
|
|
Grade A |
Depth 15-25 cm Gradient 1-8% Substrate stable cobble or boulder Some cover from banks |
|
Grade B |
Marginally outside grade A on one count only |
|
Grade C |
Well outside grade A on one or more counts |
|
Grade D |
Deep, channelised, silty etc – (useless) |
Spawning habitats |
|
|
Grade A |
Depth minimum 15 cm Flow 30 – 60 cms-1 70% of 30mm – 80mm diameter stones Near an adult holding area Nursery area within 100m downstream |
|
Grades B-D |
Failing as for nursery habitat (above) |
The feasibility study conclusions were:
Fifteen major weirs were surveyed from the tidal head to the headwaters. Most weirs were at least partly passable under some flows to adult salmon, but there were several key weir sites which were very severe obstacles and at which the feasibility study recommended demolition or fish pass construction. The most severe obstacles were encountered at sites where sluice gates for level control shut onto concrete aprons with no pool downstream and where over-spill took place some way downstream via a side channel or vestige of the former canal.
Water quality in the juvenile habitats was generally considered adequate, excepting occasional acute fish kills due mainly to silage or other agricultural effluents. There are potential problems for migrating fish in the middle to lower reaches of the river system, where summer water flows can be very low and the river sluggish for long stretches. Nevertheless there are time windows outside of the lowest flow summer periods when salmon can ascend and descend the whole river system to and from upstream spawning areas and nursery habitats. Example long section and temporal water quality records from 1989 the feasibility study are shown in Figs 2. and 3.
Fig. 2. Long section May to September Dissolved Oxygen level, River Lagan salmon feasibility study
Fig. 3. Annual temporal water quality variation, at the worst case site, River Lagan salmon feasibility study
The habitat surveys revealed that there are very significant areas of potential habitat suitable for salmon juveniles, but good spawning gravels are of very limited extent. Nursery habitat could be available to support 500,000 or more first summer fry allowing for the (cautious) holding targets of 6 per square metre in A grade and 3 per square metre in B grade. Many good nursery areas lack adjacent spawning areas. There is a need for rehabilitation to complete habitat sequences in many areas. Results on total quantities of habitat nursery available in the grades A to D were recorded per tributary or main river reach. Example data is shown in fig.4.
Fig. 4. Available nursery habitat, square metres, river Lagan and tributaries
Phase 2 – First stocking trials 1991 - 2000
Following the completion of the feasibility study, funds were sought to establish a full salmon restoration programme. Such large scale funding, however, did not prove forthcoming.
Small scale trials were begun in 1991, with spring planting of 29,000 unfed fry from the River Bush hatchery stock into Grade A or B designated juvenile habitat in the Lagan’s Ravernet tributary. The initial purpose of this work was to test the assumptions made in the feasibility study; i.e. that areas of the river system could support juvenile salmon. This proved to be the case, with an estimated 25% to 40% of spring stocked fry surviving to August and September electric fishing surveys in the summer of 1991. By April 1992, electric fishing surveys showed an estimated 2000 1+ smolts emigrating to sea, due to return in 1993 as 1SW adult salmon.
In late October 1993, reports were received of large migratory salmonids trapped downstream a weir in Lisburn , where the feasibility study had previously identified an impassable obstacle. Shortly prior to this an angler caught a single salmon on the outskirts of Belfast, an event generating considerable local media attention. Investigations revealed that the first salmon for almost 200 years had indeed returned to the Lagan and that local anglers had seen fish as early as June. These fish had not been reported to or seen by fisheries staff, despite regular visual checks at weir pools. Due to high flows permitting upstream passage, the ability of the fish to find their way upstream had been underestimated. Before the reports from the public, fisheries staff had missed these fish by inspecting weirs sites too far downstream. On collation of all reports of fish and subsequent visual counts at weirs, it was apparent that an estimated 200 fish had ascended the river in the first year of returns following re-introduction.
Since the first stocking trials, further stocking has taken place on small scale (20,000 to 50,000 fry) annually, with the exception of one season when no fry were available at the River Bush hatchery. Following the return of adult salmon to the river, stocking has been conducted at well defined discrete sites to leave large areas free to look for the progeny of any natural spawning.
During this phase of the project, the estimation of numbers of returning fish has been improved by the addition of an adult trap at a weir in the tidal head of the river, subsequently replaced by the installation of an electronic counter. The counter and trap are situated in a race feeding a fish pass, achieving partial counts as fish are sometimes able to leap the adjacent weir. Visual counts of fish have been carried out twice at the annual peak of the salmon run, allowing estimates of the adult run to date. These are based on an estimated 10% of fish using the route through the counter or trap.
Each spring, a number of smolts are released into the river, marked with PANJET™ blue dye marks, fin-clipped (adipose fin) and micro-tagged. A proportion of these are recaptured at a partial trap at the exit to saltwater, to estimate the total escapement of all smolts. In a number of years, there has been a major discrepancy between the smolt emigration estimate and that considered probable from the monitored survival of stocked fry. This has also correlated with the years when significant emigration of naturally spawned fish was expected, indicating some wild output from the river. Data for number of fry stocked, smolt output estimate and adult return estimate are presented in Table 2. Estimates of returning adult salmon are given in Fig. 5.
Table 2. River Lagan salmon stocking and returns
|
Year |
Number of fry stocked |
Total smolts grown on from planted and wild spawned fry |
Release of farm reared smolts |
Estimated return of adult fish to the river |
|
1991 |
29000 |
0 |
0 |
0 |
|
1992 |
29000 |
2000 |
0 |
0 |
|
1993 |
50000 |
4000 |
1700 |
200 |
|
1994 |
119000 |
4310 |
2190 |
100 |
|
1995 |
50000 |
6724 |
7276 |
350 |
|
1996 |
None |
4009 |
7291 |
180 |
|
1997 |
29400 |
No data |
600 |
400 |
|
1998 |
40000 |
3000*1 |
3000 |
100 |
|
1999 |
40000 |
8300*2 |
3600 |
50 |
|
2000 |
35000 |
4500 |
4000 |
800 |
|
2001 |
20000 |
1500*3 |
3500 |
450 |
|
2002 |
50000 |
3500*3 |
4000 |
*1 Includes an estimated 1000 2 year old wild smolts derived from natural spawning, the first significant proven wild production since stocking experiments began in 1991.
*2 Includes an estimated 3000 2 year old wild smolts derived from natural spawning, in Collin river
*3 Smolt recapture data incomplete due to open floodgates - based on short range recapture only over part of the run – considered to be an absolute minimum figure and probably a significant underestimate
Fig. 5 River Lagan adult salmon runs
Phase 3 – towards a self sustaining population – 2001 onward
The current emphasis of the project is on establishing a natural spawning population. Each year, up to 100 sites are targeted in a late summer electrofishing survey for naturally spawned fry in unstocked areas. Survival of stocked fry is also monitored using the same method of a 5 minute semi-quantitive backpack electrofishing technique. Fig. 7 shows the areas where wild spawned fry have been detected from 1994 onward. The numbers of wild salmon fry and numbers of sites with wild fry are shown in fig. 8.
Fig. 7. Sites with naturally spawned salmon fry since re-introduction
Fig. 8. Numbers of naturally spawned salmon fry in surveys, 1994-2001
From 2001, the tenth year from first stocking, stocking policy has changed to the planting of 40-50,000 fry in small batches of maximum 10,000 each, in the highest reach available of small tributaries. This places fry in the best areas for water quality and mitigates against the loss of all stocked fish to a single pollution incident. The stocking practice also leaves most areas where significant natural spawning has developed free of introduced fish, and uses supplementary stocking to create a small artificial population “programmed” to return to the maximum possible extent of the river system, able to use all available spawning sites.
Fish passes of Denil or Pool and overfall types have been installed on 6 of the most obstructive weirs in the lower river, in partnership with the Northern Ireland Rivers Agency, the Inland fisheries section of the Department of Culture, Arts and Leisure, and with local councils and angling clubs. Salmon have now been proven , from the distribution of naturally spawned fry, to have access to almost the entire river system.
Where opportunities present themselves, partnerships are forged with organisations making alterations to the river to facilitate habitat improvement, and this is where further work on restoring the population will have to focus in future to achieve a fully self sustaining stock. Examples include work with angling clubs and with Rivers Agency who are sometimes able to include some restoration work in with maintenance of drainage channels. Such co-operation has seen the restoration of a spawning ford by means of scouring downstream of a horseshoe groyne placed in the river to improve angling opportunity, at a site subsequently used by spawning salmon.
General discussion – Lessons learned so far
The degree of success, first in generating returns to the river, and secondly in the development of some natural spawning, indicates that most of the assumptions made in interpreting the initial feasibility study, particularly about the extent and quality of the juvenile habitats, were correct. The very simple, rapid habitat survey technique proved to be a good predictor of suitability of areas for hold stocked fry and after natural spawning, a good predictor of where naturally spawned fry could succeed. Some interpretations from the feasibility study data, however, were over-cautious these were:
1 The passability of weir sites – some salmon moved upstream over physical obstacles which at first inspection and by comparison with standard MAFF guidelines appeared to be impassable obstacles. The lesson learned is that weir inspections should be made at several states of flow, from low to spate. Assessments made under particularly low flow conditions did not relate well to fish behaviour or capability under high flows, and vice versa. Installation of fish passes at several sites, allowing upstream passage in all conditions, prevented build-up of fish below weirs.
Migrating smolts appear to be able to cope with lower dissolved oxygen levels than the minima suggested by the literature. Downstream smolt migration must have occurred through river reaches of several km, where DO levels below 6mg/l are routinely recorded during the latter part of the smolt season. Adults moving upstream are able, by migrating on high flows, to avoid the worst areas of the rivers main stem.
Salmon clearly spawned and produced fry, albeit in small and variable numbers on stream beds which bore no resemblance to classical rounded gravel, 30-80mm diameter substrates. The only possible conclusion from this is that the salmon are able to remove some silt and re-grade gravels at the act of spawning, to an extent permitting some egg survival and fry emergence before gravels re-silted in low summer flows.
The feasibility study focussed on some very obvious problems, i.e. those associated with urban and industrial impact on the river, including tidal impoundment, enrichment leading to low dissolved oxygen, and weir passability. The success of the project to date demonstrates that this initial focus was essentially valid – fish passage issues due to industrial and navigation developments were the essential cause of the absence of salmon. However, having now gone a long way to removing these essentially urban problems, the future of salmon in the river relies on tackling a more difficult issue – that of agricultural enrichment, pollution , habitat loss and a general lack of good spawning substrates. These are issues of recent decades, and additional to the problems which initially extirpated the salmon from the river Lagan in the 1700s and early 1800s.
Natural spawning sites have now begun to be used, and it is particularly interesting to note that no natural spawned fry have been found where tributaries have not first been stocked. There would hence seem to be little or no straying even from one tributary to another. The order of first observation of natural spawning follows the order in which tributaries were stocked, with a three year lag, i.e. the shortest possible generation cycle accounting for stocked fry to S1 smolt to 1SW adult to 0+ fry.
The general degree of success encountered in the programme to date indicates the validity of the approach taken toward restoration of the salmon population i.e. taking a fish-centred approach, using trial stocking to find out how bad (or good) the state of the river was, and building upon success when encountered. Had the initial approach adopted following feasibility study been continued, i.e. to seek funding for a major, fully funded, restoration project based on a “habitat first” basis, it is possible that very little progress would have been made. The general public impression of the river Lagan before salmon re-introduction and surrounding publicity was of a polluted, unhealthy river, and this impression influenced opinion to the extent that many were generally not prepared to believe in the possibility of salmon restoration. Once the first adult salmon were seen and reported following trial stocking, other parties beyond the fisheries agencies became keen to take a part, in addressing the need for fish passes, building in habitat works into drainage maintenance, and in considering salmon as a real issue in deliberation over water quality and general river management. It is also clear that there was an element of good fortune in the timing of the first experiments – the first stocked salmon survived well and returned to a tributary which is now under severe stress as a result of silage-effluent fish kills. Had one of these incidents occurred in 1991 the salmon programme might not have been continued, such was the air of scepticism at that time.
The programme to date also demonstrates what can be achieved on a relatively low budget commitment. At no time have more than a total of three scientific staff been engaged part-time on the work. Adult counting systems and smolt counting have been carried out with partial counts and traps, and while this creates a degree of uncertainty over the exact numbers, over time the trends are valid and the partial counts give an indication of relative numbers year on year.
This programme began before much of the current debate over genetic integrity of stocks. However, as there were no salmon present, the issue is not a major one, except to note that some genetic characteristics of the introduced stock – a mainly 1 sea winter “grilsing” stock with a relatively late summer run from sea, appear to be ideally suited to the recipient river.
The future – a truly self-sustaining population
The task of restoring a self-sustaining salmon population to the Lagan system is far from complete – and the issues of water quality and the need for full habitat restoration, somewhat ignored from the outset, in this stocking based experiment, will now have to be addressed. Spawning substrates in particular will need improvement and replacement where nursery areas exist without adjacent spawning areas. Supplemental stocking is likely to be needed for several salmon generations at least. Pressures include the ambitions of some local groups to re-build the former commercial navigation system for recreational purposes, and discussions are on-going as to whether this can be done without the fish passage problems and habitat damage created by the harsh 1700s engineering solutions. Poor water quality, now in the upper, non-urban, agricultural parts of the catchment, is also a major issue. It is hardly surprising that, having eliminated the critical blockages for salmon and begun to restore the population, the problems now faced are the same as for freshwater phase salmon populations throughout much of their range. Nevertheless, the development of some natural spawning, increasing year on year since 1994, indicates that here at least is the possibility of an Atlantic salmon population being brought back from extinction to viability.
