The Eider Spider: Development and Experimental Testing of a Novel Method to Eliminate Sea Duck Predation on Mussel Farms

Final Report
Valeo Management LP
AIMAP 2012-Q02

1. Background

The cultivation of mussels is a growing industry worldwide. Predation by birds such as migrating sea ducks however poses an ongoing challenge to mussel growers and is often associated with significant financial losses. Numerous bird deterrent methods have been tested but limited success has been achieved under the pressure of intensive predation. Ice coverage and the ducks' seasonal migration patterns limit their predation activity in most areas of Atlantic Canada, but with the prospect of milder winters due to global warming, losses are expected to increase in the future. Small mussel farms in Nova Scotia currently estimate their annual losses at between $100,000 and $200,000. The following project describes a new deterrent strategy (hereafter referred to as the Spider), that is not only cost-effective but is also conservation-friendly. Preliminary trials with a home-made prototype in a tank set-up had successfully demonstrated the efficacy of the Spider. The following project describes the results of testing this innovative approach in field trials at two commercial sites in Nova Scotia.

2. Location of the mussel farms in Nova Scotia

The original plan was to set up this project at Atlantic Aqua Farms in Whitehaven Harbour, NS and at Indian Point Marine Farms in Mahone Bay, NS. However, Atlantic Aqua Farms, a PEI-owned company, decided to scale back their operation in 2012 due to a heavy tunicate infestation, and the test site was relocated to AquaPrime Mussel Ranch in Ship Harbour, NS. Both farms depend on the purchase of mussel seed from external sources, an operation which is typically done in November. In these trials the seed used at both test sites was purchased from St. Anne's Bay in Cape Breton, NS.

2.1 AquaPrime Mussel Ranch Ltd.

This company has been operating a mussel operation in Ship Harbour, NS since 1995. It is one of the few bays which have so far escaped the tunicate invasion, although in 2012 a few specimens were noted at the entrance of the bay. They primarily use the continuous sleeving method for their culture operations. Apart from sporadic closures due to algal toxins, sea duck predation is the major biological issue which undermines their operation. Their strategy to control ducks has been to engage in high speed boat chasing, encourage hunting, and to set up visible "sacrificial" mussel lines which are offered as food to ducks in an area where they can be controlled.

2.2 Indian Point Marine Farms

This company has been growing mussels in the waters of Mahone Bay, NS since 1982. The goal of the company is to produce high quality, safe and nutritious food while minimizing the impact on the marine environment. They primarily use the continuous sleeving method as their deep site is well-suited for this type of culture method. Over the past seven to eight years, this company has had to deal with a heavy infestation of the vase tunicate Ciona intestinalis. The overwhelming biomass of this pest organism is forcing the company to scale back their operation. Sea duck predation was the primary biological problem prior to the tunicate invasion, but it still remains an issue from December to mid-April.

3. Description of the work and experimental protocol

3.1 Experimental set-up

Both operators independently decided that the Spider technology in its current state would not be suitable for the continuous sleeving culture method, especially in terms of their removal during the harvesting phase. However, they were both interested in testing this technology to determine whether it had a significant deterrent effect on sea ducks. At Ship Harbour and Mahone Bay, we chose to use 2.5-m single socks as the experimental unit to evaluate the impact of this new technology on duck predation activity. At both farm sites, two longlines each with 96 single socks were deployed. Two Spiders were installed on the first 12 socks on each longline, followed by 3 Spiders on the next 12 socks. The next set of 12 socks was designated as a control and the following 12 socks were left untouched to produce a buffer zone between this first block of 48 socks and the second replicated sequence of 48 socks on the same line. This experimental design (2 blocks of 48 socks) was repeated on another line approximately 50 m away. 

3.2 Ship Harbour

The Spider technology was deployed at the Ship Harbour site on November 16, 2012. The assembly time (i.e., inserting the arms into the disk) was estimated at one Spider per min per person, and the deployment time (i.e., installing and locking the disk in position) at four Spiders per min per person. The initial number of mussels per sock was estimated at 3000 (approx 500/ft x 8 ft). Attempts to install 4 Spiders on 1 sock failed as the sock floated upon deployment; hence, we had to readjust our initial objective of comparing 2 versus 4 Spiders, to 2 versus 3 Spiders per sock. It should be noted that these mussel socks had already been in the water for 7 d and the mussels were well hydrated. Given the positive buoyancy of the Spider unit, attempting to deploy this technology on new socks with mussel seed harvested 2-3 d previously would not be advisable. Given that the mussels tend to lose a substantial amount of water, it is likely that new socks would float for at least 10-15 min on the surface and the Spiders would quickly become entangled. To reduce the probability of entanglement between adjacent socks, the spacing between the socks was increased to 1 m, as opposed to 0.75 m, as in the original plan.

3.3 Mahone Bay

This site was set up on December 16, 2012 following the same experimental layout as Ship Harbour (2 lines, 96 socks per line). The set-up had to be delayed until the operator was confident that the settlement of tunicates had ceased; the tunicate settlement window has been steadily expanding over the past decade with the warmer fall seawater temperatures. On the deployment day, the air temperature was at -30C with a wind speed of around 10-15 km/h. The set-up went smoothly but there were issues with socks floating at the surface and the Spiders cracking, possible due to the cold temperature.

4. Results

The complete dataset including the two experimental sites, the four experimental lines, the two blocks within each experimental line, and the three treatments (Control versus 2-Spiders versus 3-Spiders) was evaluated by means of a mixed model analysis of variance. Only the Spider Treatment was declared a fixed effect and was shown to be significant for the net weight and the percent of mussels remaining in a mussel sock. The least square means did show higher mean values in terms of the proportion of mussels remaining and the net weight for those socks fitted with the Spider technology than the Control socks. 

From a commercial context perspective, the sleeves in December 2012 were stocked with 3000 mussels at Ship Harbour and 1600 mussels at Mahone Bay. There was a substantial decline across all treatment groups over the winter months and this loss is most likely attributable to duck predation. The statistically-significant experimental effect of the Spider technology could be related to a reduction in the number of mussels lost due to slippage rather than a reduction in duck predation. According to the mussel producers, the mussel socks at the end of the study period were considered of marginal commercial values.

4.1 Ship Harbour

Staff at AquaPrime reported that significant numbers of sea ducks, in particular the common Goldeneye, began to arrive at the Ship Harbour site in mid-January. Eider ducks and scoters were also present but in smaller flocks. The assessment of the experimental lines was conducted in late April 2013. In the first experimental line surveyed, every single mussel sock had been severely affected by duck predation. At times, there were small sections at the bottom of a sock below a Spider which appeared to have been either missed by the ducks and/or protected by the Spider. However, in commercial terms the impact of the heavy duck predation translated into a near complete loss of mussel inventory. When the second experimental line was evaluated, generally full mussel socks with either two or three Spiders were observed at the beginning of the first Block followed by decimated Control socks similar to the first experimental line. Moving on to the second experimental Block on this second line, we found that all the socks, with or without Spiders, had suffered heavy duck predation. It would appear that there was a position effect and we surmised that the ducks had somehow missed the beginning of second experimental line. We suspect that some surviving mussels which had been trapped/protected inside the mesh over the winter did move out after the ducks had departed.

A serious design flaw with the Spider is the positive buoyancy which leads to problems of entanglement with neighboring sleeves and additional mussel losses. We would also surmise that, as the mussels are being consumed, the sleeves with the Spiders become progressively lighter and thus tend to float making them an easier target for ducks. In any case, this issue of positive buoyancy would need to be addressed if additional testing were to be carried out. It should be noted that the Spider technology was tested in a tank using scoters as the model duck. At Ship Harbour, the primary duck predator was the common Goldeneye which is considered a medium-sized sea duck compared to scoters which are considered large sea ducks. It is possible that the spacing of the arms on the Spider technology was not a suitable deterrent either in terms of the size or the behaviour of the common Goldeneye.

4.2 Indian Point Marine Farms

According to Peter Darnell, the mussel producer at the Mahone Bay location, the level of predation was substantially less than usual. It is possible that the lower production level or the high abundance of tunicates on the mussel sleeves is becoming less attractive as a food source for ducks. Two species of ducks were observed near the mussel lines, Old Squaw and scoters, from March onwards. There was generally higher mussel survival than had been observed at Ship Harbour, but overall no obvious advantage to employing the Spider technology for the purpose of duck deterrence. The producer noted that the mussels had been slow to move out of the socks and the buoyancy from the Spiders caused issues with the socks becoming entangled either among themselves or along the main lines. 

5. Discussion

The original plan was to deploy the Spider technology on continuous mussel sleeving but both Nova Scotia farm operators stated that this technology in its current state would interfere with harvesting using their hydraulic field gear system. Despite this issue, there was still considerable interest in determining whether this technology had a deterrent effect on sea duck predation, and thus the choice was made to undertake the trials with single socks. 

At an experimental level, the presence of the Spider technology on the mussel socks was statistically shown to increase mussel retention and did result in a higher net weight at the end of the study. It was obvious that the Spider did provide sheltered pockets under the disk which were less accessible to the ducks. However, at best, the Spider technology may have produced a 12% higher retention of mussels compared to the Control socks, i.e., an increase from 23% to 35%. Unfortunately, apart from this one anomalous zone, substantial predation was noted on all the mussel socks, including the Control and the Spider-fitted socks. 

The deployment of the Spider technology would be commercially feasible if the return on the investment was significant. For the deployment, our observations suggested that the assembly time is the limiting factor; basically, one person installing Spiders on the mussel socks could keep up with four persons engaged in the assembly of the units. The reduction on the number of arms or having one assembly piece with multiple arms would be one way to decrease the assembly time. Other issues related to this technology were identified during the set-up phase: the need to increase the spacing between socks in order to preclude entanglement; the tendency for the socks to float due to the positive buoyancy of the Spiders; the limited working area due to the high space requirements of the units; and the increased rate of disc breakage at low temperatures. The single sock culture method is extensively used in PEI which would potentially be the major target market for this technology in Eastern Canada. However, spacing between socks is typically set at 0.75 m; a greater spacing would translate into a lower production per longline. A reduction in the diameter of the central disk might help resolve this issue. The positive buoyancy of the Spider is a serious issue in that neighbouring socks rapidly became entangled, especially under conditions of high current or wave action. Switching to a negatively-buoyant material would help to alleviate this problem. The issue of the high space requirement of these units would need to be addressed in terms of the logistics of deployment and subsequent retrieval, or through changing the design so as to render the units "stackable". Finally, identifying a material that will perform better at low temperatures is crucial for the deployment and harvest phase. Winter harvesting at sub-zero temperatures is a normal practice in PEI and this technology must be sufficiently pliable to sustain rough handling under these conditions.