Voles on rolls

Where to begin...  It's been four years since I last wrote a blog post and I feel that it's about time that I started with a clean slate. Those that used to follow my blog back in 2012/2013 may recall my pre-university Kalahari* Desert adventures. I spent six months working at the Kalahari* Meerkat Project assisting with the day-to-day recording of meerkat behaviour and life history data. Since then I have completed a Bachelor's degree in Zoology at Swansea University and a Master's degree in Conservation Science at Imperial College London. It was the search for a suitable MSc thesis project which drew me to the Fens* of Cambridgeshire; where I'm now employed as Assistant to the Environmental Officer of the Middle Level Commissioners.

        *Some technical terms, geographical locations and other explanatory information
          are listed in a glossary at the foot of the page.

For my thesis project, I researched the effectiveness of coir roll revetment* installations to create suitable habitats for breeding water voles (Arvicola amphibius) in the Fenland* main drains. Coir rolls are a type of organic bank toe revetment taking the form of long cylindrical meshed nets packed with coir* and pre-established riparian* plants. The coir roll revetments in question were installed by the Middle Level Commissioners between 2009 and 2015 and incorporate water vole food plants such as sedge (Carex acutiformis), based on the idea that they reduce bank erosion whilst encouraging water vole occupancy. To the best of my knowledge, my project is novel, since no one has scientifically quantified the effect of these installations on water vole occupancy until now.

The plight of the water vole

The European water vole, fondly known as 'Ratty' from Kenneth Grahame's Wind in the Willows once used to be a common sighting on Britain's waterways. Sadly, this lovable creature is now Britain’s fastest declining native mammal, reportedly vanishing from 70% of sites surveyed by the National Water Vole Monitoring Programme between 1989-1990 and 1996-1998, and is thus listed as a Biodiversity Action Plan (BAP) Priority Species. The introduction of the invasive* American mink (Neovison vison) and adjacent land-use changes are largely blamed for this decline, however little research has investigated how riparian erosion-protection practices may also affect vole populations by making banks inhospitable.

A water vole (Arvicola amphibius). Note the blunt nose, short tail, small ears and lack of neck, 

distinguishing it from a rat (Rattus norvegicus).

Data collection
I was keen to find out whether the use of coir roll revetments is a more water vole-friendly riparian management alternative (as anecdotal evidence suggests) to other commonly used strategies. In order to do this, I surveyed for the presence of water vole latrines*, droppings and feeding signs at 22 coir roll revetment sites to assess water vole occupancy. For comparison, I also surveyed for water vole signs at 22 sites assigned to each of the following riparian management categories:

• Bare bank: devoid of any management (hard engineering or bioengineering). <50% of the bank toe’s length was vegetated.
• Natural sedge bank: >50% of the water’s edge vegetation was comprised of sedge species (Cyperaceae). Devoid of bank management.
• Sedge plug ledge: A 30 cm wide mud shelf (otherwise known as a berm*) installed just above the summer water level and planted with pre-established lesser pond sedge plugs.
• Hard engineering: Corrugated metal sheets and wooden boards enclosing flint stones at the water’s edge.

In total, I had 110 sites (22 for each category) spread across Cambridgeshire and Norfolk. At each of these sites, I surveyed an 8 m stretch of bank for water vole signs, plant species abundance and vegetation height every two weeks between mid-May and early July 2017. To facilitate the recording of signs, Cliff Carson (the Environmental Officer) introduced me to the idea of installing floating wooden rafts in the centre of each 8 m section (tethered to the bank using string and bamboo canes) to act as artificial latrine and feeding platforms (see trail camera pictures below). This ingenious method proved to be extremely reliable for detecting these elusive, semi-aquatic and largely crepuscular* or nocturnal rodents, allowing for some interesting results.

                                        Above: a water vole using one of my 110 survey rafts (June 2017)                                

                                        Above left: one of my 110 survey rafts: note the presence of tic tac sized droppings and feeding remains (sedge                (Carex acutiformis) and willowherb (Epilobium sp.)) (May 2017). Above right: An example of the water vole bank                  signs which I was recording: note the latrine located in the burrow entrance (20 pence for scale) (April 2017).

Some of the most rewarding moments during my fieldwork were checking my trail cameras for water vole footage. I just love the way this charming vole cleans its ears in the first video and nibbles on the reed sweet grass at the end of the second.

        

Above: trail camera videos showing a water vole going about its daily business on one of my rafts (June 2017).

My findings
For the purpose of this blog post, I've excluded the more complicated stats involving occupancy analysis in R*, and refer only to percentage data and pie charts. If you'd like more detail about my statistical methodology and findings, please feel free to drop me a line.

Broadly speaking, my project presented three key findings; firstly, water voles appeared to favour coir roll revetments over the other four management types. A staggering 21/22 coir roll revetment sites (95.45%) were found to have latrines during at least 1/4 survey visits (see pie charts below). Natural sedge banks revealed similarly high water vole occupancy, with latrines recorded at 19/22 sites (86.36%). The management type with the lowest number of occupied sites was bare bank (n = 9/22).

Secondly, the abundance of sedge (which is particularly prevalent in coir roll revetments) appeared to be an important factor influencing water vole occupancy. Sedge was the most frequently recorded plant from feeding remains at coir roll revetment, sedge plug ledge and natural sedge sites, agreeing with the literature that it is a dominant food source. Where management types presented a very low abundance of sedge (hard engineering and bare bank), water voles appeared to compromise by eating a greater variety of other plants. Reed sweet-grass (Glyceria maxima) was a particular favourite at such sites where sedge was rare.

Thirdly, water vole occupancy was positively related to vegetation height at the water's edge, i.e. the taller the vegetation along a section of bank, the more likely it was that water voles were there. In most cases, bare banks had the least and shortest vegetation, which may further explain its comparatively low occupancy.

Water vole occupancy (evidenced by the presence of latrines observed during at least one of the four survey visits) for each of the management types.

    

The percentage of survey sites at which water vole latrines were detected (during at least one of the four survey visits) by management type.

The data displayed in the pie charts above suggest that, in order to encourage and/or maintain high water vole occupancy, waterway management bodies should refrain from leaving banks 'bare' or installing sedge plug ledges. These management options lack erosion protection infrastructure, which subsequently reduces the ability of vegetation to persist (required by water voles for both food and shelter). Contrastingly, the use of hard engineering practices such as metal sheeting should be avoided as it is likely that these cause the opposite effect of too much erosion protection, resulting in reduced bank penetrability for both burrowing water voles and the plants on which they depend. Instead, waterway management bodies and wildlife authorities looking to improve banks for water voles should consider the use of coir roll revetments and encourage the growth of natural sedge banks.

Above: (left) the end of a section of coir roll located along Fillenham's Drain in Chatteris and (right) a sward of well-established vegetation (predominantly sedge) growing from a section of coir roll revetment just three years after installation (2017).

It appears that erosion protection is an important and pivotal factor in determining bank vegetation cover, and therefore suitability for water voles. It is possible that a trade-off between the effective erosion protection offered by harder engineering options, and preferential bank/vegetation characteristics offered by softer approaches must be met, in order to achieve high occupancy. Coir roll revetments possibly meet this compromise effectively, as the dense, protective coconut husk material and meshed netting, supported by wooden posts may absorb the hydraulic power of the river. These features have likely allowed the plant plugs within the coir rolls, and the surrounding bank vegetation to avoid erosion damage and become established, whilst simultaneously still offering a penetrable bank to burrowing water voles. 

Another explanation for the success of coir roll revetments (and indeed natural sedge banks), is that they both have a high density of sedge, a key provider of food and shelter for water voles. Bare bank and hard engineering sites were distinctly lacking in sedge, whilst the sedge which had been planted at sedge plug ledge sites had largely been eroded away or replaced by common reed (Phragmites australis). The lack of sedge (and the reduced vegetation diversity in general) at bare bank sites is likely due to there being no protection from erosion. On the subject of vegetation, a likely reason for occupancy being positively related to vegetation height is that taller vegetation is generally associated with reduced predation risk due to the provision of more shelter.

Not only do my results show that coir roll revetments have the ability to improve habitats for pre-existing water vole populations, but they also suggest that coir roll revetments can encourage water voles to previously unoccupied sites. This was illustrated by five of my coir roll revetment survey sites located along a diverted section of river called the Fillenham's Drain in Chatteris (Cambridgeshire). Water voles were absent from this location according to a survey undertaken by Cliff Carson in 2014 prior to the river diversion and subsequent installation of coir roll revetments. In just three years, water voles have successfully colonised this stretch of drain, resulting in some of the most occupied survey sites.

At a regional scale, the high water vole occupancy of the Fenland main drains revealed by my research (even within less suitable habitats: bare banks, sedge plug ledges and hard engineered banks) is encouraging considering the background of national decline. The results from this study concur with those of a recent water vole survey of two Cambridgeshire drainage districts* (Curf Fen and Ransonmoor) undertaken by the Wildlife Trust in 2015, in that the Cambridgeshire Fens are a stronghold for water voles. Whilst the 2015 study surveyed only the smaller Fenland drains, and my study surveyed just the main (larger) drains, it is interesting to see that water vole occupancy appears to be just as high on the main drains. Previous papers report that narrower river courses are generally much preferred by water voles, however my findings indicate that the larger main drains are generally still occupied by water voles.

Through this project, the success of coir roll revetments to increase riverbank occupancy by water voles has been quantifiably demonstrated for the first time, and previous anecdotal evidence is now supported. As well as the clear benefits to biodiversity, partly due to their ability to reduce erosion, coir roll revetments are an unobtrusive feature, which may appeal to stakeholders wishing to install a riparian revetment with high aesthetic value. As a result, I strongly recommend the use of coir roll revetments by water management bodies, conservation organisations, and developers looking to encourage water voles or mitigate the negative effects of human development such as river diversions.

My advice to anyone installing coir roll revetments is to incorporate pre-established lesser pond sedge plugs (as a valuable food source) within the coir, and where possible, allow the vegetation to grow to at least 30 cm, especially at the water's edge (to provide shelter).

Glossary

• Berm: a flat area of earth bordering a man-made waterway. A berm can be as narrow as 30 cm or large enough to allow tract machinery to carry out waterway maintenance.
• Coir: a fibre made from the outer husk of a coconut; its durability makes it an ideal material for making anti-erosion matting and ropes. Coir is also incorporated in potting compost, since it provides structure and protection to growing plant roots.
• Crepuscular: describes an animal which is predominantly active in the evening after sunset.
• Drainage districts: low-lying areas of land which are administered to by operating authorities known as internal drainage boards (IDBs) to manage flood risk and land drainage.
• Fens/Fenland: the flat low-lying areas of eastern England; these include Cambridgeshire, Lincolnshire, and Norfolk, which were formerly marshland but have largely been drained for agricultural use since the 17th century.
• Invasive: describes a species which is not native to the location in question, and that has a tendency to spread its range quickly, undesirably or harmfully.
• Kalahari: a vast desert located in southern Africa, including the majority of Botswana and parts of South Africa and Namibia.
• Latrine: An aggregation of >6 droppings which have been flattened by the ‘drumming’ of water voles’ hind feet during scent marking. Latrines differ from droppings in that they are indicative of territorial breeding individuals at occupied sites.
• Middle Level Commissioners: The statutory authority that works "to protect life and property from flooding", in the "central and larger section of the Great Level of the Fens", a drained area of East Anglia".
• R: a statistical software in which occupancy analysis can be undertaken using the package unmarked.
• Revetment: man-made structures placed on banks or cliffs in order to absorb the energy of incoming water, thus reducing and/or preventing erosion.
• Riparian: relating to or situated on the banks of a river.