Technical, Logistical, and Economic Considerations for the Development and Implementation of a Scottish Salmon Counter Network: Scottish Marine and Freshwater Science Vol 7 No 2

2.0 Technical Considerations and Capital Costs

2.1 Counter Technologies

There is a diversity of fish counter technologies and manufacturers (Fewings 1998), each with their own set of technical advantages and limitations (Eatherley et al. 2005). Common counter technologies used to enumerate salmonids in North America and the UK include multibeam sonar, resistivity, and optical beam counters. Less commonly used technologies also exist, including video and splitbeam sonar (see Section 2.2 for descriptions). Selecting the appropriate technology can pose challenges without the proper information to evaluate trade-offs between the strengths, limitations and costs of each technology. This section provides an in depth review of the most common fish counter technologies, along with their advantages and limitations. First we briefly review how each counter technology works, and their optimal working conditions. Each technology is also summarized in Section 2.2 and references from our literature review are provided in Appendix 1. All of the information discussed was used to inform the decision and cost model in Section 5 but is also useful for evaluating the appropriate counter application on its own.

2.1.1 Hydroacoustic Counters

Multibeam Sonar

Multibeam, or imaging sonars, function by emitting numerous small acoustic beams (i.e., sound pulses) at a fixed frequency from a transducer and converting the returning echoes into high quality images. Software applications interpret these images to generate a video-like image that can be played back using various types of software. Two primary manufacturers, Teledyne BlueView and Sound Metrics Corp., supply the sonar transducers and their associated software. Each manufacturer offers different models that range in cost and technical specifications. Through literature research and conversations with fisheries researchers that operate sonar equipment, we found that the most widely used and effective models are the DIDSON 300 produced by Sound Metrics Corp. and the M900 series from Teledyne BlueView. From here on we refer to these two models when we provide any specifications or details. The cost of these highly specialized units is generally high (i.e., between £20 000 and £52 000). Table 2.1 shows the approximate equipment costs for multibeam sonar. The costs presented in the tables were prices quoted to IFR by the manufacturer in US dollars. These prices were converted to British pounds using the conversion rate at the time of writing this document. A significant advantage of this type of technology is that limited mounting and diversion structures are required to operate the counter. The minimum equipment investment for this type of technology includes storage for all the equipment, the sonar transducer and mounting device, a data logger (i.e., personal computer along with operational software and hard drives), and onsite power. Additional accessories such as sonar lenses and lighting boards are recommended. Mounting the sonar transducer must be tailored to each site and can be designed and fabricated locally. The transducer mount can be fixed to a pre-existing structure (Baumgartner et al. 2006), or it can be free standing, such as an adjustable pole mount ( Figure 2.1) or a modified stepladder ( Figure 2.2) (see Enzenhofer and Cronkite 2005).

Figure 2.1. Sonar unit mounted on the bottom of a tripod at Spius Creek, held to the substrate by sandbags. This installation allows adjustment with a winch above water surface. Photo courtesy of H. Olynyk - DFO.

Figure 2.2. A simple stepladder mount for a DIDSON sensor. Photo courtesy of H. Enzenhofer.Table 2.1. Equipment costs for all major types of counters. List may not be exhaustive, nor may all components be required at all site locations. Cost ranges encompass the base cost through to the highest cost likely to be incurred.

Table 2.1. Equipment costs#for#all major types of counters. List may not be exhaustive, nor may all components be required at all site locations. Cost ranges encompass the base cost through to the highest cost likely to be incurred.

Technology	Counter	Sensor	Structure	Mounting bracket	Power supply	Computer	Data storage media	Equipment storage on site	Video validation	Remote access software
Multibeam sonar	£20 000 - £60 000	NA	Fence: £1000 - £50 000	£5000 - £8 000	£500 - £2000	£400 - £000	£50 - £500	£200 - £5000	NA	NA
Resistivity counters	£10 000 - £20 000	Crump weir: £20 000 - £750 000	Fence: £1000 - £50 000	NA	£500 - £2000	NA	NA	£200 - £5000	£2000 - £10 000	£400 - £1500
		Flat pad: £300/m
		Tube: £1000	Fish pass insert: £7 000 - £10 000
		Flume: £4000
Optical beam counters	£20 000 - £30 000 a	NA b	Fence: £1000 - £50 000	NA	£500 - £2000	NA	NA	£200 - £5000	NA	NA
			Fish pass insert: £7 000 - £10 000
Video counters	£1000 - £20 000	NA	Fence: £1000 - £50 000	£500 - £8 000	£500 - £2000	£400 - £1000	£50 - £500	£200 - £5000	NA	NA
Splitbeam sonar	£35 000 - £70 000	NA	Fence: £1000 - £20 000	£1000 - £5000	£500 - £2000	£400 - £1500	£50 - £500	£200 - £5 000	NA	NA

^aCost of a video validation box for optical beam counters have been included in the counter costs.
^bSensor costs for optical beam counters have been included in the counter costs locations.

Due to the nature of multibeam sonar, a blind spot is present directly in front of the transducer; the size of this blind spot is dependent on the type of sonar being used. Blind spots can range from a minimum of 0.83 m in DIDSON 300 or 2.0 m in M900 series. As all fish must pass through the sonar beam to be detected, a small diversion fence is often required behind the mount and beyond the blind spot on the downstream side of the equipment to prevent fish from passing behind the equipment or within the blind spot. The range of operation for each sonar device is different and is a function of the operation frequency. Single transducers have been used to enumerate fish in cross-sectional areas ranging from 2 to 35 m (Belcher 2004). In most cases, evaluation of fish passage beyond 20 to 25 m becomes increasingly problematic (see Case Study 1, Section 3.1.2). Due to these limitations, combining sonar counters with partial fences to constrain fish passage, or using two counters each on alternate riverbanks, can improve detection accuracy. Orientation of the sonar beam can also be problematic if water levels fluctuate during operations. These situations may require continuous realignment, which can be time consuming if water height fluctuations are common. In instances of high flow conditions a remote operable pan or tilt mount may be required for alignment. Passing objects through the beam can calibrate settings and check for blind spots. Validation methods to determine counter accuracy are restricted to clear, calm streams with laminar flow where fish passage can be visually assessed by an observer (H. Enzenhofer, pers. comm.). Other disadvantages of sonar counters include: (1) the inability of sonar to identify fish species, and (2) the high cost and portable nature of the equipment making it susceptible to theft, vandalism or loss in flood events. Careful selection of appropriate sites is also required to ensure the substrate or riverbed profile does not cause blind spots in the beam, that fish are not spawning in the beam area, and that fish are encouraged to pass through the beam one time rather than recycling through prior to spawning. In general, sonar counters do not function well in areas with high sound reverberation such as fish passes (Baumgartner 2006, Baumgartner et. al. 2012).

Data collection and storage is undertaken on a local computer hard drive using proprietary software, which requires a user to set up and maintain. Data storage demands are high (0.5 to 2 GB/hour), although current data storage media is relatively inexpensive to purchase. Power demands can also be high (up to 130 Ah/day, Pipal et al. 2010), so connecting to mains power is optimal. Some installations have been operated on large battery banks, solar panels or other types of electrical generators, but this reduces reliability and increases costs and labour.

Data analysis can occur through manual observation and can be reviewed in real time or at faster speeds. Teledyne BlueView and Sound Metrics Corp. offer playback software with the purchase of a counter. Third party software can also be used to automate the data analysis process and provide single fish traces or counts. Sound Metrics proprietary software is also capable of automating the process and producing counts. The Teledyne BlueView software is playback only. Echoview 6.1 is a third party software that can read and analyze hydroacoustic data, and perform a semi-automated process to clean, analyze and detect fish in data files. The efficiency of Echoview's analysis is dependent on data quality and provides time savings compared to manual analysis methods in most cases. Further overview of manufacturer and third party software can be found in Section 3.1 (Existing Methods).

Potential Engineering Requirements

Engineering costs for deployment of sonar counters are generally limited. Specific brackets, mounts or fence designs may be required, but in general these can be generated from existing designs at lower costs unless there is a specific engineer sign off requirement. Structural engineering is rarely required unless the project includes modifications to an engineered structure such as a fish pass, in which case the requirements would be very site specific.

M900 Series by Teledyne BlueView

BlueView Technologies, Inc. is a large company based out of Seattle, Washington, United States that focuses on creating high-resolution underwater acoustic measurement tools for navigation, monitoring, and detection. The company supplies the US Navy, Coast Guard and port authorities. BlueView do not specifically advertise their systems for fisheries management applications but the basic principles are similar to those of other multibeam systems and have been used to enumerate migrating salmonids (Cronkite et al. 2008, D. Ramos-Espinoza, unpublished data). The price for BlueView transducers range from £15 350 to £22 110.

Teledyne BlueView offers a variety of transducer models, however the P900-45 is the model that has been used in fisheries applications (Cronkite et al. 2008, D. Ramos-Espinoza, unpublished data). The P series of sonars have now been replaced by the M series of sonars and are the new generation of transducers from Teledyne BlueView. The technical specifications remain almost identical and thus from here on we refer to the specifications of the M series transducers. The M900-45 unit operates at a single frequency of 900 KHz and has an optimum detection range between 2 and 60 m. The M900-45 uses 256 beams (beam width of 1° horizontal by 20° vertical, spaced 0.18° apart) for a total 45° field of view.

Advantages for the M900 series are similar to those of the DIDSON. The equipment is very adaptable to various conditions and set up and operation is relatively simple and requires minimal training. Playback software is also intuitive and easy to use.

Due to the lower operating frequency of the M900-45, the quality of the video is not as high as the DIDSON and manual enumeration is not as quick and easy. The M900-45 has not been widely used by researchers and thus the level of accuracy for its sizing tool is unknown. Proprietary software does not have fish counting or tracking capabilities and third party software must be used to generate and analyze count or trace data. Through our own work we discovered that Echoview software was capable of analyzing M900-45 data and that data quality is similar to DIDSON on the low frequency setting (0.7 MHz). Teledyne BlueView has recently developed a new dual frequency model (M900-2250) that aims to increase the quality of output data when used at the high frequency (768 beams with a width of 1° horizontal [H] by 20° vertical [V], spaced 0.18° apart). Like the DIDSON at higher frequency, this unit produces higher quality images but has a similar operating range to that of the DIDSON at its high frequency setting (M900 = 8 m, DIDSON = 10 m).

DIDSON by Sound Metrics Corp.

The DIDSON is a multibeam sonar system that was developed by the Applied Physics Laboratory at the University of Washington in 1999. It is now manufactured and sold by Sound Metrics Corporation in Bellevue, Washington, United States, a large company specializing in hydroacoustic equipment for commercial applications.

Sound Metrics' sonars range from £48 030 to £51 236 depending on the unit in question. Data can be collected between 4-21 frames per second providing near video quality imaging once it is stitched together by the equipment's proprietary software. The DIDSON can operate at two frequencies which determine the detection range and video resolution. At high resolution (1.8 MHz), the DIDSON can be deployed between 0.42 to 26.1 m away from the target region, and detect fish in a target zone up to 10 m wide. At low resolution (0.7 MHz), it can be deployed between 0.83 to 52.3 m away from the target, and detect fish in a target zone up to 40 m wide (see Case Study 1, Section 3.1.2). Additional proprietary lenses can be used to enhance image quality. The high frequency mode uses 96 beams (beam width of 0.3° horizontal [H] by 14° vertical [V], spaced 0.3° apart) for a total field of view of 29° horizontal by 14° vertical; the low-frequency mode uses 48 beams (beam width of 0.4° H by 14° V, spaced 0.6° apart) for a total view of 29° horizontal by 14° vertical.

A major advantage of DIDSON counters is the high quality images produced by the small angle and spacing of the beams. In addition the DIDSON hardware and software are relatively intuitive to use and, the training involved for operating the equipment is minimal (see Section 3.1.1 for a review of DIDSON software). This equipment is very adaptable; applications of DIDSON can be used in different environments with the use of proprietary or third party accessories. At high resolution, manual enumeration of fish can be performed accurately, as image quality is high, although, sufficient time should be allocated to achieve high enumeration accuracy.

The DIDSON can be used to measure the length of individual fish but accuracy is variable depending on site conditions. Burwen et al. (2010) found that the length of manually measured fish is highly correlated to lengths measured by the DIDSON (R ² = 0.90, Root mean square error = 5.76 m). It should be noted this study used the DIDSON LR, a model that was not advertised on the Sound Metrics website. Burwen et al. (2007) compared length estimates of tethered and free-swimming fish using a DIDSON 300. The study found that tethered fish were subject to positive bias that increased with range from the transducer (1.3 cm/m of range). Measurements of free-swimming fish had a slight positive bias for individuals under 68 cm and slight negative bias for those greater than 68 cm. One disadvantage with the DIDSON at low resolution is the probability of fish detection declines as distance from the DIDSON sonar increases (Burwen et al. 2010). In our experience, we found fish swimming beyond 18 m from the sonar head may have a lower probability of detection and length inaccuracies. Data generated in Case Study 1 suggested objects beyond 18 m from the sonar head may generate signal strengths below 80 dB; these objects were either undetected by the DIDSON counter or impossible to detect by both manual analysis and Echoview's software analysis (see Case Study 1, Section 3.1.2). Another major disadvantage with sonar technology is the large amount of data generated. The high resolution setting collects approximately 1 GB of data per hour. This means data typically needs to be uploaded daily unless a large databank is present onsite.

Splitbeam Sonar

Splitbeam sonars or echosounders have been used periodically in fish passage assessments (i.e., River Spey, Brotherston, C.E 2002). The principle behind all echosounders is that a pulse of high voltage energy is transmitted from the echosounder control unit to a transducer which in turn changes this to a sound pressure wave. This wave radiates spherically from the transducer into the water. If the wave impacts an object in the water with a different density than the water surrounding it, some of the wave energy is bounced back to the transducer. The transducer converts the returning pressure waves into electrical energy, which are magnified and filtered by the echosounder receiver. The receiver then provides an output signal, which can then be measured and assessed for additional information. Table 2.1 shows the approximate equipment costs for splitbeam sonar.

Two manufacturers, BioSonics and Hydroacoustic Technologies Inc. (HTI), market equipment that can be used for such applications. Equipment costs for a basic splitbeam echosounder and transducer range from £30 000 to £54 000 (as quoted by BioSonics Inc.). The sensor unit can be deployed on fixed concrete structures, fish passes, or mounted using a portable setup. Proprietary software and a data storage device are required, and all have to be present during data collection. In addition to the sensor equipment and computer, a mounting and deployment rig, power source, and diversion fence are required to operate at a particular location. Typical power consumption for the echosounder, not including a computer, is around 2.4 Ah.

The advantages of using splitbeam sonar are similar to those of multibeam sonars. A permanent structure is not required in the river, the equipment is relatively easy to use (plug and play), comes with its own logging software (potentially analysis software), and can be downloaded remotely. Unlike multibeam sonars, the file sizes are smaller reducing storage media requirements. Typically units require low maintenance although in some rivers such as the River Spey, where river discharge and debris can be substantial, this might not be the case. Limitations include the need for a specific riverbed profile that is gently sloping and triangular in cross-section. A bed profile with small substrate sizes is also required as it provides less "noise" or backscatter and helps to avoid fish passing undetected in the acoustic shadows. Other limitations include, software accuracy in tracking fish versus other debris and the need for additional hardware for remote operation and download (BioSonics DT-X Automated Monitoring System, price included in cost estimate). There is a need for a computer to run concurrently with the echosounder to collect data, thus power requirements are high enough to consider mains power as a requirement. Post processing of data with or without additional software is a time consuming endeavour.

Potential Engineering Requirements

Engineering costs for deployment of hydroacoustic counters are generally low. There may be the need for specific brackets, mounts or fence designs, but in general, these can be generated from existing designs and or fabricators at lower costs unless there is a specific engineer sign off requirement. Structural engineering is rarely required unless the project includes modifications to a fish pass.

2.1.2 Resistivity Counters

The Logie and Mark series resistivity counters are most commonly used in Europe and North America with one additional counter (the Pulsar) used in a few Canadian locations. These counters can be connected to multiple sensors comprised of electrode arrays. Sensor configurations include: Crump weirs, flat pads, boxes and tubes (see section 2.3 for descriptions). Advantages of these counters include low operational costs following installation, minimal annual maintenance and robust data collection (> 90% accuracy; McCubbing and Ignace 2000, Simpson 1978). This technology can also be designed to cover the full river widths which can avoid delays in fish migration caused by fences or other large structures (Aprahamian et al. 1996). The main disadvantage is that the initial installation costs can be high when a Crump weir is used as the sensor structure. Sensors also need to be validated (Dunkley and Shearer, 1982, Forbes et al. 1999) to determine accuracy under a range of environmental conditions (e.g., Bray 1997). Table 2.1 shows the approximate equipment costs for resistivity counters.

Required equipment includes a resistivity counter, electrode sensors, onsite power, lightning board, and onsite equipment storage. Power use is generally low, typically around 0.3 Ah per channel/pad deployed (for Logie 2100C). Resistivity counters also have the benefit in that they can be operated using mains power or battery banks. When the sensor electrodes are submersed in water, they create a resistive transducer (Aprahamian et al.1996). For a given water depth, temperature, and conductivity the resistance measured between a pair of electrodes will be constant. This is called bulk resistance and is independent of water velocity. Changes in water conductivity, and volume alter the measured bulk resistance, which generally decreases with increasing water depth and conductivity. Counter designs must be sure to avoid situations where inter-electrode resistance is low, which is typically the result of low water levels over the electrodes, high conductivity and/or long electrode lengths. Specialized computing equipment with data storage monitors the bulk resistance of electrode pairs and evaluates these changes against firmware algorithms to establish fish passage events. Data are stored internally for later retrieval, usually through an onsite laptop computer or through remote access software.

Potential Engineering Requirements

Engineering costs for deployment of resistivity counters can be limited or substantial if a full river span weir is proposed or requires modification. For the purposes of this report, we provide the costs of installing a resistivity counter in an existing fish pass, as a flat pad with or without a diversion fence and in conjunction with a full river span Crump weir. The latter option will require substantial engineering work including: scope and survey, design, permitting, construction oversight and sign off. Users should also be aware that the passage of objects, such as debris and ice, and changes in water depth over the electrodes as a result of wind or air entrainment, could cause temporary changes in the resistance between the electrodes. Counter design must account for these factors to avoid them being interpreted as fish movements and consequently causing false positive counts.

Logie Counter Series

The Logie 2100C was created in 1996 and is currently being manufactured by Aquantic Ltd. in Scotland. Note that there has not been a new hardware or software release since the early 2000s. The Logie 2100C is operated in conjunction with an electrode sensor to detect the upstream and downstream passage of fish as they pass over the sensors. Electrode sensors are manufactured by a third party and are comprised of three corrosion-resistant metal conductors placed in a parallel alignment to form an open array sensor configuration appropriate for weir, flat pad or closed tube configurations (Figure 2.3).

Figure 2.3. Logie resistivity counter setup with battery bank. Photo courtesy of InStream Fisheries Research.

An extremely important subsidiary function of the Logie fish counter is its capability to regularly measure factors affecting bulk resistance (every 30 minutes) and to automatically adjust the sensitivity of its signal-processing path to compensate for any changes. At sites where moderate to high water conductivity prevails and remains constant, the standard counter is not required to make these adjustments. At sites with low conductivity, these adjustments are made with the aid of more precise conductivity data from the optional environmental card and associated conductivity probe.

The Logie 2100C counter continuously monitors the bulk resistance between pairs of electrodes and from them derives a signal, which defines the instantaneous relative magnitude of one to the other. As a fish swims through the sensor array it displaces water. Because the body mass of the fish is considerably less resistive than the volume of water, which it displaces, this passage causes a temporary reduction in the resistance measured between an electrode pair. Fish direction is determined by the order in which the pairs of electrodes measure a change in resistance. Thus, the perturbation of the background signal by a fish swimming through the array allows the counter to detect its passage and direction.

Different relationships exist between the resistance a fish creates and its mass when passing over the counter electrodes, at varying bulk resistances. At bulk resistances below 100 ohms, there is little contrast between the resistances created by a 20 cm fish compared to one in excess of 1 m in length. However, as bulk resistance increases above 100 ohms, fish at a constant swim height will show significantly larger resistances as they increase in size. Thus, the counter can establish the direction of fish passage, and in some cases establish relative fish size based on the strength of resistance signals (McCubbing et al. 2000).

Data are collected using Logie's proprietary software (refer to Section 3.1.6). Each change in resistance is logged on a counter buffer file. This record includes the date, time, conductivity (if a conductivity board and probe are present), counter electrode pair (channel), direction of passage (or event) and estimated peak signal size (Table 2.2). The Logie counter can also produce a graphical data output. This is generated for all records if the counter is set up to capture this data, but is not recorded within the counter itself. Instead it is captured on a laptop or data logger through the printer port. Graphics data can be visually inspected using Aquantic's 2100C Graphics Program software for compliance with typical fish events and can be used as a form of pseudo-validation (see Section 4.2). An example of the graphical output is provided in Figure 2.4.

Table 2.2. Example of data output from Logie resistivity counter.

Date	Time	Conductivity	Channel	Direction	Peak signal size
04/29/1999	19:44:24	142	1	U	93
04/29/1999	19:46:32	142	1	U	43
04/29/1999	19:47:17	142	2	U	66

Figure 2.4. Example signal traces from graphics programs operated by the Logie 2100C resistivity fish counter with a flat pad sensor. The top plot shows the trace of a fish moving upstream over the counter and the bottom plot show the trace of a fish moving downstream over the counter.

Mark Counter Series

The Scottish and Southern Energy ( SSE) counter was developed in the late 1940's by Norman Lethlean for what used to be the North of Scotland Hydro-Electric Board. It was first operated at Clunie and Pitlochry salmon ladders in 1951 (R. Gardiner, pers. Comm.). Since its inception, the SSE counter has been subject to a number of upgrades and modifications. Commercial sale of these units has not occurred, so full specifications are not readily available. Use of these counters has been restricted to fish passes with limited passage areas such as the top cell of a Borland lift. It is capable of operating two sensor channels each 1 m wide. The Mark 12 ( MK12) unit is most commonly used in fish passes. MK12 has removed the Wheatstone Bridge electrical circuit to measure changes in resistance and instead monitors the resistance across pairs of electrodes and not the resistance of the body of water above them. Briefly, as an ascending fish enters the downstream counting zone (between downstream and centre electrode) it causes the current flowing between the electrodes to increase. Once the current exceeds the minimum fish current (defined by user) the downstream yellow LED illuminates, indicating that a fish has entered the lower counting zone. As the fish proceeds upstream it moves away from the lower counting zone and enters the upper counting zone (between centre and upstream electrodes) and causes an increase in current between those electrodes. Once the current exceeds the minimum fish current the upstream green LED illuminates, indicating that the fish has entered the upper counting zone. If the fish continues to move upstream and leaves the counting zone within a user defined time period, an upstream count is recorded and both LEDs light up. For a fish moving downstream the above sequence is reversed. The MK12 counter now also has the ability to automatically compensate for changing conductivity. The new MK12 unit uses proprietary software to link image records to environmental and fish signal data, count records, and allow users to set the minimum current before a fish is detected. Linking all the various data together allows users to greatly reduce the amount of time needed for validation. Due to the limited width of each independent sensor, these counters are restricted to fish passage areas less than 1 m wide.

Pulsar Resistivity Counters

Pulsar resistivity fish counters have been used by the government of Canada for many years in the enumeration of sockeye salmon and other species. These counters use the Wheatstone Bridge principle similar to the Logie counter, but only work with a series of small (30 cm width) Perspex tunnels. While a few sites still utilize this older technology on Vancouver Island (Stamp River), most locations have moved to alternate more updated technologies that create digital records of fish passage. No current manufacturer of equipment could be sourced for these counters that historically were produced in British Columbia, Canada.

North West Marine Technologies

North West Marine Technologies ( NWMT) produces an adult fish counter and sensor tube for its utilization. Information on the use of the adult fish counter has been very difficult to source and existing operational sites have not been located.

2.1.3 Optical Beam Counters

Optical beam counters use vertically arranged optical infrared beams to count fish as they pass through the counter ( Figure 2.5). Due to the limited distance light can travel through water, the emitting diodes and receivers must be placed closely together (< 1 m) and thus usually require a fish pass or fence system to divert fish towards sensors. The pattern of disruptions to light beams enable the fish counter to calculate the shape, size, profile and direction of motion of a fish passing through the sensor ( Figure 2.6). The equipment required for this type of installation includes the optical counter, onsite power, and a mounting structure. Additional structures such as a lightning board and onsite equipment storage are preferred. Power consumption is generally low, typically at 0.2 Ah per channel deployed. Advantages to using these counters include low cost operation following installation and robust data. The largest disadvantage is the initial installation costs, particularly if the sensors are to be placed in a new location where an existing diversion weir is not present but is required. Also, weekly maintenance may be required for debris removal. As this type of technology does not cover the full width of a river, it also has the potential to delay fish migration if it is used in conjunction with any type of diversion structures such as fences and weirs. Sensor units should be submerged at all times and air entrainment avoided if large quantities of noise-related events are to be avoided. Limited publications exist on the effectiveness of optical beam counters despite worldwide use. According to the manufacturer, the Riverwatcher can count fish at 98% accuracy and measure size with more than 95% accuracy. A study by Shardlow and Hyatt (2004) achieved > 90% accuracy with careful installation and operation of a Vaki Riverwatcher counter. Values can approach 100% in ideal conditions (low fish abundance and clear water) and decrease as conditions become more difficult (high fish abundance or debris load). Table 2.1 shows the approximate equipment costs for optical beam counters.

Figure 2.5. Virtual illustration of Vaki Riverwatcher's migration corridor. Photo courtesy of Vaki.

Figure 2.6. Virtual illustration of fish migrating through Vaki Riverwatcher. The left illustrates a side profile view of the counter. The right illustrates how a fish disrupts the optical beams. Photo courtesy of Vaki.

Potential Engineering Requirements

Engineering costs for deployment of optical beam counters is variable and mostly contingent on whether a diversion weir is proposed or requires modification. There may be the need for specific weir or fence designs and these would be very site specific. For the purposes of this report, we provide the costs of installing an optical beam counter in an existing fish pass and in conjunction with a full river span fence. Neither are likely to require substantial engineering work.

Vaki Counters

Vaki, based in Iceland, is the only commercial supplier of optical beam counters. Many of their designs are used in the fish farming industry. The Riverwatcher is specifically marketed for migrant fish enumeration and is extensively used in Iceland, Norway and Sweden, as well as many other countries worldwide. This model is an integrated infrared optical beam counter unit which, when combined with Vaki's proprietary monitoring computer and firmware, will generate fish abundance data, date and time of records, digital silhouettes of fish from the infrared lights, evaluation of fish versus non fish records ( QA), species evaluation, direction of movement, and fish length. In addition, the Riverwatcher can be combined with a proprietary video light box to capture still or video images of fish movement ( Figure 2.7). This information is tagged and appended to the counters output for operator review. Data is compiled and analyzed with the provided software and allows users to check individual fish records against images provided from the optical beam counter or, when the conditions are suitable and video is installed, from the video images (Figure 2.8). Refer to Section 3.1.4 for a detailed analysis of the Riverwatcher's proprietary software, Winari.

Figure 2.7. Vaki optical beam counter with double light box video recorder. Photo courtesy of Vaki.

Figure 2.8. Screenshot of Vaki's proprietary software for the Riverwatcher counter, Winari, with corresponding fish shadow from a specific fish event. Photo courtesy of Vaki. 2.1.4 Other Counter Technologies

Video

In general, rivers and streams in Scotland experience periods of high turbidity annually, usually associated with spate events and presumably movement of salmon and trout. For this reason, full river recordings of fish movement are limited. One current pilot study on the River Deveron (equipment supplied by Lamberg Bio Marin) did not have any available data at the time of writing this report. During a site visit by IFR staff in October 2014, the river was high and extremely turbid and the problems of capturing video images under these conditions were self-evident. Table 2.1 shows the approximate equipment costs for video counters.

Video cameras have been extensively used in fish ladder applications in the North West of the US for several decades as a way of verifying manual counts at observation windows ( http://www.fpc.org/adultsalmon_home.html). A more recent evaluation of video only counters was undertaken by the UK Environment Agency (Washburn et al 2008b). In this review, the potential for capturing video images of fish with cameras and digital video recorders was reviewed, as well as the availability of motion-triggered recording and post-processing software. To summarize, video cameras used in fish passes and ladders do have utility on their own when enumerating migrant fish. Cameras were placed at a suitable location in the fish pass where fish do not hold and were typically operated both day and night. Light sources for night recording can include visible or infrared light. A flash board can be fixed to the wall or bed of a fish pass to flood the area with as much light as possible. The main advantages to this are readily available equipment that is low cost and easy to install. The disadvantages are that turbidity or light reflection can prevent collection of useable images. Cameras also need protection from flood events and images require manual processing or additional software to evaluate which is time consuming. Software such as FishTick can save time on data analysis and avoid operator error. This software has two main functions: it can remove segments of video that lack fish, and be set to record video when fish trigger the software. See Section 3.1.7 for a review of the FishTick software. A study by the UK Environment Agency found that FishTick could review 24 hours of data in 15 minutes, with a detection rate of 90% (Washburn et al. 2008a). Other preliminary reports suggest inaccuracy in enumerating high-density eel and American shad migrations. It is likely more suited to enumerate species with longer period migrations, where fish numbers are dispersed throughout the migration period. As with any other automated fish counting software, raw data quality is key to the efficiency and accuracy of the results. Turbid water and low lighting would make fish detection difficult. IFR staff have not had the opportunity to work with this software and suggest potential users consult FishTick and the UK Environment Agency report (Washburn et al. 2008b) for additional information.

The estimated cost of installing a fish pass video monitoring system is between £5 000 and £8 000 for equipment and labour. A digital video recorder ( DVR) and weatherproof cameras are the minimum equipment required. Low power DVRs or computers fitted with DVR cards can be deployed depending on location and power availability at the site. If installed in conjunction with other fish counting technologies, one dedicated computer can be used to run both the DVR and counter.

Potential Engineering Requirements

Engineering costs for deployment of video counters are generally limited. Like sonar and hydroacoustic counters, there may be the need for specific brackets, mounts or fence designs, but in general, these can be generated from existing designs and/or fabricators at lower costs unless there is a specific engineer signoff requirement. Structural engineering is rarely required unless the project includes modifications to a fish pass.

2.2 Counter Technology Specification Sheets

Counter specification sheets summarize the details of the counters and the advantages and limitations presented in the previous section that provided a more comprehensive review of each technology. These sheets are designed to be a standalone section that will provide readers an "at a glance" comparison of each counter type. These specification sheets are formatted for printing.

2.2.1 Multibeam Sonar
Multibeam sonars function by emitting numerous small acoustic beams (sound pulses) and converting the returning echoes into high quality images. The images are stitched together by proprietary software, resulting in data that resembles a video format. The data can be processed by software or manually to produce counts. Most of the specifications are similar between the two manufactures, however specific differences are noted in the tables (BlueView = BV, Sound Metrics = SM).

Manufacturers and Models
Manufacturer	Model	Cost	Company size
Teledyne BlueView	M900-90, M900-2250	£15 300 and £22 500	Large
Sound Metrics Corp.	DIDSON 300, ARIS Explorer 1800	£48 030 and £51 236	Large

Required Equipment
Equipment	Source	Comments
Sonar/transducer	Manufacturer	SM: Included is 15 m cable, other lengths are extra BV: 1.5 m cable included, other lengths extra
Sonar/transducer mount	Manufacturer and third party fabrication	May include pan/tilt rotator
Sonar lenses	Manufacturer (Sound Metrics)	Concentrator, telephoto, or spreader lens dependent on site specific requirements
Logging device	Personal computer
Download device	Personal computer
Lighting board	Manufacturer or third party
Onsite equipment storage	Third party fabrication

Power Requirements
Power source	Supply voltage	Power consumption
Mains	12 - 48 Volts	25 - 150 watts or 2.5 - 12.5 Ah
Battery	12 - 48 Volts	25 - 150 watts or 2.5 - 12.5 Ah

Data Management and Software
Application	Software and source	Comments
Logging, downloading and setup	Both: Proprietary software	BV: File recorded in . SON format SM: File recorded in . DDF, and . ARIS format
Analysis	SM: Proprietary software or third party (e.g. Echoview, R) BV: third party software	Both: Counts not provided from raw data, post-processing required, using proprietary or third party software.
Data storage	Internal and personal computer	SM: DIDSON, up to 8 GB of internal storage (4hrs of recording) or personal computer BV: Requires personal computer

Operational Requirements
Operation	Schedule	Comments
Sonar installation	Annually (user defined/site specific)	Permanent structure vs. temporary structure
Sonar setup	As conditions change	Site specific
Downloading	User defined, site specific	Dependent on files sizes, and data management requirements
Sonar maintenance	Manufacturer recommended (i.e., annually )	e.g., Recalibration, seal maintenance, cable maintenance
Calibration	User defined or internal calibration

Sensor Unit Structure and Engineering
Structures	Engineering	River position	Comments
Fixed concrete structure	Third party construction	In river anchored or mounted to a shore structure	SM: Max width 10 m at high frequency and 40 m at low frequency
Diversion fence	Third party fabrication of fence and supporting structures	In river	This is a common structure used in combination with multibeam counters
Portable mount	Manufacturer and third party fabrication	In river; aimed to scan sections of interest	This may require construction of diversion fences; maximum width limits are as discussed above

	Advantages	Limitations
Sonar	-Plug and play	-High cost for sonar transducer
Data management and software	-Potential for data backup -Software provided and able to provide count data	-Need personal computer running concurrent to log and store data -Post processing for count data- requires time -Lack of validation
Power requirements		High power requirements
Operational requirements	-Low sonar maintenance	-May require personnel to be on site to download data daily
Sensor unit structure and engineering	-Does not require permanent structures to be constructed	- SM: Limited to small and medium river sizes

2.2.2 Splitbeam Sonars
Echosounders function by transmitting a high voltage pulse to a transducer. The transducer converts the electrical energy to sound pressure, which radiates spherically from the transducer. The sound pressure wave (pulse) then travels through the water. If the wave impacts an object with a density different than that of the surrounding water, a certain amount of the pressure is reflected back and detected by the transducer. The transducer converts the reflected pressure wave into electrical energy that is amplified and filtered by the echo sound receiver. This receiver then creates an output signal that can be measured and further provide information on the target (object) that reflected the energy (pulse).

Manufacturers and Models
Manufacturer	Model	Cost	Company size
BioSonics	DT-X Echosounder + Transducer	£35 492	Medium

Required Equipment
Equipment	Source	Comments
Transducers unit	Manufacturer	Sell echosounder and transducer as kits
Logging device	Personal Computer	Remote sites require additional DT-X AMS system that automates logging and operation (cost - £17 261)
Download device	Personal computer
Transducer mount/deployment	Manufacturer and third party fabrication.	User defined
Onsite equipment storage	Third party fabrication

Power Requirements
Power source	Supply voltage	Power consumption
Mains	11-14 V DC or 90-264 V AC	30 Watts, 2.5 Ah at 12 V
Battery	11-14 V DC or 90-264 V AC	30 Watts, 2.5 Ah at 12 V

Data Management and Software
Application	Software and source	Comments
Logging, downloading and setup	Proprietary software	Included with counter
Analysis	Proprietary or user defined (e.g., Echoview)	Has two programs to analyze data VisAcq Auto Track and Visual Analyzer
Data storage	Personal computer	Data is stored on personal computers that operate the equipment

Operational Requirements
Operation	Schedule	Comments
Transducer installation	User defined	Permanent vs. temporary structure
Echosounder setup	As conditions change	Site specific
Downloading	User defined, site specific	Data files are not very large - can be exported as .csv files
Transducer maintenance	Annually	Factory calibration
In Situ Calibration	User defined	completing a field calibration before and after a survey is recommended

Sensor Unit Structure and Engineering
Structures	Engineering	River position	Comments
Fixed concrete structure	Third party construction	In river anchored or mounted to a shore structure	Max width is dependent on riverbed profile.
Fish pass	Manufacturer and third party fabrication of mount for sonar	In river; aimed to scan width of fish pass channel.	Position should take into account varying water levels.
Portable mount	Manufacturer and third party fabrication	In river; aimed to scan sections	Proper site required

	Advantages	Limitations
Echosounder, transducer	-Plug and play	-High cost for unit
Data management and software	-Automated counting and data storage -Potential for data backup -Potential for remote downloading -Software provided -Small file size and large storage capacity	-Need personal computer running concurrent -Proprietary software has limitations - third party software is expensive. -Post processing for count data maybe required
Power requirements		-High power requirements
Operational requirements	-Low counter maintenance	-May require a person to be on site at all times or purchase of expensive remote operation equipment
Sensor unit structure and engineering	-Does not require permanent structures to be constructed	-Requires ideal river profile conditions

2.2.3 Resistivity Counters
Resistivity counters function as resistivity transducers, measuring the bulk resistance of the water through a set of three electrodes. Resistivity counters rely on the body mass of a fish being considerably less resistive than the volume of water that it displaces. When a fish passes over the electrodes, a momentary reduction in resistance is detected and the counter assigns a direction to the movement. Resistivity counters are comprised of a counter box containing the instruments for measuring resistivity and an algorithm that translates changes in resistance into fish movement records, additional equipment is required for installation and use. Most of the specifications are similar between the two manufacturers, however specific differences are noted in the tables (Logie 2100C = LG, Mark 12 = MK).

Manufacturers and Models
Manufacturer	Model	Cost	Company size
Aquantic	Logie 2100C	£10 000	Small
EA Tech and Scottish and Southern Energy	Mark 12	£20 000	Large

Required Equipment
Equipment	Source	Comments
Sensor unit	Third party fabrication	See Sensor Unit Structure and Engineering section
Logging device	Not required	LG: Internal memory MK: Internal memory & Compact Flash card
Download device	Personal computer	MK: Can also have data transferred to Compact Flash card
Lighting board	Manufacturer
Onsite equipment storage	Third party fabrication

Power Requirements
Power source	Supply voltage	Power consumption
Mains	LG: 24 V or 12 V DC MK: 15 V to 75 V DC	LG: 0.3 Ah per channel
Battery	24 V or 12 V DC battery bank	LG: 0.3 Ah per channel (e.g., 12 V 40 Ah battery life for is approximately 2.5 days)

Data Management and Software
Application	Software and source	Comments
Logging, downloading and setup	Proprietary software	Included with counter
Analysis	User defined (e.g., Excel, R)	MK: has proprietary analysis software used for validation
Data storage	Internal and personal computer	LG: Stores data but personal computer required for additional backup of data; outputs .txt files < 100 kb per record MK: Compact Flash card and .txt files

Operational Requirements
Operation	Schedule	Comments
Sensor installation	Once or annually	Permanent structure vs. temporary structure
Counter setup	As conditions change	Site specific
Downloading	User defined, site specific	Max storage in the 10's of thousands or records, remote download potential
Sensor maintenance	Site specific	E.g., removal of algal growth or debris
Counter maintenance	Annually and during installation	Check connections to sensor unit
Calibration	User defined or internal calibration	LG: Internal calibration occur every 30 mins. User may also calibrate conductivity probe.

Sensor Unit Structure and Engineering
Structures	Engineering	River position	Comments
Crump weir	Fixed concrete structure	In river; up to full river width	LG: Max width 5 m per channel (max 4 channels per unit)
Fish pass (tube sensors or flat sensor pad)	Third party fabrication of insert or mount for sensors to fit into fish pass	In river; channel width	MK: Max width 1 m per channel (max 2 channels per unit)
Flat pad	Third party fabrication of sensor pad and anchors	In river; up to full river width	Max width 5 m per channel

	Advantages	Limitations
Counter	-Moderate cost for counter unit	-Can only function in waters with conductivity > 20 µS
Data management and software	-Automated counting & data storage -Potential for data backup -Potential for remote downloading -Software provided -Small file size and large storage capacity	-Need PC running concurrent to counter for data backup - LG: Software is limited in functionality. No analysis capability - LG: Poor stability at upper end of storage capacity.
Power requirements	-Low power requirements - good for use in remote locations	-Operating validation equipment will require more power
Operational requirements	-Low counter maintenance
Sensor unit structure and engineering	-Wide range of sensor array can be used with counter and can be modified to suit river conditions	-Sensor units must be fabricated by third party (high costs, e.g., Crump). -Potentially high cost of installation -Limited to small and medium river sizes or in the case of Mark 12 a fish pass structure (max width of 1 m)

2.2.4 Optical Beam Counters
Optical beam counters emit beams of infrared light from one side of the counter and are received by sensor units on the opposite side. A fish count is created when a fish breaks the two vertical beams of light while swimming in either direction. The two-beam design enables the direction of detection and produces a trace that is recorded with every count. Optical beam counters come as a complete unit with sensor, counter, and software.

Manufacturers and Models
Manufacturer	Model	Cost	Company size
Vaki	Riverwatcher with video validation box	£27 000	Medium

Required Equipment
Equipment	Source	Comments
Sensor unit	Manufacturer
Logging device	Manufacturer
Download device	Personal computer
Lightning board	Third party
Mounting box or enclosure	Manufacturer	This is included as package with Riverwatcher model.
Onsite equipment storage	Third party fabrication

Power Requirements
Power source	Supply voltage	Power consumption
Mains	12 V DC	0.2 Ah
Battery	12 V DC battery bank	One channel (0.2 Ah) (e.g., 12 V 40 Ah battery life for is approximately 8 days)

Data Management and Software
Application	Software and source	Comments
Logging, downloading and setup	Proprietary software	Included with counter
Analysis	Proprietary software or user defined (e.g., Excel, R)	Proprietary software can be used for validation and correcting counts but does not provide estimates of uncertainty
Data storage	Internal or web based server	Riverwatcher daily program can be used to upload data to server and real-time counts can be observed.

Operational Requirements
Operation	Schedule	Comments
Sensor installation	Annually	Can be installed using a permanent or temporary structure
Counter setup	Annually	Site specific - length depth ratios for fish images can be set so the counter can identify species.
Downloading	User defined, site specific	Max storage in the 10's of thousands of records, remote download capable with all units
Sensor maintenance	Site specific	E.g., removal of algal growth or debris
Counter maintenance	Annually and during installation	Very little maintenance required
Calibration	No calibration needed

Sensor Unit Structure and Engineering
Structures	Engineering	River position	Comments
Fish pass	Third party fabrication of insert or mount for sensors to fit into fish pass	In river; channel width	This is the most common structure used to operate optical beam counters
Fence	Third party fabrication of fence and supporting structures	In river; up to full river width	This is a less common structure used to operate optical beam counters

	Advantages	Limitations
Counter	-Moderate cost for counter unit which includes sensor and validation camera	-Can only function in waters with turbidity <90 NTU
Data management and software	-Automated counting and data storage -Potential for data backup -Remote downloading included -Excellent software provided -Small file size and large storage capacity
Power requirements	-Low power requirements - good for use in remote locations
Operational requirements	-Low counter maintenance
Sensor unit structure and engineering	-sensors are manufactured -Validation equipment can be added to sensor unit by manufacture.	-Sensor units are small and multiple units may be required for high migration rates. -Limited to small and medium river sizes and fish passes

2.3 Structures

In this section we briefly summarize the structures used to assist in the enumeration of fish through electronic counter technology. The structures do not count fish, but are used in conjunction with the equipment detailed in Section 2.1 to improve and facilitate the collection of data.

2.3.1 Fences

Fish fences are commonly used worldwide as a primary method for the capture and enumeration of salmon and trout. Fences generally include a full span barrier across a river or fish pass combined with a trap box or passageway into which migrating fish are directed. Upon capture, fish are often handled through netting or crowding so that abundance, species allocation and biological data can be collected. Structures used to construct fences may be temporary or permanent depending on migration timing, river characteristics and project budgets. Permanent impassable structures such as purpose-built concrete weirs, existing rock weirs, and dams with fish passage structures will also be discussed, along with fish pass box and tube counter enumeration methodologies.

Fences can be used in conjunction with remote sensing equipment (i.e., counters, tag readers, etc.) to effectively enumerate migrating populations of salmonids (Rand et al. 2007). Two common examples of temporary diversion fences are picket fences and Alaskan floating fences, both intending to restrict fish movement to a specific area ( Figure 2.9). Both fence types can be interchangeable or may be suited to specific river types and conditions. Collection areas can range from tens of meters in width to < 1 m, and are dependent on the type of technology employed. Sonar or flat pad resistivity counters, for example, can span larger sections of the river, whereas Vaki or resistivity tube counters cover only small sections of the river (< 2 m).

Figure 2.9. Alaskan floating fence used on the Stanislaus River, United States. Panels are 20 ft. long and incorporate a Vaki counter unit. Photo courtesy of Vaki.

Picket Fences

Picket fences are structures composed of vertical pickets that are held together by aluminum rails, connected horizontally between metal tripods ( Figure 2.10). Tripods are placed perpendicular to the riverbed or at an angle to improve fish passage or debris collection, and are typically weighted down with rocks or sandbags. Pickets made from steel or aluminum pipe are pushed down through the holes in the rails and into the river substrate, completing the barrier effect. Such fences can be used in high flow rivers provided debris loading is low and installation is completed during low flows ( Figure 2.11). Increased maintenance costs can be incurred with cleaning and is dependent on the amount of leaf litter and other floating debris present in the system.

Figure 2.10. A single panel 10 ft high picket fence with mounting tripod and three rails. Photo courtesy of InStream Fisheries Research.

Advantages of picket fences include the low cost of fabrication and installation, as well as the portable nature of the units. Disadvantages include the risk of fence breach, daily maintenance for debris removal, picket removal, and risk of blow out or fence loss from high flows.

Figure 2.11. Seton River picket fence with trap box. River width approximately 30 m, discharge 30 m ³/s, depth 1 m. Photo courtesy of InStream Fisheries Research.

Floating Fences

Alaskan floating fences use a combination of air-filled watertight PVC pipes held at equal spacing by a metal frame to form panels (McCubbing and Ward 2007, 2008). Each panel receives additional floatation by the addition of a planer board attached to the downstream end of the panel. This board, typically made from buoyant materials such as plywood or insulation foam, is attached by a hinge at one end under the panel and by adjustable chains at the other. By adjusting the chains, the angle of the panel can be determined by the operator. Panels are attached together and held in place on the riverbed by a wire running through cleats attached to the panels' upstream end ( Figure 2.12). The wire is fixed to a mooring point on the riverbed. Panels can be 3 to 10 m in length and should float at an upward angle from the upstream to downstream direction. As water depth and velocity increases, the panels run the risk of sinking. The planer boards provide additional lift over and above the buoyancy of the PVC pipes, thus preventing the fence from sinking.

Advantages of Alaskan floating fences include low maintenance and installation costs, as well as the semi-portable nature of the units. Disadvantages include the risk of fence sinking when too much debris builds up on the fence, the need for a mounting structure on the riverbed, and the risk of damage to PVC pickets in debris laden flood events.

Figure 2.12. Keogh River Alaskan floating fence with 3 m long panels and trap box located on the far riverbank. Photo courtesy of InStream Fisheries Research.

Fence Installation Costs

The cost of manufacturing picket and Alaskan floating fences is highly dependent on the cost of raw materials and local machine shop labor. The primary cost is in aluminum, which varies greatly depending on purchasing time. Thus, any direction provided on the cost per meter of fence construction is subject to change. The market cost of any fence installation would need to be verified if it is to be compared with alternate technologies.

2.3.2 Resistivity Sensor Structures

Crump Weir

A Crump weir is a triangular structure with a 1:2 sloping front face and a 1:5 sloping back face (McCubbing and Ignance 2000) ( Figure 2.13). Crump weirs were originally used as measuring structures in open flow channels to predict discharge and change flow characteristics in rivers. Since the Crump weir is fixed in the river, water flows over the weir without the downstream level being below the weir crest. Thus, Crump weirs have the advantage of producing a constant discharge coefficient over a wide range of natural discharge conditions. Crump (1952) developed the original design, with further investigations completed by White (1971). The application of Crump weirs in fish migration monitoring was first evaluated in the 1960's and 1970's in the UK by Simpson (1978). By pairing them with sensor units attached to a resistivity fish counter, the combination of the weir structure and sensor modified fish behaviour (i.e., fish swam at a constant height above sensors) and improved counter accuracy. Fish counts resulted in accuracies greater than 90% for validated upstream migrating salmonids (McCubbing and Ignance 2000). Crump weirs are the most common structures used to deploy resistivity counter sensors ( Figure 2.13). Sensor units made from a pad of non-conductive materials (e.g., high-density polyethylene or fibreglass) are placed on the 1:5 sloping downstream face of the weir where fish passage is generally close to the weir face. Electrodes made from a non-corrosive material such as stainless steel are set into the pad to avoid damage by debris. Spacing between the electrodes is dependent on the size of the target species. Mounting methods and sidewalls must be insulated to avoid electric conductance with sensor electrodes.

Figure 2.13. A four channel Crump weir with resistivity counter sensors installed. Photo courtesy of InStream Fisheries Research.

Crump weirs can take two forms: a mass concrete structure ( Figure 2.14; see McCubbing and Ward 2007, 2008), or a combination of prefabricated weirs made from concrete, aluminum, fibreglass, or wood that are attached to a solid sill ( Figure 2.15; see McCubbing et al. 2014). Sills are partially buried in the riverbed and can also be made from a number of materials, including a mass concrete pour, concrete lock blocks tied together with a thin pour sill, gabion baskets, or a metal substructure. Local engineering requirements, river substrate and discharge, site width and construction access would likely influence the type of Crump weir to be constructed. Life expectancy of these weirs is in the range of 15 to 20 years, after which maintenance and upgrades can be substantial.

The cost of installing a Crump weir, without sensor electrodes, can vary from £20 000 for a small stream to £750 000 depending on site characteristics. Most of this cost variance is determined by the size of the river.

Figure 2.14. Helmsdale resistivity fish counter weir with four channels at low flow. Photo courtesy of InStream Fisheries Research.

Figure 2.15. Glendale spawning channel (British Columbia, Canada) with a hybrid Crump weir and diversion fence. Photo courtesy of InStream Fisheries Research.

Flat pads

IFR has designed a number of modular flat pad sensors that can be directly placed on the riverbed with very low impact (McCubbing and Lingard 2013, D. Ramos-Espinoza unpublished data). Flat pad sensors can be used in situations where water depths at the enumeration site remain relatively low (< 0.5 m max depth). Pad frames are made from non-conductive materials and provide a mounting location for the electrodes. Structural strength is provided by non-conductive grommets with the frames open between electrodes to avoid surfing (i.e., raising of pads from riverbed). Pads can be used in series to provide multiple channels covering the desired width of the enumeration site ( Figures 2.16 and 2.17).

Figure 2.16. Chilcotin River flat pad sensor prior to (left) and after installation (right). River width approximately 24 m. Photo courtesy of InStream Fisheries Research.

Figure 2.17. Crawford Creek flat pad counter. Channel width approximately 10 m. Photo courtesy of InStream Fisheries Research.

The cost of building and installing flat pads are generally low. The equipment and supplies can be purchased for approximately £300 per 1 m section. Fabrication can be undertaken by anyone with power tool experience and IFR has developed assembly plans for this. The units can be held in place on the riverbed by a heavy-duty chain ballast, duckbill anchors or something similar.

Box Sensors

Box sensors have been developed for specific applications in fish passes where an already restricted area is available for fish passage ( Figure 2.18). As with other sensor types, the essential elements of design include electrodes set into a non-conductive base, non-conductive sidewalls (e.g., fibreglass, plastic) and a maximum water depth of 1 m. All conductive elements must be at least 1.5 times the distance between electrode pairs to avoid signal leakage.

These units are specialized and are hard to assess for budgeting purposes. Each site is unique and thus requires a configuration that fits the location. A minimum of £5000 is likely required for small fish pass sites, but this could rise substantially on larger, more complex, locations.

Figure 2.18. Box sensor unit awaiting deployment. Photo courtesy of InStream Fisheries Research.

Tube Counters

The placement of electrodes in counting tubes has been common practice for many years (e.g., Tummel fish pass, Pitlochry, UK). The design of the tube sensors involves installing the circular electrodes at suitable spacing within the non-conductive interior of the tube. More recently, tube counters have been used in Canada where high fish passage rates (i.e., up to 8 000 fish per day) have resulted in 'banks' of up to eight sensors being used in conjunction with a separation grid to provide accurate counts (McCubbing 2012, Ladell and McCubbing 2013a, 2013b, Casselman et. al. 2014). Tubes can be large in orifice (> 1 m; Tummel and Pitlochry Rivers, Scotland, UK) or smaller (~0.35 m; Seton Dam, Lillooet, BC; Figures 2.19 and 2.20). Smaller tubes are used when fish need to be in close proximity to electrodes to produce accurate counts.

These units are like a specialized box sensor and are also hard to assess for budgeting purposes. Each site is somewhat unique and thus requires a configuration that fits the location. A minimum of £4000 is likely required for small fish pass sites, and this could rise substantially on larger, more complex, locations, where multiple tubes and large separation grids are required.

Figure 2.19. Seton Dam fish pass counter with eight sensor units. Note that only the top four tubes are visible. Photo courtesy of InStream Fisheries Research.

Figure 2.20. Seton Dam fish pass sensor units prior to installation in a separation grid. Photo courtesy of InStream Fisheries Research.

References

Alaska Department of Fish and Game. 2015. Alaska Fisheries Sonar. Juneau, Alaska, United States. Available from http://www.adfg.alaska.gov/index.cfm?adfg=sonar.didson.
Aprahamian, M., Nicholson, S., McCubbing, D. and Davidson, I. 1996. The use of resistivity fish counters in fish stock assessment. In: Stock Assessment in Inland Fisheries, I.G. Cowx (ed). Fishing News Books, Oxford, 27-43.
Aquantic Ltd. 1998. 2100C Logie Fish Counter - Operating and Technical Manual.
Baumgartner, L. J.; Reynoldson, N.; Cameron, L.; Stanger, J., 2006: Assessment of a dual-frequency identification sonar ( DIDSON) for application in fish migration studies. Fisheries Final Report Series No. 84, NSW Department of Primary Industries.
Baumgartner, L. J., Bettanin, M., McPherson, J., Jones, M., Zampatti, B. and Beyer, K. (2012), Influence of turbidity and passage rate on the efficiency of an infrared counter to enumerate and measure riverine fish. Journal of Applied Ichthyology, 28: 531-536.
Belcher, E.O. 2004, Case Study: Alaska Department of Fish and Game Uses DIDSON to Count Salmon Swimming Up-River to Spawn. Sound Metrics Corp. Bellevue, Washington, United States. 4 p.
Casselman, M.T., N.J. Burnett, N.N. Bett, C.T. Middleton, E.G. Martins, D.C. Braun, D. McCubbing, and S.G. Hinch. 2014. BRGMON-14 Effectiveness of Cayoosh flow dilution, dam operation, and fishway passage on delay and survival of upstream migration of salmon in the Seton-Anderson watershed. Annual Report - 2013. Report prepared for St'át'imc Government Services and BC Hydro. The University of British Columbia, Vancouver, BC. 66 p. + 2 Apps.
Cronkite, G.M., Enzenhofer, H.J., Ridley, T., Holmes, J., Lilja, J., and Benner, K.. 2006. Use of high-frequency imaging sonar to estimate adult sockeye salmon escapement in the Horsefly River. Can. Tech. Rep. Fish. Aquat. Sci. 2647.
Cronkite, G.M., Enzenhofer, H.J., and Holmes, J.A. 2008. Evaluation of the BlueView ProViewer 900 Imaging Sonar as a tool for counting adult sockeye salmon in the Adams River, British Columbia. Can. Tech. Rep. Fish. Aquat. Sci. 2798.
Crump, E.S. 1952. A new method of gauging stream flow with little afflux by means of submerged weir of triangular profile. Proc. Inst. Civil Engrs. Part 1, Vol. 1. 223-242 pp.
EA Technology Ltd. 2006. Fish Counter System - User Guide Version 1.4, Guide produced for Scottish and Southern Energy.
Ladell, J., and McCubbing, D. 2013a. Fish counter enumeration of steelhead and rainbow trout on the Bonaparte River in 2011. Report submitted to Ministry of Environment, Fisheries Branch, Kamloops, BC. 25 p.
Ladell, J., and McCubbing, D. 2013b. Fish counter enumeration of steelhead and rainbow trout on the Bonaparte River in 2012. Report submitted to Ministry of Environment, Fisheries Branch, Kamloops, BC. 29 p.
McCubbing, D., and Ignace, D. 2000. Salmonid escapement estimates on the Deadman River, resistivity counter video validation and escapement estimates. MEOLP Project Report Number 2000, Vancouver, Canada.
McCubbing , D.J.F., Ward, B., and Burroughs, L. 2000. Salmonid escapement on the Keogh River: a demonstration of a resistivity counter in British Columbia. Province of British Columbia Fisheries Technical Circular no. 104, 25 pp.
McCubbing, D., and Ward, B. 2007. Adult steelhead trout and salmonid smolt migration at the Keogh River, B.C., during winter and spring 2007. Report prepared for Habitat Conservation Trust Fund. Contract #CBIO4051.
McCubbing, D., and Ward, B. 2008. Adult steelhead trout and salmonid smolt migration at the Keogh River, B.C., during winter and spring 2008. Report prepared for Habitat Conservation Trust Fund. 41 p.
McCubbing, D. 2010. Fish counter enumeration of steelhead and rainbow trout on the Bonaparte River in 2009. Report submitted to Ministry of Environment, Fisheries Branch, Kamloops, BC. 30 p.
McCubbing, D., and Lingard, S. 2013. Coho salmon adult spawner escapement estimates for the Lemieux River based on resistivity counts, 2012. Technical report prepared for Fisheries and Oceans Canada, Kamloops. 24 p.
McCubbing, D., Melville, C., Wilson, LJ, and Ramos-Espinoza, D. 2014. Lower Bridge River adult salmon and steelhead enumeration 2013. Report prepared for BC Hydro. 57 p. + 1 App.
Rakowitz, G. 2009. Didson counting - manual or automatic? Proceedings of the underwater acoustic measurements: technologies and results. Nafplion, Greece, 6pp.
Simpson, D. 1978. Electrical resistivity fish counters. In Rhythmic Activity of Fishes, J.E. Thorpe, ed., pp. 259-268. London: Academic Press.
Sound Metrics Corp. 2015. DIDSON Sonars. Bellevue, Washington, United States. Available from http://www.soundmetrics.com/products/didson-sonars/Software.
White, W.R. 1971. The performance of two-dimensional and flat-V triangular profile weirs. Proc. Inst. Civil Engrs. Suppl. (ii) Paper 7350 S. 48 pp.
Washburn, E., Gregory, J., and Clabburn, P. 2008a. Development of a fish counting system for fish passes. Environmental Agency Internal Report.
Washburn, E., Gregory, J., and Clabburn, P. 2008b. Using video images for fisheries monitoring. Environmental Agency Internal Report.