Follow our progress running the nursery, watching wild bees, keeping honey bees and creating our own bee-haven in south Oxfordshire

Our compost - always peat-free


We always use peat-free compost, which we do not find causes any issues, but since we installed to our new posh flood-bench irrigation system last year the compost we were using tended to hold too much water and something more free-draining would be better. When we were originally choosing our compost we attended the Four Oaks commercial growers show as a good way of speaking to the suppliers and seeing their products, and from there identified two worth testing. 

We chose Petersfields peat-free supreme and have since changed to thier peat-free, high coir version as we found we wanted the extra drainage.

Its proved to be excellent as it is open enough for quick rooting. The low loam content does not seem to leak out of the holes at the bottom, which is important as we water from below in flood benches.

Its also a consistant product so well done to all those at Petersfields/Hewitts Turf.


Plant trials: garden centre annuals


It is well known that most commercially sold annuals are not very good for bees but I wanted to see just how true this was so this year we decided to plants them and observe how the bees reacted. We created two beds each 2x3 meters and planted one with a selection of standard garden centre annuals and the other with an annual seed mix supplied by Flowerscapes, a company that know their stuff about pollinating insects.

The 'six pack' annuals we tested were lobelia, stocks, pelargoniums, petunias, verbena and salvia.

We then observed the beds about once a week, on a dry bright day, and simply counted the number of bees, an other pollinating insects we saw. The results were dramatic.

The only plants from the six pack annuals that we ever saw any bees on where the stocks and the lobelia but out of those the lobelia was the clear winner. Typically we say one bee, usually a bumblebee, somewhere within the 12 plants we grew.

The seed mix did better, the average sighting was 2.3 pollinators in the 6 square meter area; often these were butterflies and hoverflies with some bumblebees.

From a gardening perspective, both beds provided a good continuous splash of colour but the annual seed mix established and flourished with only some initial watering where the garden centre annuals needed regular (once a week) watering or they stopped flowering.

Personally I greatly preferred the visual effect of the seed mix with its mixed pallet of very pretty little jewels and think this is a really beautiful way to fill any temporary spare spaces in a garden, perhaps between newly planted perennials while they grow. I should be noted that neither area in this trial attracted as many bees as either a range of our perennial plugs or a solid patch of phacelia grown from seed, either of which recorded been numbers 5 to 10 times greater.



Learning to count bees


Today I attended a workshop at the Laboritory of Apiculture and Social Insects - LASI to thier friends or anyone looking for snappier title - at Sussex University. The workshop focused on garden plants for bees and included practical sessions to learn their technique for assessing how attractive plants are to various different bees. Professor Francis Reitniks put it very simply; bees are very sophisticated at assessing plants value for nectar and pollen so the easiest way to assess the plants is to count bees.  This is exactly what I have been trying to do at Rosybee but with a much more disciplined process - fab, and just what I needed to learn.

LASI has a trial garden set up where each of 32 types of plants (selected based on anecdotal evidence for lack of anything else) are each planted in a 1 square meter patch to make the counting easier. We spent a happy hour counting, cataloging, and learning (see picture)

It was also really good to meet representatives from stately-homes, local councils, beekeeping and major horticultural organisations such as Wisley and Kew, all coming together with a common goal.


Nectar rich seed trials: year 2 - June


The plan is to run the seed trials for 3 years and to see how the flowering behaves over time.  This will be our second year and I have just finished the very minimal amounts of maintenance that I think give the mixes a fair shot at success. As you can see there are some flowers now but also a lot of weeds. All the mixes we are trialling are sold by their various suppliers with claims that the flowering should last multiple years; some are mixes of annuals that are meant to self-seed to renew and others have perennials that should flower every year.

The maintenance has involved the following:

* we let the sheep graze all the mixes in September; a bit later than the suppliers recommend but definitely should have allowed any seed to drop

* the worst of the invasive weeds - patches of solid dock and areas of 5ft thistles - have been removed with herbicide

* we strimmed any tall tough stems the sheep but left the strimmings on the ground

* a few large bare patches have been reseeded

In addition I have now sown a new patch of the 'French mix' in an area near the road that was grass last year. I collected some of the seed by hand to use for this and have done this for purely aesthetic reasons to have a reliable splash of colour at the entrance but it will also be interesting to see how the collected seed works.

Having had a close inspection at the weekend have a few observations:

1, I can see that the 20% wildflower mix has now established quite a few perennials amongst the grass but oxeye daisy and forget-me-not and a few campion appear to be the dominant flower species. (that is what the picture shows)

2, The bumblebee mix area has been very swampy during the winter and has had several weeks of being underwater so I am not sure it will survive. This was not something I was expecting to test but unforeseen things will happen. The results is grass, with the dock and nettles coming through strongly but no sign at all of any self-seeded borage or phacelia

3, the French mix was completely bare dry ground until very recently but now there are signs of a few calendula, cornflowers (see picture - right), oh and the ubiquitous nettles. We resowed one section with seed collected towards the end of last years flowering and this looks like it has a wider range of seedlings coming through but probably not the same ratios of the original mix.

4, the DEFRA recommended mix looks like pure grass at the moment with about 10% dock

The dock and nettles are going to continue to be a problem and I am not sure if we should leave them or remove what we can; torn between letting nature do its thing and wanting it to look nice.

I will update again when everything is in full flower and hopefully also has bees to report!


Bee plants for acid soil


I needed to research which of my plants would be ok in acid soil for one of my customers and it seemed daft not to document the findings. In addition to rosybee plants, the top acid soil plant for bees has to be heather. Beware of rhododendrons as, even though the bees are quite happy on them, the resulting honey is toxic. Try buddleia as an alternative if you want a large shrub.

The following plants will be tolerant of acid soil up to a ph of about 6.

  • Agastache
  • Aster (actually prefer slightly acid soil)
  • Echium
  • Geraniums
  • Helenium
  • Malva
  • Papaver (poppy)
  • Salvia
  • Teucrium

(thanks to for the graphic)





Plants for bees; Doronicum - a march-flowering yellow daisy for b


At this time of year the bees are just beginning to emerge more regularly and very soon they will find much of their food from trees and shrubs. Most perennials and particularly most of the daisy family tend to flower later. Doronicum is a significant exception flowering as soon as the temperatures begin to head above 10 degrees. My plants are just showing buds now. Honeybees love all types of daisy-shaped flowers as they are very open and accessible. They also tend to flower for weeks opening up new nectaries on a daily basis until the centre of the flower has a fuzzy appearance.

Dornonicums are  meant to like damp conditions but I have found they also cope well in quite light soils.

They clump up relaibly but are not invasive; all in all a very well behaved but tough garden plant.

Bee plants for Denman College (WI HQ)

Rosybee is located in south Oxfordshire and not far from Denman College which is the head office  and training college of the Womens Institure.  The WI have been great supporters of the 'bee cause' since they launched thier 'SoS for Bees' campaign at thier 2009 AGM. Based on this I thought it would make sense to and see if we could share ideas. Its been a very busy year but in October I finally managed to find time to contact their head gardener, John Towler, to offer some of my seasons-end stock.  John is a real enthusiast about wildlife gardening and maintains Denmans garden with an inspired balance of providing masses of colour for the many visitors and sensitivity to a birds, beasts and insects. They even have a few bee hives tucked in under a group of fruit trees.

I was very kindly given a tour of the 14 acre garden; the long border is a joy for bee-lovers and I began to feel that my extra plants resembled the proverbial 'coals to Newcastle' but John is on a tight budget so I think my extra stock will find a good home.

John is particularly proud of his herb garden that has a sunny spot with tall hedging protecting it from the north to create a really sheltered corner. It comprises a set of lavender edged beds brimming with mature herbs that John has allowed to flower for pollen rather than keep tidy for the kitchen.

I have agreed to keep in touch with John and supply plants when I can as its such a wonderful

Based on this visit I have been inspired to work out a 'bee-friendly garden audit' that will allow me (or anyone else interested) to objectively assess to what extent a garden provides all round support for our buzzy friends. I am still working on the details but so far I think that Denman College will score a whapping 9 out of 10.


Nectar rich seed mix trials - first year results

In spring this year we planted 5 different seed mixes which are marketed as being good for pollinating insects: 1. an 80% grass, 20% wildflower commercial mix to the DEFRA specification advised for field margin Environment Scheme

2. a ‘bumblebee’ mix of 100% flowers including borage and phacelia

3. a more expensive 30% wildflower mix with a much wider range of perennial wildflowers

4. a mix that is widely used and recommended in France

5. a ‘nectar and pollen’ 100% legume mix with lots of clovers - this was planted on a different and poorer soil to the others

Now in September I can report the first years results. As you would expect from the range above the results are very mixed.  Weeds that we failed to fully irradicate first have been the main challenge which is very common story with seed mixes.  We had sprayed the area with a general weedkiller the previous autumn so our perennial weed has not been too bad except for the nettles; our challenge has been germination of both annual and perennial weed from seeds we exposed when we ploughed in spring. We should have waited longer for that to germinate but then you loose the ideal time to sow the seeds you want. The lesson is to fully prepare the ground in autumn or sow later in early summer after the weed seed has sprouted and been removed.

Flowering results:

a. Growth and weed suppression

Mixes 1, 2 and 4 all grew quickly which reduced the content of other species of plants. The most effective at weeds suppression appeared to be mix 2, the bumblebee mix, as the phacelia grows very fast and then the clovers pack out any remaining space. The wildflower mix has the most weed as even the grass elements are slower to get established. The sellers have deliberately chosen slower growing grass types to avoid the perennial wildflowers being crowded out; the problem is that allows other weed plenty of opportunity to get in there. This patch now looks terrible but we will leave it to see how if fairs over time.

b. Flowering

The mixes with a high content of annual flowers not only won the war against the weeds but also produced the most flower this year. The two with the most flower are the 4, the French mix (pictured above in Sept) and 2, the bumblebee mix.

The least flower was produced by mix 1, the DEFRA spec, which is only now producing a few flowers from 10% the clover content but the overall effect is 99.9% green.

The French mix has been the most dramatic and colourful and has changed every month going from purple phacelia in June, through blue cornflowers in July and on to yellow calendula and coreopsis in August.  However the phacelia, then borage, then clover content of the bumblebee mix has been more muted in shades but a very similar flower density up to August.

c. Bee attraction

Both the Bumblebee and French mixes did very well for insect attraction. In part this was because they both contained phacelia which was clearly the winning plant for be density.  Our method of assessing is the pick a sunny day, at times between 12 and 3 pm and then count bees per square meter.  The full results were as follows:




How quickly do our plug plants grow?

To show how quickly our plug plants grow to fill the space I planted up 10 of each of our stock plants and will take photos to show how they progress. I recommend that our plugs are planted roughly 40cm apart (less for some of the smaller plants) and in 18 months the emerging clumps should halve the bare ground between each plant and when in full glory will be touching.  This is what you want - solid planting but not so crowded that they begin to fight each other for space. This also keeps most of the weeds under control.  I cannot bring myself to leave a trial bed completely unweeded but I find I can get away with just removing the worst weeds in spring and then just pull out any horrors when passing - our plants are tough enough to win through the rest.

So, if the plugs look a little sparse when you first put them in then I suggest adding some annuals (seed is the easiest) just to cover some of the space.  I recommend, phacelia or borage as they will maximise the nectar and pollen supply until the perennials get established. Calendula is also useful as it will fill space quickly and flower for 4 months.

This is how the look in early September - only 3 months later

Result: the will double  or triple in size within 3 months from spring to autum or from autumn to the following summer.


Creating a wildflower meadow - pros and cons

I have to confess feeling a little conflicted on the topic of wildflower meadows; on one hand I totally buy into the romantic notion that it would be brilliant if we could reinstate great swathes of the British countryside with expanses of native grasses dotted with our native wildflowers. However, being a pragmatist to the last I see several major flaws in the plan: - having lost 94% (who counted that?) of our original meadowland, its unlikely we can really find much spare land to reinstate - especially if we need a commercially viable option - food production pressures have never been greater so this is in conflict with the romance - recreating wildflower meadows is not that easy  as they are happiest on poor soils.  Get the conditions wrong and you just end up with a weed patch (see rosybee trial wildflower mix year one!)

If you accept point one above, then we either do what we can or find alternatives. My view is the latter but I guess it depends on your objectives. If you want the pastoral image then you will need to work hard to acquire the right sort of land and then after several years of careful treatments to remove the plants you dont want and cultivate the ones you do then you may get something approximate ...and good luck. By the way, dont believe any seed packets or distributers that imply you just chuck in a seed mix and away you go.

If you want to support wildlife then I recommend you study the foods that the critters you favour need and focus on planting the right things, regardless if the effect is meadowesk.

If you want to try and preserve our native wildflowers then you can do that in a garden setting where they do not need to compete so hard with the grass for survival as they would in a meadow. Or, work to eliminate patches of grass and put in wildflower plugs rather ie avoid ploughing or rotivating a big area and this will throw up all the buried generations of weed seed.

So, overall my recommendation is to work out what matters to you and focus on that.


Nectar rich mix trials; the 'french' mix

On holiday in France last summer we found noticed that as you drive through the countryside, you quite often find sections of field margins or areas at the edge of villages where there is suddenly a blaze of colour.  This is because the French government have been very active in encouraging seed mixes in spare land to provide environmental benefit.  (the French government is also one of those in Europe that have banned certain pesticides after working out they are harmful to the bees as well and the unwanted insects) We tracked the seed mixes down in a standard garden centre where we found a variety of mixes: some just for colour, others for birds, butterflies or pollinating insects.  We bought a big bag of the bee-friendly one and this became one of the mixes we are now trialling at rosybee.

I am writing this in July, having sown the mix this April, and can report:

1. it has grown well just scattered on the surface of the soil

2. it grew better on slightly raised areas of soil and I suspect that the very wet April may have inhibited (rotting) the gemination of the seeds that had landed in the furrows.

3. it contains (in order of flowering from June onwards) phacelia, poppy, cornflowers, echium, wild carrot, camomile, corn cockle and calendula.

4. bee numbers in June- July range from 3 to 5 per square meter in warm, dry weather conditions during day-time, which makes this a highly successful mix for bees.

5. the bees almost ignore the cornflowers and are mainly attracted to the phacelia and echium (not a big finding there!)

6. it is pretty!!! (I know, not a very scientific comment but I am sure most people we appreciate the benefit)

I will continue to monitor as the phacelia is now almost over and it will be interesting the track the bee-count.


Nectar rich seed mix trials: 'bumblebee mix'

Right now in June of our first year of these trials, the bumblebee mix is a clear winner.  This is a mix from Bright Seeds which contains: - phacelia - borage - red clover - birdsfoot trefoil

From the look of both the seed and the results the above order reflects the ratios as it is really dominated by the phacelia with the borage coming through and odd bits of clover peeking out in places where the floods earlier in the year stopped the phacelia germinating. I cant see any sign of the birdsfoot trefoil but will give it time.

The phacelia is great as it grows so fast and makes that whole area look lush and established compared to the other seed mix trial plots where they are still just struggling to make the ground look green rather than brown (except where the nettles and dock have pushed back up!!)

Brights claim that this mix will last for 4 years with the phacelia and borage self-seeding. My guess is that will depend on how much of the indigenous weed returns. We will track it each year and record the results.

Interestingly I put a patch of the same seed mix in a newly ploughed section of the main field where the soil has a much higher clay content and the plants are not nearly as high or lush. It appears that soil type will also be a factor. These two pictures where both taken on the same day in early June and you can easily see the difference.

Nectar rich seed mix trials at rosybee

At rosybee we are undertaking research as well as providing plug-plants.  Over the last few weeks we have been ploughing up sections of the site and preparing the ground for sowing a range of different commercially available seed mixes.  The picture here, shows the bumblebee mix, lush and green in early June just before flowering. Our preparation was to plough, then harrow sections of our field - that has been grazing land for several decades - and then wait for the surface to get 'crumbly'. Ideally I would have liked to allow for the seed-based weeds to germinate during March it was too dry for any wees to emerge. We sowed in early April when the forecast was for rain.  The area we have prepared is approximately two and a half acres which is too small for tractor sowing but too big to rake over by hand so, with the help of my daughter, we spend the weekend sowing onto the surface.  We found a very handy  'spreader' with a winding handle that smoothly distributes the seed in a circle as you walk along and covers about a 3 meter wide area - very useful, but still quite a lot of walking up an down.

We have sown seed mixes with the aim of

  1. observe which mix the bees like best (and which type of bees) and
  2. over the next 3 years, how do the mixes last over time

The mixes are

  • an 80% grass, 20% wildflower commercial mix to the Defra specification advised for field margin Environment Schemes
  • a 'bumblebee' mix of 100% flowers including borage and phacelia
  • a more expensive 30% wildflower mix with a much wider range of perennial wildflowers
  • a mix that is widely used and recommended in France
  • a 'nectar and pollen' 100% legume mix with lots of clovers
  • a solid strip of borage
  • a solid strip of phacelia

All of these will be compared with a large area (approx 30% of an acre) of our plug plants.

Most of the mixes germinated  quite quickly when the rains came but the borage only started to show (pictured right) now that it has warmed up again.

The phacelia in the bumblebee mix has grown the most but is much more advanced in the more loamy soil near our access way than in the heavy clay soil in the main field.  We can also see a certain amount of dock and cow parsley coming up too but that was inevitable and we will just have to live with that.

We now need them all to flower so that we can count the bees as the main focus of the research.



Planting for bees in our village

One house in our village has decided to take out the grass in front of their house and plant a nectar rich meadow area instead.  Penny and Dudley are well known in the village and actively involved in a number of environment related activities: Dudley is a biologist by training and now, retired, has become the local wildlife expert; Penny is a passionate gardener. Together, their vision is to try and encourage the entire village to make more environmental use of any patches of spare land to benefit pollinating insects.  Fabulous! The picture shows how thier front garden looks in May, about 8 weeks after planting this seed mix and a few of my plug plants to give it a boost.  The dry weather in March delayed germination and then the grim April delayed growth but it is now coming on well as you can tell be the green fuzz effect.

In parallel with their activity we have also been planting a range of nectar rich seed mixes at Rosybee which I will be tracking and comparing.

Anyway, "well done" to Penny and Dudley.

Site prep for bee friendly planting at Rosybee

Most of the infrastructure we need for nursery is now in place so it is time to start developing the site itself. Our vision is to fill four acres with all types of bee-friendly plants, shrubs and trees. At the heart of this will be the research plots where we will test out some of the best selling 'nectar rich' seed mixes and compare them with plug grown perennials.

The first step has been to get the plough in to break the turf in 3 big sections.  Our soil is very heavy compacted clay so the ploughing has raised it up like giant, 60cm high, butter  curls. (see the big ridges at the edge of the picture above) In this condition its hard to imagine how it will ever be a workable texture that I can so seed into. But they assure me that once it has dried out a little then next stage, harrowing, will break it down  much more. 

In the mean time our daughter thinks is a new adventure playground!


I wrote the above text before we had got round to taking a picture and 2 hours later I find that they have already been back with the harrier!! Speedy.  So here is how the 'avenue' looks all nice and freshly tilled; ready to sow. This area will be planted with 3 different commercial seed mixes so should look pretty until the grass begins to dominate.


Bee-plant trial plot - February - coming back to life

It so exciting to see the the new growth pushing its way up and to know that the plants have made it successfully through another winter. The first trial bee plot is now entering its third season and so the perennials as well established (as are a few weeds) and I can also see how the self-seeding annual plants are working. Both the borage and the mysotis have self seeded well and seem to work fine together covering early and late season. The Chieranthus Chieri was three years old last season and getting a bit woody but produces a carpet of seedlings which are now growing on beutifully and I expect will be flowering by April.

The only plant which does not seem to have seeded is the biennial echium vulgare. I think this is becuase it was planted in the shade of some taller plants and did not get enough direct sun. I will restock with some spare plugs in a sunnier area.


Geology in our field - Vale of White Horse

In the process of doing the groundworks and laying trenches for water/power connections we have dug up quite a few areas of our field with some interesting results. We first noticed the interesting alluvial layers when we dug the 200m long services trench; we found we a layer of clay on top but under that it is mainly a one meter layer of ballast made up of small mainly chalk pebbles, sand, grit and some clay. Below that we hit the thick blue clay 'Kimmeridge' that made this area so attractive to Thames Water for a reservior (they would have had a natural waterproof base).  We also hit spots of more dense clay and sometimes sandy stuff.  Clear signs of the ancient riverbed that is now the Thames valley.  This picture shows the layers in the hole we dug for the sump.

You can just see grass at the top then 30cm of light clay, then pale ballast, the dark clay below. The red mark in the centre is where we have cut through a nineteenth century land drain which indicates they has issues with the site too.

This study has been made even more interesting by trying to work out how porous - or not- the topsoil layer is and to predict how we can manage rainwater or risk of rising water table. Our entire village is in a flood plain area and this field anecdotally suffers from standing water as times during the winter.

There are two theories about the cause of this, each would require a different strategy to manage:

1, the light clay layer just below the surface absorbs so little water that rain sits in puddles till it evaporates

2, the ballast layer underneath gets saturated above  heavy Kimmeridge clay and has nowhere to go so is forced to the surface

Next to us, a housing developer is trying to work this out as part of thier planning process. They have employed geologists and all the neighbours have developed theories so its become a hot and contentious subject. Establishibng the facts has been made more difficult by the near-drought conditions this year which have resulted in there being very little ground water to observe.

We favour theory 1 as it seems to reflect what we witness; when it rains the water just sits there; where we have dug down into the Kimmerigde clay, water trickles slowly into the hole from the base of the ballast layer.  On this basis, to reduce standing water on our turning circle we have drilled large holes through the surface clay and back-filled with ballast to act as a series of small soak-aways.

For the polytunnel foundation we need to a, avoid floodwater coming in from the surrounding field and b, make sure that irrigation water has somewhere to go.  So, having removed only a very little topsoil to achieve a level surface we have added a 20cm layer of more ballast and will add drainage channels down each side.

We really have no way of knowing if this will be sufficient and will just have to wait for the rain. Who knew that so much geology would be required for horticulture.



Bee plants - the trial plot in July/August

Its been so dry for the last few months and I had intended not to water the bee plant trial plot at all, leaving it to fend with the elements as it would in a field margin.  This has proved to be too tough a test as the soil is also very light and sandy so any moisture dries out very fast.  So I have to confess that I have provided some water but not a lot; about 6 watering cans in total. Just enough to save the plant but not enough to really encourage much production of flowers with nectar.  Depressing as a test but educational. The plants that have done well and provided some flower through this period and

  • echium vulgare
  • borage
  • oxeye daisy
  • malva moschiata
  • althaea officianalis (marsh mallow)
  • agastache foeniculum

And now the solidago (golden rod) is coming into flower.

What has also been very interesting is to observe that the bees pay no interest at all to some of the plants that are recommended for bees. This included the achilleas.Sadly the monarda (beebalm) flowered so poorly with the draught that the bees really didnt have much chance to show thier interest. This plant is one that will need more moisture to thrive.

Bee plants - the trial plot in June

The idea of the bee-plot is the design a planting scheme that will ensure continuous flow of nectar and pollen throughout the season. So naturally when choosing the plants for this trial plot I was quite determined to make sure that I had a range that would ensure flowering right through the 'june gap'. Up till the first week in June this was going well but now most of the earlier flowering plants have almost finished and the next batch are still to bloom. It think the very warm spring weather had some bearing on this as it accelerated the flowering of the poppies and geranium pratense.  And now it is much cooler than expected. The borage is just beginning to flower but the agastache and monarda are not quite ready.  The only thing really flowering happily is the oxe-eye daisy which may not self seed when everything else gets established.

I think that there will be lots of  flowers by the end of the month so hopefully this is just a two week lull.  All the same, it shows that the trial was necessary and I now know what to adjust in the design.

Phacelia - good for lots of different bee species

We are finding that this plant is really good for many different types of bees. Here are results of our brief study: Our phacelia plot is about 24 square meters and we first studied the bee activity  in June at midday in full sun for about 20 minutes.  We then continued observing for more types of bee at various other times of day for 5 minutes at a time.

Bee density: approximately 9 bees per square meter when in full sun, 5 per sqm at when overcast - the difference being reduced numbers of honeybees when its cloudy.

Bee diversity: & species observed, most of them repeatedly:  honey bees were the most common (but our hives are only 50 meters away)but we also saw the following:









  • buff-tailed bumblebees
  • red-tailed bumblebees
  • early bumblebees
  • brown banded carder bees
  • common carder bees
  • field cuckoo bumblebees
  • tree bumblebees
  • garden bumblebee (just one)
  • hairy footed flower bee (just one)

Quite a haul!

Observed behaviour:

The honey bees were purely collecting pollen but most of the bumblebees were after the nectar.  The latter have to push their faces past the forest of stamens to get into the small centres to the petals.

It was quite common to see multiple bumblebees in a single flowerhead but the honey bees will not go near a flower that has any other bee on it.

The pollen is dark purple/blue and makes the bees legs look like they have a plaster  on.

Growing phacelia

We  cast the seed on the surface in a rough patch of cleared earth in our field. Then we totally ignored it and left it to fight amongst the native weeds and grass.  It has faired well with this treatment and represents just over half of the vegetation in the plot.