Sustainable Beekeeping with Warré Hives -- David Heaf's Warré Project
Please note: This page is no longer
Abbé Émile Warré (?-1951) experimented with over 350 hives of various types over a period of 50 years. During that time he developed a bee-friendly fixed-comb hive designed for minimal intervention, easy harvesting and enlargement as well as for producing honey at minimal cost of labour and capital. He called his hive la Ruche Populaire, which could be translated as 'the People's Hive'. He describes the origins and development of this hive, how to construct it and how to manage it through the beekeeping year in his book L'Apiculture por Tous. We have translated this book as Beekeeping For All and it is available in print and for free download.
Pros and Cons
A criticism of hives with moveable frames is that the conditions in the colony are far from natural, even if the hive is not repeatedly opened and the frames moved around. A wild colony building in a cavity, or a managed colony building in a skep, attaches comb not only to the top of the cavity, but also to the sides. This creates vertical cul-de-sacs between adjacent combs that retain the heat and the scents (including pheromones and possibly other volatile substances necessary for full colony health) produced by the colony (German: Nestduftwärmebindung). With framed hives, beekeepers strive to keep the sides and tops of the frames away from the hive body walls and from whatever is above, thus allowing heat and scents to escape from the nest. As fast as the bees try to build in gaps round frames, as diligently does the beekeeper try to keep ahead of the bees by scraping away so-called 'brace comb'. The main point of the Warré hive is to make it bee-appropriate or bee-friendly by eliminating the gaps so as to conserve nest heat and scent. The Warré hive is a vertical (tiered, storified) top-bar hive..
Another artifact introduced by modern beekeeping is queen excluders. They are not used in Warré hives. The queen has access, at least in principle, to the whole comb chamber. Having no excluder is made workable by allowing the bees, to extend the combs downwards by nadiring fresh boxes, i.e. adding boxes underneath. The brood nest gradually moves downwards as brood at the top hatches leaving honey-filled comb above. This overcomes another artifact, namely that of forcing the bees to build comb for honey storage above the brood nest, i.e. in supers. As plenty of space is always kept available in the Warré hive below the nest for further building, two 'risk factors' of swarming, i.e. no work for the wax makers and overcrowding, are minimised.
The bees are allowed to build comb freestyle, subject only to the positioning of top-bars and beads or starter strips of wax. There is no need for foundation. Thus the bees determine their own cell size, the proportion of drone to worker cells and where to put them.
Feeding sugar is frowned on by followers of natural beekeeping and, in any case, if sugar is introduced into the Warré hive, some would almost certainly end up in the harvested honey. Thus, if feeding is necessary, diluted honey is used either with an Ashforth feeder or from a jar on the top bars. Sugar should be used only in exceptional circumstances.
Some criticisms that have been made of the Warré hive include (together with our response):
Some advantages of the Warré hive include:
All the information needed to construct and manage Abbé Warré's own hive is given in his book Beekeeping For All, which we offer for free download. His hive, comprising a minimum of three stacked boxes in spring, is designed for opening only once a year at the honey harvest. At the spring visit, boxes are merely inserted under the brood nest and so far as the bees are concerned this is hardly a hive opening as they appear to continue their work undisturbed and no heat is let out. However, with this extremely low frequency of intervention or interference, in Warré's hive, the beekeeper has only the entrance activity to tell him what is going on inside. To provide more information, the hive was modified by Jean-Marie Frères to include a window across the entire back of each box. This modification is described in his book L'Apiculture Ecologique de A a Z, edited and illustrated by Jean-Claude Guillaume (1997). Details of the book and how to order it are at http://groups.google.fr/group/ruche-ecologique/.
Making the the original Warré hive-body box involves a lot less time and cost compared with say a UK 'National' brood box and there is only one type of box to make, which serves for both brood and honey harvest. Furthermore the original Warré box is merely butt-jointed and nailed. The finger jointing on the boxes shown here is unnecessarily robust and was dropped in favour of butt joints in later hives. The modified Warré hive, i.e. with a window, is not as economical as the original and thus runs counter to one of Warré's aims for the 'People's Hive', namely simplicity and cheapness. What is saved in not having to make the rebated, tenoned and bevelled bars of the 'National' brood box is probably cancelled out by the time and cost that goes into the rebates, tenons and glazing of the back of a modified Warré box.
For the first batch of boxes (12) for a trial of the Warré hive concept in Wales, the boxes were made with not only windows, but also interlocking (castellated) joints. All joints are glued with exterior grade woodworking adhesive. These are very time consuming to make. A second batch had half-jointing at the front with two rows of 50 mm galvanized nails at right-angles. Subsequent batches were windowless and butt-jointed with 65 mm galvanized nails, 7 to each joint.
The prototype illustrated here is made almost entirely of driftwood (washed ashore from the Irish Sea), some of which was already painted -- hence the 'castellated' corners. Before being deployed, all outer surfaces were painted with a suitable breathable exterior wood stain. For hot climates a white or light-coloured paint is advisable. Subsequently boxes were made of larch and finished with two coats of raw linseed oil.
Front view Rear view
The pictures above show a complete hive with three comb chambers (hive-body boxes) with handles, a roof, an open-mesh floor and a stand. The last two items are not in Warré's design.
An important addition to the Frères & Guillaume design is having open-mesh floors to monitor Varroa (the sample tray goes in the rear slot just above the stand). The Warré floors most of the time as they are less draughty and the mesh floor temporarily substituted for monitoring Varroa. As can be seen from the rear view image above, another modification of Frères & Guillaume's hive is removal of the right and left battens of the shutter surround. The shutter runs the full width of the back.
The rear view (right) shows the shutters covering the windows of the comb chambers. The roof, this design based on a UK 'National' hive roof, is ventilated on all four sides and covered with aluminium sheeting (for example, from a recycled caravan). The following two images show the quilt (100 mm deep) with the roof removed.
Quilt normal position Quilt underside
The quilt -- an insulating, but vapour permeable box -- is slightly modified in this version in that the insulating material is polystyrene pieces (used in packaging) instead of straw or dried leaves. It is covered top and bottom by thick cotton canvas (closely woven hessian is also suitable). As there are no vents in the quilt, heat loss is minimized, yet vapour and gases can find their way through the two layers of cloth and amongst the intervening fragments of polystyrene. A solid polystyrene sheet would not be suitable here as it would present a significant barrier to moisture and air diffusion. Hives made later have a quilt to Warré specification: sackcloth fixed only underneath, fill of wood shavings (from an electric planer) and open at the top so the insulation can be turned over, and if necessary replaced.
Warré recommends placing directly on the top bars a piece of coarse weave cloth such as jute sacking stiffened with flour paste. The bees propolise or unpropolise the holes in the weave to increase or reduce the amount of top ventilation. J-M Frères & J-C Guillaume suggest using a 'moustiquaire' (mosquito net). The advantage of this is that the cloth can be gently peeled off without upsetting the bees, whereas removing a crown board often involves cracking propolised joints which often agitates the bees.
With the roof, quilt and cloth removed, the top bars are exposed (see above). Internally, the chamber is 300 mm square (floor plan) and 210 mm high. The walls should be at least 20mm thick; here they are 25 mm. Two 10 x 10 mm rebates support the top bars which are 24 x 9 x 320 mm at 36 mm centres, each with a 2 x 2 mm grove down the centre of the underside into which beeswax is poured to form a starter strip (Waxing top-bars). The 1 mm gap above the bars is essential for making removal the box above easier. The bars are held in place with one gimp pins (the same 20 x 1 mm lacquered pins as used for frames) at each end. Latterly, the heads of the pins are cut off before finishing hammering in. This eases removal of top-bars. All boxes have top-bars. The bars are rough sawn underneath to help comb adhesion, but above they are planed and coated with linseed oil to minimise comb adhesion from above.
Above left: inside view of window; above right: window shutter opened
With the top bars of the top chamber removed (above), its window and the top bars of the next chamber down are visible. The window glass is 4 x 300 x 114 mm let into 4 x 7 mm rebates in the back bars and secured with gimp pins. The shutter comprises a piece of polystyrene 21 x 98 x 298 mm nailed to a wooden, 20 mm thick, outer cover. The right-hand image above shows the shutter removed. Note that the comb will align 'cold-way', i.e. with midrib planes at right angles to the front/back. However, if required, for wintering the boxes can be rotated 90 degrees so the comb lies 'warm-way'.
Scale drawing of cross-section of window of modified Warré hive. The top and bottom window frame bars of this version are mortise and tenoned into the side walls.
Above shows a detail of a window. The shutter is retained by two wooden latches, each secured by a stainless steel or brass screw, and mounted on 20 mm thick battens bevelled to shed water.
Roof inside Mesh floor
Original Warré floor design (two pieces of fallen comb have been fixed to this used floor by the bees)
The inside view of the 'National' roof (above, left) shows all four vents. To make a vent, a hole is drilled from the outside, sloping upwards to prevent water seeping in, and emerges in the middle of the spacer battens which create a ventilated void space under the roof. The holes are closed on the inside with perforated zinc or other wasp/bee-proof material. This roof type has proved unproblematic with the Warré hive. No condensation forms under it. In the UK climate, it is customary to place a brick on the roof of each hive. With the Warré hive at exposed sites, particularly when four or more boxes are in place, two bricks are used.
The open-mesh floor (above right) is adapted from a design used by the author for National hive floors. In this case, woven galvanized metal mesh is used with 8 wires to the inch, although usually expanded galvanized steel is used (N6664F12502500, www.expandedmetalcompany.co.uk). The mesh slopes upwards to the rear of the hive and the part of the back bar projecting into the area accessible to bees is bevelled to allow debris to slide out through the mesh to the ground. The same applies to the beading at the edges. This floor is fitted with a 7 x 100 mm entrance. The entrance block can be removed when there is a nectar flow on. The 7 mm aperture is generally regarded as mouse-tight, but the author adds metal mouse guards in October, each with a row of 9 mm circular holes. The type shown in the image below is a modification of Warré's combined mouse guard and robber guard. Here it is made from scrap aluminium angle, but it can easily be cut from a food can.
3-box hive with Warré roof
The illustration above shows a hive with a Warré pattern roof ventilated under the eaves and gable solely to protect the hive from the sun. It contains a board to prevent mouse access to the quilt. This means that in Warré's design there is little or no passage of air or moisture through the top of the quilt. It thus must contain hydrophilic material such as wood shavings etc. so that excess moisture can be absorbed. The quilt probably has a certain amount of moisture buffering properties.
As the footprint of the hive is smaller than many other hive designs, a firm base is required. I use a recycled paving slab and a 300 mm high stand with legs outside the four corners of the hive. But any kind of secure stand is suitable. It must be levelled with a spirit level to ensure that comb is built parallel to the sides. Levelling is greatly facilitated if the stand is on a firm base, such as a concrete slab. The levelling can then be readily completed with pieces of slate or other weatherproof shims. The author prefers stands that fit the hive footprint exactly, allowing easy water runoff, and providing a 100 mm skirt all round as extra draught protection below the Varroa floor, when used. The Varroa floor sample-tray is inserted only when sampling Varroa.
Warré recommends having the hive entrance much closer to the ground so as to make it easy for laden foragers who miss the entrance to climb in. Bricks or blocks would suffice for this.
Principle of Warré hive use
Starting with only two chambers (hive-body boxes), both fitted with top bars in identical alignments, a swarm (2 kg minimum) is introduced in the usual manner, either by running it in up a board or by tipping it into an eke temporarily inserted (until the following morning) between the boxes. A colony may be transferred from a framed brood box by placing it over a Warré box with a suitable adapter board and leaving it until the colony has established in the Warré (see, for instance http://ruche-warre.levillage.org/Christophe Certoux.htm).
Guided by the starter beads in the top bars, the swarm or transferred colony draws comb 'cold-way'. (New comb in a Warré hive-body box) Progress of the colony is viewed through the window in the rear. Eventually the comb reaches down to the upper surfaces of the top bars of the box below and building is resumed below these, again guided by starter beads. When the colony is observed through the window to have developed somewhat in the bottom box, a third is added underneath it. This may present a lifting problem for a single operator, but this can be overcome with a suitable lifting device. (Floor change with lift)
As brood hatches at the top of the nest, the cells vacated are gradually filled with honey. The brood nest moves steadily downwards and eventually comes to occupy only the bottom two boxes. Depending on the nectar flow, further boxes may be added in succession at the bottom. Warré reached a maximum of seven in total. When the flow is over and the brood nest well clear of the upper box(es) the honey boxes are removed one by one after smoking the bees down leaving two boxes -- the upper mostly honey and the lower mostly comb and diminishing brood -- and a minimum of 12 kg honey. The top-bars are scraped clean. The cloth and quilt insulation are renewed and a mouse guard fitted. This is the only hive opening in the strict sense in basic Warré management as it is the only time when the hive atmosphere is let out by the beekeeper.
The comb is cut out and the honey harvested. The box is cleaned and prepared for re-use. Re-use (below the brood nest) can be immediate if in a good season an early harvest is taken and insufficient spare boxes are available. The wax is extracted from the combs in a solar extractor.
We mentioned above the principle of the retention of nest scent and heat (German: Nestduftwärmebindung). This concept is taken from a book by the Austrian beekeeper Johann Thür published in 1946. In it he presents the hive of Abbé Christ which is almost identical in concept to Warré's 'People's Hive'. A translation of Thür is available for free download.
April/May 2007 -- The author's first Warré hives are populated
Two of the author's British 'National' hives were preparing to swarm on 26 April and 5 May 2007, and were artificially swarmed into Warré hives 1 and Warré 2 respectively without searching for the queen and using the following method. The Warré roof and quilt were removed and an aluminium funnel designed to receive bees shaken or brushed from National hive frames was placed on the top box. Bees were shaken from each brood frame of the hive to be swarmed and when clean of bees the frame was placed in a fresh brood box. Bees on any frames with queen cells with larvae were brushed, not shaken, into the Warré hive. When all 11 frames were clear of bees, the funnel was shaken free of bees, the bees on the top were smoked down and a Warré-to-National adapter board was placed on the top Warré box. A queen excluder was placed on this followed by the National box containing brood, its super and the crown board. Many bees took to the air and shortly after this operation almost the entire front of the Warré was covered in bees. About 10 minutes later it was as shown in the photo below.
The brood in the National hive was left on the Warré for two hours to repopulate the combs with nurse bees. It was then moved onto its own stand a metre from the Warré and facing the same way. The Warré was given a feed in a Warré-sized Ashforth-type feeder placed on the brood box. The feed was 900 g honey (from the same apiary) diluted 2 parts honey to 1 part water by weight. The quilt and roof were placed on the feeder which was removed when empty. Six days later the brood in the National was moved in the early afternoon to another site in the apiary leaving the flying bees to strengthen the Warré.
By 7 May 2007, 10 days later, the top box of Warré 1 viewed through its window was as shown in the following photo:
Six combs have been drawn wide enough to be touching the window glass and have been extended downwards to or below the window sill, i.e. at least 160 mm from the top bars. A seventh comb is occasionally visible as the bees move around, and the eighth, the smallest, is obscured by bees.
By 19 May 2007, seven combs of Warré 1 top box were developed to their full widths against the window, down to about 25 mm from the sill. Most of the end cells were filled with honey and cells further in were capped. In box 2, five combs were visible, the longest extending half way to the floor of its chamber. For about a fortnight, the weather had been cool, wet, often overcast and windy. In Warré 2, seven combs extended to below the window sill and the eighth, still quite narrow and thin, was extended to about 100 mm below the top bars. A slightly slower initial development of Warré 2 compared with Warré 1 is probably attributable to the poorer weather since Warré 2 was populated.
The Warré hives were visited by the author and Jean-François Dardenne (http://ruche-warre.levillage.org) who advised removing the third (bottom) brood box until the brood had developed to near the bottom of the second box. The rationale is that the bees have less distance to climb after entering the hive. This was done for both Warré hives.
Warré hive 1 is a strong colony occupying three boxes and foraging well. Warré 2, after a good start and drawing one and three quarters boxes of comb, greatly slowed down. By 14 July Warré 2 had seven advanced queen cells, one sealed, in the top box close to the window. On 19 July the cells were hatched or opened at the side and new queen was seen stinging the side of an unhatched cell, helped by workers. This hive superseded. Foraging from it was less than half that of Warré 1 by early August. A further four Warré hives were populated. The details are summarised in the following table:
Hives 1, 4, 5 and 6 are still drawing comb. It remains to be seen whether they will build up comb and stores to winter well. Warré in Beekeeping For All recommends that the colony should winter on two boxes of drawn comb and 12 kg stores. In mid August the important nectar flows that remain in the immediate locality are Himalayan balsam (Impatiens glandulifera), Japanese knotweed (Reynoutria japonica) and ivy (Hedera helix). The latter provides nectar and pollen into early November.
Warré reports that a healthy 2 kg swarm will fill two boxes with comb in 15-20 days, if hived at the start of the main nectar flow. Hive 1 did not receive its third box until 24 May, i.e. a week later than expected, by which time the second was almost full. Development seemed relatively rapid in the warm weather at the end of April and early May but since then it has been relatively wet and cool with rain on 50 days to date. This has more severely affected the hives populated since April. The experiment will not be complete until the colonies have overwintered and gone through at least one full season.
2 October 2007
All hives were fitted with mouse guards and hefted in preparation for wintering.
Hefting was done by weighing each side of the boxes containing comb using a 60-pound spring-balance suspended from a sash cramp (weighing device). The cramp screw was turned so that the box, just lifted off its seating. The weights from each side were combined. Boxes to be hefted were freed with a hive tool and if necessary rotated slightly in both directions to break any comb bridges to the box below. The stores were estimated by subtracting weights of the box, the top bars, the comb and an estimate for the bees. The results are shown in the following table.
Warré recommends leaving 12 kg. stores for overwintering. Only two hives approximated to this, namely 1 and 4. However, hive 3 and 6 had over 18 pounds, which is more than the amount on which several of the author's National (framed) hives overwintered successfully from 2006 to 2007. Furthermore, in the week since hefting, the ivy (Hedera helix) flow was vigorous and even Himalayan balsam (Impatiens glandulifera) was still supplying nectar. The author's National hives generally gain weight in October. Hives 2 and 5 almost certainly have too low a weight of stores to survive the winter. It was hoped that hive 1, with comb filling nearly 3 boxes would yield honey for harvest. However, when this box was removed, the top edge of the brood nest extended into two of the combs. It was therefore left intact. Later calculation also confirmed that the total stores were anyway below the amount that Warré recommended for overwintering.
It is interesting to note that the colony that performed best, number 4, was a natural swarm. It put on more weight than the first colony which was hived 6 weeks earlier. Also, the last colony to be hived, on 11 July, has done relatively well too. This was also a natural swarm.
The author has abandoned swarm control in favour of catching swarms. A high proportion are caught in bait hives placed not far from the apiary and transferred on the evening of arrival to a Warré hive. (see 'Hiving a swarm from a bait hive into a Warré')
11 February 2008
Beekeeping For All
Details of and how to order the printed edition of the book Beekeeping For
All by Abbé Émile Warré from Northern Bee Books:
Direct link to Northern Bee Books online sales page for Beekeeping For All
by Abbé Émile Warré:
Download for free the e-book (PDF) Beekeeping For All by Abbé Warré.
English Warré beekeeping web portal:
The author would be interested in corresponding with anyone about the principles, design, populating and management of Warré hives. Please contact David Heaf at:
david (at) dheaf (dot) plus (dot) com [please reconstruct this antispam e-address].
Please note: This page is no longer