NOTE: This is from the biological report on the status of Atlantic Salmon -
see Table of Contents and News Release for additional information.

 

SECTION 6: DISTRIBUTION AND ABUNDANCE - STATUS OF ATLANTIC SALMON IN THE GULF OF MAINE DPS

6.1 DESCRIPTION OF THE HABITAT WITHIN THE GULF OF MAINE DPS

The Gulf of Maine DPS encompasses all naturally reproducing remnant populations of Atlantic salmon from the Kennebec River downstream of Edwards Dam northward to the mouth of the St. Croix River. As a complex, these rivers are typically small to moderate sized coastal drainages in the LMFP ecoregion. This commonality of zoogeographic classification makes coarse level descriptions of watersheds very similar between the rivers. The watershed structure, available Atlantic salmon habitat, and abundance of Atlantic salmon stocks at various life stages are best known for the seven largest rivers with extant Atlantic salmon populations. The habitat and population ecology of populations in smaller rivers is less well known with the possible exception of Cove Brook (Meister 1962; Baum 1997). This section focuses on the seven core rivers where the most comprehensive and quantitative information is available.

The Dennys River originates in Lake Meddybemps in the town of Meddybemps, Washington County, Maine. The drainage area of the Dennys River is 34,188 ha, and it flows a distance of 32 km to Cobscook Bay. In addition to Lake Meddybemps, Cathance and Little Cathance Lakes are located in the headwaters of the drainage. The confluence of Cathance Stream, a major tributary, is located approximately 1.0 km upstream from tidewater. The upper reach of the river, from Lake Meddybemps to the falls is flat and slow moving. The reach from the falls to Cathance Stream has flat water stretches and a few riffle areas. The estuary is large, has numerous coves and bays, and numerous peninsulas and islands between Dennysville and the ocean (Beland et al. 1982). Lands within the drainage are sparsely populated and managed for the growth and harvest of forest products and lowbush blueberries. Water quality is generally good, but logging throughout the area has resulted in an abundance of woody debris in some reaches of the river.

The East Machias River originates at Pocomoonshine Lake in the towns of Princeton and Alexander in Washington County, Maine. The river has drainage of 65,009 ha that contains 26 lakes and ponds, and over 50 named tributaries. It flows a distance of 59.5 km to Machias Bay. The watershed is sparsely settled and forested with a mix of spruce and fir. Organic materials from wetlands and bordering lakes and ponds discolor the waters of the river. The East Machias and Machias Rivers enter the same estuary and the lower 3.2 km of the estuary is common to both rivers (Dube and Fletcher 1982).

The Machias River drains an area of over 119,140 ha. It originates from the five Machias Lakes and flows 98 km to Machias Bay. The watershed is located in Washington and Hancock Counties and more than 160 tributaries and 25 lakes and ponds exist in the system. A natural gorge at the mouth of the river in the town of Machias may impede the passage of salmon during periods of extreme high flow. The gorge is being studied by the State of Maine to determine if passage can be improved as part of State rehabilitation efforts for Atlantic salmon in that river. The Machias River headwaters are characterized by rolling hills with forested stream valleys and a number of barren areas, with ground cover typically consisting of shrubs. The lower portion of the basin is composed of large forested areas (Fletcher and Meister 1982). The Machias and East Machias Rivers share a common estuary. The estuary is elongate, approximately 9.6 km in length, but relatively narrow.

The Pleasant River watershed in Washington County originates at Pleasant River Lake in Beddington and drains an area of 22,015 ha. It flows 45 km to the head of tide in the town of Columbia Falls. There are few lakes in the watershed, and the tributaries are a network of small feeder streams with a combined length of 109.4 km (Dube and Jordan 1982). The headwaters are composed mostly of hills and ridges, with forests of spruce, fir, and hardwoods. The river water exhibits a high degree of red-brown coloration caused by leaching of roots, leaves, and other organic materials that originate from extensive peat bogs in the drainage. The bogs provide water during dry periods, storage during wet periods, and moderate discharge in the basin (Dube and Jordan 1982). The Narraguagus River originates at Eagle Lake, flows through Washington and Hancock Counties, and drains an area of approximately 60,088 ha. The main stem drops a total of 124 m over a distance of 69 km to the head of tide in Cherryfield. The West Branch of the Narraguagus, a major tributary, has a drainage area of approximately 18,100 ha and reaches the main stem 3.2 km upstream from the head of tide. There are more than 402 km of streams and rivers in the drainage and about 30 lakes and ponds, with three of the lakes exceeding 162 ha in size (Baum and Jordan 1982). The topography of the headwaters consists of rocky hills and ridges, and forests that are primarily a mix of spruce and fir interspersed with hardwoods. There are large blueberry barrens in the watershed, and lands are primarily managed for berry production and forest products.

The Ducktrap River is relatively small compared to other Atlantic salmon rivers in Maine. It originates in Tilden Pond in Belmont Township, Waldo County, has a drainage area of approximately 9,324 ha, and flows for a distance of 10.7 km to Lincolnville where it enters Penobscot Bay. There are four ponds in the drainage and two major tributaries. The two tributaries, Kendall and Black Brooks, enter the main stem in the lower portion of the drainage. The surrounding area is sparsely settled, and former agricultural lands are either overgrown or reverting to early successional growth. The drainage is rugged and hilly, and in the lower portion the riverbanks rise sharply from the stream to heights that exceed 30.5 m (Bryant 1956).

The Sheepscot River originates as a series of hillside springs in West Montville, Waldo County and flows a distance of 54.7 km to the estuary near Alna. The West Branch of the river originates at Branch Pond in Kennebec County, flows a distance of 24 km and enters the main stem in Sheepscot. The Dyer River, the largest of the tributaries, has a length of 27.3 km and flows to the estuary. The Sheepscot River drainage includes 24 lakes and ponds and encompasses an area of 59,052 ha. The upper portion of the Sheepscot River estuary resembles a fjord, whereas the lower portion is typical of Gulf of Maine DPS watersheds, with mud flats and salt marsh covering large areas. Sheepscot Falls, located in the upper estuary, is an area composed of ledge, and the site of a former dam (Meister 1982). Land within the watershed was once intensively farmed, but the majority is now forested. Deposited glacial material provides a source of boulder, rubble, and cobble in the drainage.

6.2 SURVEYS OF AVAILABLE FRESHWATER PRODUCTION HABITAT WITHIN THE GULF OF MAINE DPS

The ASA and USFWS have conducted extensive habitat surveys on several rivers within the DPS (Horton et al. 1998). GIS coverages and habitat estimates are now available for all major watersheds of the DPS, with the exception of some minor tributaries with limited juvenile production habitat. Horton et al. (1998) report that from 1994 to 1997, more than 330 km of Atlantic salmon habitat have been surveyed within the DPS. Habitat maps for the rivers within the DPS are critical tools for population assessments, habitat restoration and conservation, watershed management, and fisheries management (Figures 6.2-1 to 6.2-7). Table 6.2 provides a profile of the rivers within the DPS and is derived from several habitat surveys (Beland 1982; Horton et al. 1998; S. Davis, Maine DIFW, Pers. Comm.; E. Baum, ASC, Pers. Comm.). Table 6.2: Juvenile Atlantic Salmon Production Habitat (units = 100m²) within the Gulf of Maine DPS including estimated habitat within the geograpical range for smaller coastal drainages presently accessible to Atlantic salmon. (a) see section Figure 5.2.2-1 for complete listing of smaller coastal drainages. (b) see section 6.2.1 for definition and calculation of Conservation Spawning Requirements.

The ASA and FWS are presently conducting more thorough habitat surveys of Cove Brook and other smaller coastal drainages to better quantify available habitat in those drainages (Buckley, FWS, pers. Comm.). Additionally, the FWS Gulf of Maine Project is currently undertaking a study on the Narraguagus River to use airborne infrared imagery to identify cool water refugia for Atlantic salmon. A spatially continuous temperature profile can be obtained for the river using this technology. One objective is to determine if cold water inputs are predictable on an annual basis regardless of stream flow and water temperature. The imagery is currently being processed so it is too early to state whether this technology will be useful in the future in the Narraguagus River and other rivers within the DPS (Wright 1998).

Figure 6.2-1: Dennys River Watershed
Figure 6.2-2: East Machias River Watershed
Figure 6.2-3: Machias River Watershed
Figure 6.2-4: Pleasant River Watershed
Figure 6.2-5: Narraguagus River Watershed
Figure 6.2-6: Ducktrap River Watershed
Figure 6.2-7: Sheepscot River Watershed

6.3 DESCRIPTION OF POPULATION ABUNDANCE WITHIN THE GULF OF MAINE DPS

Species abundance in a DPS is a critical concern in assessing the population status of a species under the ESA. While the Services have no quantitative guidance for determining if a species is endangered or threatened, an examination of current abundance compared to historical levels and analysis of recent trends was used to determine the status of Atlantic salmon within the Gulf of Maine DPS. As with most anadromous species, Atlantic salmon frequently exhibit large temporal changes in abundance. While the high level of variation that these populations exhibit makes quantitative assessments of changes in abundance difficult, trends in these indices are evident. The relative abundance of Atlantic salmon in several U.S. streams was examined using data from the ASA and the U.S. Atlantic Salmon Assessment Committee (USASAC 1999). Information on adult returns, redd counts, and juvenile population abundance is presented in this section as well as information on broodstock collections and subsequent stocking.

6.3.1 ADULT ABUNDANCE

Documented returns of adult Atlantic salmon to the DPS rivers surveyed remain low relative to conservation escapement goals (USASAC 1999). In considering the numbers reported for the seven wild river populations, one must be cognizant that these data represent minimal counts and not total numbers owing to counting inefficiency and observational bias (USASAC 1999). This is true whether considering adult numbers or redd counts of spawning adults. These numbers can vary greatly year to year in the same river due to factors not related to actual abundance, and probably always underestimate the actual numbers present. In some years, there was no effort to collect count data for several rivers, even though some adults probably did return to those rivers. The accuracy of returning adult counts is a function of the extent and nature of sampling effort (trapping, observation, etc.) on each of the rivers (USASAC 1999; Beland and Dubé 1999). The best and most comparable year-to-year data available are on the Narraguagus River because of a long-term intensive study. Other river data on adult counts are the result of vastly differing sampling efforts year-to-year and river-to-river within a year. Direct comparisons of adult counts of one year to another year are subject to the potential for significant and unmeasured error.

For an individual river, most past counts of returning adult wild salmon are of limited value in assessing the population trends because of the inconsistent methods and discontinuous nature of gathering adult return counting data. The exception is the Narraguagus River, where adult counting effort and technique have been relatively consistent over recent years (Beland and Dubé 1999). The Narraguagus River adult return data are useful for showing that these wild populations have been low this past decade relative to observations of earlier decades and reasonable expectations of habitat carrying capacity. Given the limitations described, absolute values of adult estimates in a single year cannot be used to assess the immediate condition of a stock in a river or to determine if listing is warranted. However, when individual river data are combined in aggregate they provide a relatively robust indication of population trends within the Gulf of Maine DPS. The development of fixed weirs and other improvements in population assessment included in the Atlantic Salmon Conservation Plan for Seven Maine Rivers will, by the year 2001, upgrade the effectiveness of monitoring adult returns to more accurately index the true state of the wild populations. This will also permit a more precise evaluation of short-term fluctuations in a population. It is important that this increased precision be achieved by that time, because 2001 is when the first significant impact of fry stocking on adult returns should be realized in the Gulf of Maine DPS.

Atlantic salmon documented to have returned to rivers within the DPS through angler catch and trap data from 1970 to 1998 provide the best available composite index of recent adult population trends (Table 6.3.1-1; Figure 6.3.1-1). These indices indicate that there was a dramatic decline in the mid-1980s and populations have remained at low levels since that point.

Figure 6.3.1-1 and Table 6.3.1-1: Total Documented Natural (Wild & Fry Stocked) Spawner Returns from USASAC (1999) data (minimal* indices) and escapement goal (e-goal) for each river in the Gulf of Maine DPS.

YEAR	DENNYS	E. MACHIAS	MACHIAS	PLEASANT	NARRAGUAGUS	DUCKTRAP	SHEEPSCOT	TOTAL
1970	49	1	226	1	132	-	6	415
1971	19	5	147	1	73	-	30	275
1972	61	3	191	1	244	-	20	520
1973	40	5	28	2	142	-	20	237
1974	49	1	26	30	137	-	20	263
1975	40	22	41	8	109	-	11	231
1976	20	0	18	1	28	-	10	77
1977	26	20	15	3	117	-	24	205
1978	38	46	90	16	98	-	35	323
1979	38	18	58	8	49	-	8	179
1980	73	38	65	5	115	-	30	326
1981	46	29	34	23	51	-	15	198
1982	20	22	55	7	67	-	15	186
1983	28	5	17	38	71	-	12	171
1984	68	38	25	17	58	-	22	228
1985	14	30	27	31	57	15	6	180
1986	8	8	28	19	25	15	11	114
1987	1	6	4	5	26	0	6	48
1988	6	5	8	-	27	0	0	46
1989	1	9	9	0	27	0	3	49
1990	11	17	1	0	28	3	0	60
1991	6	2	1	0	68	0	0	77
1992	6	0	0	0	47	0	4	57
1993	7	0	13	0	74	0	0	94
1994	6	-	-	1	50	-	15	71
1995	5	-	-	-	56	-	22	83
1996	10	-	-	-	56	-	8	74
1997	0	-	-	1	35	-	0	35
1998	1	-	-	-	22	-	-	23
Total	697	330	1127	216	2089	33	353	4845
Average	24	14	47	9	72	4	13	167
E-goal	139	145	463	72	385	39	111

* These are considered minimal estimates, a "0" means that no fish were documented to have returned a "-" indicates that no quantitative data were collected in that year. However, it is critical to note that as displayed in table 6.4, the presence of redds indicates that adults were present in the rivers during those same years. To compare rivers within the DPS to each other and to the escapement required for adequate egg deposition, a method developed by Elson (1975) and adopted by the North American Salmon Working Group (NASWG) has been used to estimate escapement goals for each watershed in the DPS (USASAC 1999). Since historical Atlantic salmon abundance data are available, this methodology allows for comparing population abundance to habitat potential within the constraints of adult return data mentioned previously. This method assumes a target egg deposition of 2.4 eggs/m² is needed to fully seed a river (Elson 1975). An average female fecundity of 7,200 eggs/female (Baum and Meister 1971; Baum 1997) and a 1:1 male: female ratio (Baum 1997) was used to determine optimal escapement. For example:

With 100,000 m² of accessible habitat, target spawners would be:

100,000m² X 2.4 eggs/m² = 240,000 eggs;

240,000 eggs / 7,200 eggs/female = 33.333 females; and

33.333 x 2 = 66.67 = 67 Atlantic salmon

Once the escapement goal is calculated, a standardized comparison can be made among the rivers of different size. The return was calculated as a percentage of the escapement goal to standardize among rivers and compare run size to optimal escapement. This value was simply the percentage of the abundance index (trap count or extrapolated adult return from redd counts) divided by escapement goal. For example:

 

An escapement goal of 67 spawners and index of 35 spawners:

(35/67) x 100 = 52.23% of escapement goal

Over the past 24 years, the Dennys and Narraguagus Rivers have had the best returns relative to available habitat. However, adult returns in these rivers still average less than 20% of their escapement goal (Table 6.3.1-2 and Figure 6.3.1-2). The Pleasant, Sheepscot, and Machias Rivers averaged between 10% and 12%. However, recent downward trends in abundance put most rivers at less than 10% of their escapement goals. Only the Narraguagus River has exceeded 10% in the past 7 years. While these estimates are based on counts that give a minimal estimate of run strength, the low levels of abundance are disturbing, given the recent declining abundance. It is important to note that during most of these early time series (1970-1985), recreational fisheries were still harvesting adults from these rivers. Thus, these percentages of run size to escapement goals are extremely optimistic because these percentages represent the potential spawning contribution of returning spawners (run size), not actual escapement. Since exploitation during this time series frequently exceeded 50%, spawning escapement through the 1980's was likely significantly less than the escapement goals summarized in Figure 6.3.1-2 and Table 6.3.1-2. However, the contribution of precocious parr to spawning in these rivers is unknown and could serve to increase the effective spawning size of these populations.

 

 

Figure 6.3.1-2 and Table 6.3.1-2: Natural Run Size as a Percentage of Escapement Goal for Available Habitat

 

YEAR	DENNYS	E. MACHIAS	MACHIAS	PLEASANT	NARRAGUAGUS	DUCKTRAP	SHEEPSCOT
1970	35%	1%	49%	1%	34%		5%
1971	14%	3%	32%	1%	19%		27%
1972	44%	2%	41%	1%	63%		18%
1973	29%	3%	6%	3%	37%		18%
1974	35%	1%	6%	42%	36%		18%
1975	29%	15%	9%	11%	28%		10%
1976	14%	0%	4%	1%	7%		9%
1977	19%	14%	3%	4%	30%		22%
1978	27%	32%	19%	22%	25%		32%
1979	27%	12%	13%	11%	13%		7%
1980	52%	26%	14%	7%	30%		27%
1981	33%	20%	7%	32%	13%		14%
1982	14%	15%	12%	10%	17%		14%
1983	20%	3%	4%	53%	18%		11%
1984	49%	26%	5%	24%	15%		20%
1985	10%	21%	6%	43%	15%	38%	5%
1986	6%	6%	6%	26%	6%	38%	10%
1987	1%	4%	1%	7%	7%	0%	5%
1988	4%	3%	2%	0%	7%	0%	0%
1989	1%	6%	2%	0%	7%	0%	3%
1990	8%	12%	0%	0%	7%	8%	0%
1991	4%	1%	0%	0%	18%	0%	0%
1992	4%	0%	0%	0%	12%	0%	4%
1993	5%	0%	3%	0%	19%	0%	0%
1994	4%				13%		14%
1995	4%				15%		20%
1996	7%				15%		7%
1997	0%				9%
1998	1%				6%
Average	25	14	47	9	74	4	13
E-goal	139	145	463	72	385	39	111
Avg/goal	18%	9%	10%	13%	19%	9%	12%

The pre-fishery abundance index of North American salmon stocks that migrate to the Greenland region of the North Atlantic Ocean continues to be low in spite of apparently improving marine habitat conditions as reflected by ocean surface temperature data in the past few years (NASWG 1999). The apparent non-response to improving marine habitat is believed to be due, in part, to generally depressed spawning populations in North American home rivers and resultant low number of juvenile salmon entering the ocean. Without adequate numbers of emigrating smolts, North American populations have not responded to improved marine growth and survival conditions. Based on estimates of the pre-fishery abundance of North American salmon stocks in the West Greenland Sea provided by the ICES, relatively low adult returns should be anticipated in many North American salmon rivers again in 1999 (NASWG 1999).

Fall redd counts are also used to estimate adult returns. These counts are particularly useful for rivers that do not have trapping facilities for returning adults. The accuracy of the counts of redds created by spawning adults varies due to water conditions (visibility, discharge, water temperature, etc.) and the amount of observation effort (Beland and Dubé 1999). Low, clear water conditions in the fall can provide a high level of efficiency in counting redds by both enhancing river accessibility for observers and the visibility of redds; while high, turbid water limits access and visibility and greatly lowers counting efficiency. Such sampling conditions can vary day-to-day and river-to-river. The value of the past redd counts for assessing the wild populations lies more in the trends of changes in their relative values over a period of years, and less in their absolute values of particular years or rivers. Recent increases in redd counts can be attributed to increased coverage of watershed and supplemental broodstock releases (Table 6.3.1-3; Figure 6.3.1-3). Those releases will be discussed later in this section under the heading of stock enhancement programs.

Declining adult Atlantic salmon returns of the last three decades are best characterized by an early period of relatively high fishing mortality that has declined to minor levels while marine habitat suitability declined severely. As marine habitat indices improved in the early 1990's, the ability of the stocks to respond has been hindered by low spawner abundance caused by previous marine mortality factors. The ability and resilience of Atlantic salmon stocks to return to high abundance is strongly related to the abundance of spawners (i.e., Myers and Barrowman 1996). Since 1970, there has not been a substantial period of time where: marine habitat indices were high; fishing mortality was low; and spawner abundance was at conservation targets for any Atlantic salmon stocks in the Gulf of Maine DPS.

 

Figure 6.3.1-3 and Table 6.3.1-3: Redd Counts within the DPS

YEAR	DENNYS	E. MACHIAS	MACHIAS	PLEASANT	NARRAGUAGUS	DUCKTRAP	SHEEPSCOT	TOTAL
1973	160	58	137		97			452
1974	100	31	94		26			251
1975		5	110		80			195
1976	45	8	80		27			160
1977	202	18	298		277			795
1978	540	94	227		4			865
1979	165		128					293
1980	217	31	361		113			722
1981	150	117	162		298			727
1982	117	17	23		67			224
1983	25	16	83		22			146
1984	253	102	289		259	6	80	989
1985	249	89	234		201	18		791
1986	143	91	273		345		145	997
1987	108	130	274		210	61	79	862
1988	112	18	50		100	8	46	334
1989	148	8	129		163	29	20	497
1990	89	2	113		201	37	17	459
1991	81	0	121	44	186	54		486
1992	63	2	83	17	131	18	40	354
1993	25	17	45	22	105	20	33	267
1994	15	19	50	0	57	36	41	218
1995	49		21	8	61	15	2	156
1996	30	41	102	41	161	44	12	431
1997	35	11	59	1	78	2	8	194
1998	32	74	74	2	63	9	2	256

* The % of the drainage surveyed varies by river and year

6.3.2 Juvenile Abundance

The ASA and FWS annually conduct juvenile population surveys utilizing electrofishing at over 100 index sites within the DPS (Horton et al. 1998; Beland and Dubé 1999). Generally speaking, densities of young-of-the-year salmon (0+) and parr (1+ and 2+) remain low relative to potential carrying capacity. These depressed juvenile abundances are a direct result of low adult returns in recent years. The increases reported by the ASA in juvenile populations (0+, 1+ and 2+) in 1997 within the DPS indicate that fry stocking can increase in-river juvenile (parr) populations. Juvenile densities in the unstocked rivers, the Pleasant and the Ducktrap, remained relatively stable compared to the time series of data (Table 6.3.2-1; Figure 6.3.2-1).

 

Table 6.3.2-1 and Figure 6.3.2-1: Average Densities of Young of the Year (0+) And Parr (1+, 2+ Years) per habitat unit from Horton et al. (1998) and Beland and Dubé (1999).

 

 

YEAR	DENNYS		E. MACHIAS		MACHIAS		PLEASANT		NARRAGUAGUS		DUCKTRAP		SHEEPSCOT		AVG DENSITY
	YOY	PARR	YOY	PARR	YOY	PARR	YOY	PARR	YOY	PARR	YOY	PARR	YOY	PARR	YOY	PARR
1990	3.6	3.4	18.9	6.5	7.1	4.4	7	4	1.1	1.4	15.1	8.2	3.5	1.6	8.0	4.2
1991	3.3	2.8	12.8	4.7	14.1	3.7	14.9	4	3.4	1.5			3.3	2.5	8.6	3.2
1992	2.4	2.2			6.7	3.7			4.4	1.9		10.8	2.8	8.3	4.1	5.4
1993	1.9	1.5	10.9	5.6	10.4	5.3	2.6	1	1.8	3.1	27.7	8.2	0.1	1.1	7.9	3.7
1994	1.5	2.4	7.2	6.3	3.8	2.6	0.5	2.4	2.5	1.2	2.1	1.7	0	0.5	2.5	2.4
1995	1.1	1.2	3.4	3.5	3.9	1.2	1.1	3	3.4	1.7	28	2.4	1.1	0	6.0	1.9
1996	0.9	0.5	18.7	2.1	3.9	2.3	3	1.1	4.7	2.7	8.9	6.9	1.3	0.9	5.9	2.4
1997	5.5	1.9	8.7	7.3	6.5	3.4	9.1	3.7	4.3	3.8	11.1	5.2	1	2.4	6.6	4.0
AVERAGE	2.5	2.0	11.5	5.1	7.1	3.3	5.5	2.7	3.2	2.2	15.5	6.2	1.6	2.2	6.2	3.4

A total parr population estimate is not available for the entire DPS; however, the ASA and NMFS have conducted a drainage-wide parr population study on the Narraguagus River since 1991. Electrofishing population estimates at select index sites are extrapolated to a basinwide estimate using a habitat based stratification design (Beland and Dubé 1999). The numbers presented in the table below are obtained by electrofishing up to 45 sites annually. The parr population in the Narraguagus River has increased in recent years. This increase is most likely attributed to the river-specific stocking program including both fry stocking and adult broodstock releases (Beland and Dubé 1999).

 

Table 6.3.2-2: Drainage-wide parr population estimates for the Narraguagus River

 

YEAR	Large Parr (age >1 + 95% CL)
1991	15,863 " 1,687
1992		14,915 " 1,815
1993		22,901 " 6,916
1994		9,536 " 660
1995		12,737 " 2,962
1996		11,073 " 1,196
1997		26,775 " 4,016
1998		25,382 " 2,832

The 1997 parr population estimate in the Narraguagus River was the highest estimate in the time series of data. In 1997, the basin-wide population estimate of large parr in the Narraguagus was 26,682, an increase of 113% from the 1996 estimate (Beland and Dubé 1999). The drainage-wide population of age 1+ and older parr on the Narraguagus River in 1998 was approximately 25,382, a 5% decrease from the 1997 high (USASAC 1999).

6.3.3 Smolt Production and Outmigration

The NMFS and the ASA have been conducting a study on the Narraguagus River monitoring outmigration of smolts by documenting timing of migration, survival, length, weight and number of smolts (Kocik et al. 1998a). This study was initiated to obtain smolt counts on one of the DPS rivers to better partition mortality between freshwater and marine ecosystems. Starting in 1997, four rotary screw fish traps were deployed from mid-April to early June to sample emigrating smolts. The locations of the traps are downstream of approximately 85% of the juvenile rearing habitat in the basin. In 1998, a total of 974 smolts were captured and the emigrant smolt population on the Narraguagus River was estimated at 2,925 ( 273, which was not significantly higher than the 1997 estimate of 2,871 ( 539. Smolts trapped in 1998 were significantly shorter and lighter than those observed in 1996 or 1997. Average overwinter survival from large parr to smolt was significantly lower in 1998 (11%) compared to 1997 (24%). There is a 99% probability that overwinter survival from large parr to smolt was less than 30%, the minimum estimate cited in previous studies. Survival estimates in both years are substantially lower than estimates previously reported in scientific literature and previously accepted estimates for this region (Bley 1987; Bley and Moring 1988; Baum 1997; Kocik et al. 1999). These substantially lower survival rates could be negatively impacting population recovery. It is unknown whether these overwinter survival rates are typical for the Narraguagus River on a long-term basis or if they are comparable to other rivers in the Gulf of Maine DPS. The NMFS and ASA expanded smolt trapping to the Pleasant River in 1999 and plan additional coverage in 2000 to answer some of these questions. In addition, smolt traps will be deployed in autumn 1999 in an attempt to investigate the possibility of a fall smolt outmigration which, if present, would lead to an overestimation of mortality. Additional years of study will enhance this database, allow testing of environmental correlates to overwinter survival, and facilitate analysis of a potential overwinter production threshold. If suspect relationships are found, probable causes of mortality can be investigated and studies can be conducted to identify possible habitat rehabilitation or enhancement that could increase survival through the smolt stage. McCarl and Retting (1983) have illustrated the importance of evaluating smolt production as it relates to adult returns to better understand population dynamics and recovery of salmonine stocks.

In addition to using traps to estimate total smolt production in the Narraguagus River, researchers implanted ultrasonic pingers into wild smolts and hatchery smolts to investigate their behavior and fate as they moved downriver and into the estuary in 1997 and 1998 (Kocik et al. 1998). Preliminary data on detection percentages between transects averaged 90% in the riverine section, 91% in the estuary, and dropped to 83% at the marine array, suggesting a zone of increased mortality in Narraguagus Bay. The median transit time for wild smolts from the first detection unit to the marine array (21 km) was 84 hours, yielding a median speed of 0.3 km/h, slower than the speed observed in 1997 (0.5 km/h). Slower outmigration in 1998 may be due to lower stream flows and warmer temperatures (Kocik et al. 1998b). Preliminary estimates of survival indicate that roughly 50% of smolts are presently emigrating from the outer waters of Narraguagus Bay and entering the Gulf of Maine.

NMFS and ASA researchers illustrated that between the 1997 and 1998 smolt runs, a 126% increase in large parr production resulted in less than a 2% increase in smolt production. Additionally, these researchers found that approximately half of these emigrating smolts do not reach the Gulf of Maine. These preliminary data led the BRT to conclude that low overwinter and emigration survival rates may be impeding the recovery of these populations and are an issue of concern.

6.4 Conservation Hatchery Programs

Atlantic salmon stocking in rivers of the Gulf of Maine DPS prior to 1991 is discussed in section 4.2. These programs were conducted primarily using stocks from the Gulf of Maine DPS and neighboring river systems and stocking occurred at relatively low levels. The river-specific stocking program for Atlantic salmon in the Gulf of Maine DPS was initiated in 1991 when the USFWS designated these river populations as Category 2 candidate species under the ESA. Craig Brook National Fish Hatchery (CBNFH) was converted from a single stock smolt production facility to a multiple-stock fry production facility. Broodstock collections began in 1991 and initially focused on the collection of sea run returning adults. However, due to insufficient adult numbers, parr began to be collected in 1992 (Table 6.4-1)(Buckley 1999). The majority of the collections have been parr that were subsequently raised to maturity at CBNFH. Mating is conducted according to protocols developed and adopted by the Maine Technical Advisory Committee (Beland et al.1997; Copeland et al. 1998). Broodstock for the Dennys, East Machias, Machias, Narraguagus and Sheepscot Rivers are held at CBNFH. These collections have increased the effective population size of these populations (wild and captive) and provide a buffer against extinction for these populations. Parr were collected from the Pleasant River and were transferred to the North Attleboro National Fish Hatchery. These fish were later destroyed due to the presence of a newly discovered Atlantic salmon disease-SSSV (Section 7.3).

 

Table 6.4-1: River-Specific Broodstock Collections

YEAR	DENNYS		E. MACHIAS		MACHIAS		PLEASANT		NARRAGUAGUS		SHEEPSCOT		TOTAL
	Adult	Parr	Adult	Parr	Adult	Parr	Adult	Parr	Adult	Parr	Adult	Parr	Adult	Parr
1991					11								11
1992	6	249				414				232			2	895
1993	6	182		239	11	280				174		87	17	962
1994	4	151		166		313				165		84	4	879
1995		234		145		375		200		361	20	107	20	1422
1996				132		238		81		361	8	87	8	899
1997		150		125		250		-		250		150		925
1998		150		125		250		-		250		150		925

The focus of the river specific program is to produce fry that are then stocked back to the river of their parent's origin. In 1992, fry stocking began with the release of less than 14,000 fry in the Machias River. It was only in 1997 that stocking reached levels where most of the suitable and unutilized habitat is fully stocked at a target density of 100 fry per habitat unit (100 m²) in the five target rivers (Copeland1998). Egg take and subsequent fry stocking has increased significantly as broodstock numbers increased at CBNFH (Copeland 1998)(Table 6.4-2). Fry stocking occurs in May after most or all of the yolk sac has been absorbed and the fish are ready to begin actively feeding (Copeland et al. 1998). Each year the ASA makes a recommendation to the TAC regarding fry stocking. The TAC then reviews the recommendation and forwards a final recommendation to the ASA, USFWS and NMFS. In 1999, ASA staff provided the TAC with a detailed rationale for their fry stocking recommendations. Primary considerations for selecting river reaches for fry stocking include habitat quality, avoiding direct competition between stocked fry and emerging wild fry, and finally, logistics (ASA Rationale for Fry Stocking Recommendation, ASA staff, 2/1/99). River specific fry releases are displayed in the table below (Table 6.4-3). Parr were collected from the Pleasant River but were not stocked later back into the Pleasant River due to the presence of a disease. This will be discussed later in Section 7.3.

The response of Atlantic salmon populations to supplemental stocking programs can be partially evaluated based on juvenile production but adult returns are the ultimate evaluation stage. It takes about 4 years from initial stocking to evaluate population level responses since there is a lag between removal of parr for brood stock development, the subsequent stocking of their offspring, juvenile assessments, and adult returns. The first opportunities to make a comprehensive evaluation will be when adults of fry-stocked origin (as 2 SW fish) potentially contribute to the 1999 spawning run that ends in October. The 1999 returns are from the moderately high fry stocking levels of 1995 for the Dennys, Machias, and Narraguagus Rivers. It will not be until 2000 that fry-stocked fish will contribute a potentially substantial element to all five rivers with river specific stocking programs in them.

Table 6.4-2: Egg Production at CBNFH (Copeland et al. 1998; Copeland, Pers. Comm.)

YEAR	DENNYS	E. MACHIAS	MACHIAS	NARRAGUAGUS	SHEEPSCOT
1991			13,789
1992	32,700
1993	23,572		47,119
1994	109,625		157,476	114,472
1995	171,797	111,922	332,228	235,660	98,029
1996	231,630	137,961	285,000	297,146	126,362
1997	494,000	394,000	602,600	516,800	375,800
1998	443,200	362,300	547,600	490,000	524,800

 
 
 
Table 6.4-3: River specific fry releases (Copeland et al. 1998; Copeland, Pers. Comm)

YEAR	DENNYS	E.MACHIAS	MACHIAS	NARRAGUAGUS	SHEEPSCOT	TOTAL
1992			13,789			13,789
1993	32,700					32,700
1994	19,963		49,969			69,932
1995	84,000		150,000	105,000		339,000
1996	141,602	114,880	232,812	200,808	102,388	792,490
1997	191,552	112,600	235,999	196,319	63,896	800,366
1998	234,000	190,000	300,000	274,000	256,000	1,254,000
1999*	173,000	210,000	169,000	156,000	302,000	1,010,000

*provisional data

Due to space constraints, there is a need to annually remove a portion of the broodstock held at CBNFH. These fish were spawned the previous year in the hatchery and are released prior to the spawning season. Experimentation on the Narraguagus River has verified that these fish do spawn after being released to the wild and that the fry survive to the parr stage. Broodstock for recent releases are provided in the table below.

Table 6.4-4: River specific surplus broodstock releases (Copeland et al. 1998)

YEAR	DENNYS	E. MACHIAS	MACHIAS	NARRAGUAGUS	SHEEPSCOT
1996	180	0	215	108	0
1997	118	91	231	127	16
1998	126	119	245	222	37

During the development of the Conservation Plan, the Governor's Task Force voted to transfer eggs from the CBNFH to private hatcheries operated by commercial growers. The transfer was made after the aquaculture industry offered to assist in the recovery of the DPS by raising smolts and/or adults as a supplement to fry releases. A total of 3,000 eggs from three strains were transferred from the CBNFH to the aquaculture industry in 1996, 1997, and 1998. Smolts of the following strains were placed into sea cages in the spring of 1999: Narraguagus River, Machias River, Dennys River and Sheepscot River. Not all of these smolts will be moved to cages, as the adults produced would be far in excess of what would be biologically appropriate for use in the river. The additional smolts could be released in the river. Most hatcheries produce smolts in one year, but within a year class or cohort many fish remain as parr and smoltify in the second year. Thus, parr are a "by-product" of smolt production and are available for stocking into the river. To date, age 0+ fall parr were stocked into the Narraguagus, Machias, Dennys, and Sheepscot Rivers.

Five river-specific stocks (Narraguagus, Machias, East Machias, Dennys and Sheepscot) are currently being reared in sea cages and 2SW adults will be available the next two years. Estimated production of adult fish will be 900 for the Narraguagus (available in 1999); and 1,800 each for the Machias, East Machias, and Dennys Rivers (available in 2000). These numbers are far in excess of any realistic biological needs that would be prudent for experimental release of adults into their rivers of origin. Their numbers are excessive because it is necessary to produce fish at these levels to acclimate them to feed in the sea cages, to make optimal use of cage rearing space, and as insurance against catastrophic losses (e.g. loss of 1,000 Narraguagus River fish last winter to seal predation). The NMFS, USFWS and ASA will tag these fish so that release options can be evaluated. The TAC has advised that the release of adults in rivers with limited adult assessment capabilities should be restricted since little is known of impacts (positive or negative) upon ongoing restoration efforts.

6.5 Determination of Population Status

Given the data reviewed in this section, the BRT concludes that naturally producing Atlantic salmon populations in the Gulf of Maine DPS are at extremely low levels of abundance. This conclusion is based principally on the facts that spawner abundance is below 10% of the number required to maximize juvenile production, juvenile abundance indices are lower than historical counts, and smolt production is less than a third of estimated capacity.

Adult counts and redd counts in all rivers continue to show a downward trend from these low abundance levels. Given recent estimates of spawner-recruitment dynamics some researchers suggest that adult populations may not be able to replace themselves and populations would be expected to decline further (Beland and Friedland 1997). Preliminary evaluations indicate that fry stocking is enhancing juvenile production in these rivers and utilization of available nursery habitat has increased. While hatchery supplementation is an important demographic and genetic conservation tool for these stocks, the evaluation of the status of these populations need to be based on the population trends of wild stocks. Because the present hatchery program utilizes primarily fry stocking and no effective non-lethal fry mark has been developed, the BRT could not assess only the wild component of juvenile or adult populations. However, given that the overall status of these stocks is so poor that the BRT concludes that the wild element of combined natural (wild and fry stocked) Atlantic salmon presmolts are at precariously low levels of abundance. Additionally, data from Kocik et al. (1998a) suggest that presmolt overwinter mortality may be substantially greater than values used in previous population modeling exercises (Beland and Friedland 1997). Given this information, the BRT concludes that the abundance of naturally produced Atlantic salmon in the Gulf of Maine DPS is continuing the downward trend in abundance that began in the late 1980's and is characteristic of the entire North American stock complex (NASWG 1999; USASAC 1999).

The demographic and genetic consequences of these low abundance levels coupled with declining abundance trends leads the BRT to conclude that the conservation status of the population segment in relation to ESA listing standards is in danger of extinction.

 

---

Previous section: SECTION 5: CONSIDERATION AS A "SPECIES" UNDER THE ESA
Next section: SECTION 7: ANALYSIS OF LISTING FACTORS
Biological report on Atlantic Salmon

---