Sunday, 26 January 2020

Crepis mollis. Northern Hawksbeard

Crepis mollis ( Northern Hawksbeard ) and adjacent Hawkweeds. 
Notes and photos from Mid July 2019. A visit to Allendale, Northumberland.
Crepis mollis with phyllaries highlighted.
Over the last few years I have been attempting to learn to identify all the 'yellow dandelion' type flowers that I could find in Cambridgeshire. This has excluded the dandelions and hawkweeds which are way too difficult, with their hundreds of apomictic microspecies. One species that I was never going to see in Cambridgeshire was Marsh Hawksbeard as the nearest site is in the Peak District. A trip to the Peak District failed to find any.  With a planned trip to visit my daughter who had moved to Newcastle, I decided to track down not only Marsh Hawksbeard ( Crepis paludosa) but also the much less common Northern Hawksbeard ( Crepis mollis) which  has its main distribution in Northern England.

 As mentioned in my previous blog on Smooth, Rough and Beaked Hawksbeards, there is a really excellent article on the identification of Marsh, Northern and associated Hawkweeds on the Cumbria Botany blog written by 'JR'.  This article also covers the potential confusion with the Hieracium vulgatum, a Hawkweed that often grows with Northern Hawksbeard. Without this article written by Jeremy Roberts and supported with photos by Phil Brown, I think I might have struggled. Both these botanists are BSBI recorders for Cumberland with years of experience, so it is interesting to see what they found as good ID features and what they left out.   REF

The second advantage was that I had found several articles on the occurrence of Northern Hawksbeard, written in the South Northumberland Newsletter by BSBI recorder for vc67, John Richards. An e-mail requesting good locations to John Richards, provided two sites and this was much appreciated. I was also lucky in that 2019 was a good year for Northern Hawksbeard. See article :-

So with a good ID article and good site locations I headed north to see and photograph two new species. This blog covers Northern Hawksbeard and the nearby Hawkweed which could be confusing unless you really look at the differences in detail.

The Cumbria Botany article picks out leaf shape as a key difference to separate C. mollis from Hieracium and also secondary features like colour of the pappus and achene shape.

It is interesting they don't really look at the involucral bracts (phyllaries), which are supposed to be the best differentiator between Crepis and Hieracium , ref Sell Vol 4 Page 202.  Crepis are supposed to have a long and a short row of involucral bracts ( phyllaries) which form a cup and saucer. The key here, is two distinct rows. The species Smooth, Rough and Beaked, as covered in my previous blog, all show this feature well but I found that Marsh and Northern are not very good at following this rule.
 Possibly this is due to both Marsh and Northern being sectioned off into different grouping by Sell, outside the section 'Crepis' that includes Rough, Smooth and Beaked etc. Although all have a scientific name starting with 'Crepis' apparently large groups, like Crepis (which has 200 species worldwide) are often split into sub-groups.

 Stace is using the term 'phyllaries' instead of 'involucral bracts' so from now, I am changing over to this term as it is more precise, rather than use just the word 'bracts'.

Crepis paludosa with phyllaries highlighted.
Crepis mollis with phyllaries highlighted.

Hawkweed at Allendale Town

The main point is that the crepis species here,
does not clearly have two rows of phyllaries, with the outer row being adpressed and of different lengths.
As a main differentiator, this feature is a failure as far as I am concerned. See highlighted phyllaries in photos.

Both Crepis paludosa and Crepis mollis have phyllaries with prominent dark hairs with glandular tips. The hawkweed phyllaries do not have obvious glandular hairs however if you look very closely, a few can be found.  Marsh Hawksbeard sometimes can lack the glandular tips  on these hairs so it's not an absolute ID feature. The Northern Hawksbeard according to Sell, always has glandular hairs on the phyllaries.
The dark hairs on the Marsh Hawksbeard (C. paludosa) are very long and definitely look for dangerous from an insect's point of view!.

My first Northern Hawksbeard.

C. mollis near Allendale Town. 17th July 2019
1) Basal Leaves.
C. mollis lower leaf, not the best example of impressed veining.

C. mollis  lowerl leaf underside paler and hairy. 

The lower basal leaf shape is elliptical with a blunt end, entire in that it has no protrusions or lobes on the margin unlike the hieracium which has toothed edges. This particular leaf above is not showing the impressed vein pattern that the Cumbria blog suggest is a typical feature, but this could be down to variation and lighting conditions. The next photo below shows this feature a bit better.

C. mollis, lower leaf showing leaf surface impressed veining.
Note the pale midrib and the leaf shape narrowing down to a narrowly-winged petiole. Next compare with a Hieracium leaf from a plant growing 15m up the road from the C. mollis.

Hieracium basal leaf

Main difference is the toothed margin and the way the leaf does not taper down to a winged petiole.
The teeth may be more or less developed.

Hieracium lower leaf underside, paler and hairy.
2) Mid-stem leaves.
While on the subject of leaves , next up are mid stem leaves.
C. Mollis, mid stem leaf.

C. mollis, mid stem leaf with Semi-amplexicaul base
The mid stem leaves partly clasp the stem, no leaf stalk. This is called a Semi-amplexicaul base.
Note the tiny red protrusions which are hydathodes which are spaced along the leaf margin and can give a slight toothed edge to the margin but not to the extent of the Hieracium toothed edges.

Hieracium mid-stem leaf, toothed margin prominent.
Hieracium Upper leaf showing no clasping to the stem. 

Hieracium mid stem leaf showing semi-amplexicaul clasping of the stem.
As usual just when you think you have found a reliable feature to separate these species you find an example that makes it more complicated. This photo above is a Hieracium with a mid stem leaf showing semi-amplexicaul clasping of the stem. This is not really surprising as the Hieracium group is split into several hundred variants all of which have slightly different characteristics, plus there is variation as well within a microspecies. Nearly all have the toothed leaf margins to a greater or lesser extent (but not absolutely all have this feature).

In conclusion, the C. mollis have leaves that are without the teeth of the Hieracium, have a narrowly-winged petiole on the basal leaves and can have impressed veining.

3) Pappus and achenes.
 Next feature mentioned in the Cumbria blog is that C. mollis has very pure white pappus hairs.

C. mollis pappus hairs bright white.
Hieracium at Allendale Town.
These photos don't quite show the difference but you can just see the left flower head on the hieracium is not the pure shinning white that C. mollis displays. In real life there was a difference if you could directly compare the two species.

C.mollis showing pure shinning white pappus hairs.
C. mollis achenes, 
The achenes are brown, banana shaped in that they are curved and taper down at each end.
No sign of beaking and with tiny ribbing. I can just about count 8 on the visible side so at least 16 ribs.

Hieracium achenes, dark brown and pappus hairs not pure white.

These Hieracium achenes are typical of all the Hieracium I have ever seen, not that I am very familiar with this complex group. Typically they are very dark and only slightly curved. Main feature is that the abrupt pappus end is not tapered at all. The ribbing is uneven and I counted ten ribs only.

Conclusion. The Cumbria blog is very keen on the achene difference as being totally distinctive in terms of achene shape . Interestingly they show Hieracium achenes that are brown in colour, so achene colour may not be as useful as I thought, having believed all Hieracium had almost black achenes. With achene shape and the leaf shape conforming to the Cumbria blog, I left happy that I had seen Northern Hawksbeard plus an interesting but similar Hawkweed.

4) Flowers.

C. mollis, showing two pronged stigma and styles are not pure yellow- more greenish.
C. mollis showing the stigma but very little of the style as it's enclosed in the stamen/anther.
The first photo shows the style well extended from the anther structure but in the second close up the anther structure almost comes up to the stigma.  The anther structure has five red stripes which is a feature shared with Marsh Hawksbeard Crepis paludosa on some flowers. I assume the style may grow in length as the flower develops. Does anybody know?

C. mollis, phyllaries with glandular hairs.

Hieracium Flower showing stigma slightly darker than petals.

Hieracium from Sinderhope site, with anther structure yellow and slightly darker stigma.

Hieracium from Sinderhope site.

The hawkweeds seen all had two or three stem leaves. Leaves had thin simple hairs on the margins and lower surface, a few simple hairs on the upper surface. I did not see any stellate hairs on the leaves.
Detail of hairs on phyllaries showing long black (but white tipped) non glandular hairs plus a few short glandular hairs.
Glandular hairs are hard to see, outnumbered by the long eglandular hairs but stellate hairs are numerous.   I think this conforms with the suggestion in the Cumbria blog, that this is a Hieracium vulgatum , Common Hawkweed.   However the hawkweeds did look slightly different between Allendale and Sinderhope so more help would be needed. Maybe next season but probably not, as the Hawkweeds are the most difficult group, with the possible exception of the other apomictic group, the dandelions. I have lots of much easier flowers to learn first.

Shining white pappus hairs
on C. Mollis
Off white pappus hairs on Hawkweed

The End
24th January 2020
Peter G Leonard.

Thursday, 14 November 2019

Cranesbills with long tips to sepals

Geranium rotundifolium  Round-leaved Cranesbill   
Geranium columbium       Long-stalked Cranesbill    
Geranium dissectum         Cut-leaved Cranesbill       

These are small flowered cranesbills which have a long tip to the sepals, unlike Dovesfoot Cranesbill  (G. molle ) and Small-flowered Cranesbill (G. pusillum) which were covered in a previous blog.

These are additional photos and notes as backup to the ID Key Features blog of August 2019.

1)  Geranium rotundifolium  Round-leaved Cranesbill

14Jun2019 Cambridge North Station
The above photo shows several key features, the pink petals have a pale white base which gives the flower a pale centre. The petals can occasionally have slight notches. The pale centre is the standout feature as otherwise it is easy to walk past thinking its just another Dovesfoot Cranesbill, which is much more common.

The leaf lobes are cut up to a half way and the shape gives the plant its common name 'Round-leaved'.
The leaves are variable in shape having from  5-9 lobes ( three at the apex). The apex leaves at the top of the plant can sometimes look very different to the rounded shape shown in the ID Key. The 'cut up to a half ' is a consistent feature.
Lower leaf as used on ID Key Feature Blog with 5/6 lobes.

The photo of the flower below shows the sepals have quite a thick point extension and it is just possible to make out the red-tipped glandular hairs.  The sepal extension is a variable feature, occasionally some plants only have a tiny extension.

The most important secondary feature is shown above and is the red-tipped glandular hairs which stick out from the sepals, flower stem, leaf stem and leaf edges.  There are in addition non glandular hairs. No other small flowered geranium has these red-tipped glandular hairs except Herb Robert but that has white lines on the petals and a completely different leaf shape.  The sepal has a darker green centre line and edge lines giving it a striped appearance. This is quite common in many Geraniums and Storks-bills.

Main stems also have red-tipped glandular hairs plus non glandular hairs mixed in.

The petals have three pale veins and the anthers are purple. Hard to see the stigma in this photo but just possible to make out it is pale pink. The anther colour can be variable from purple through to almost white.  The mixture of long non-glandular and red-tipped glandular hairs can be seen on the sepals.

6th June 2019 Cambridge North Station

The cranesbill as it springs its seeds out, with two of the five mericarps sprung. Sepals have turned slightly red at this stage showing the three stripes. The stems are often red.

2) Geranium columbium      Long-stalked Cranesbill

3 Jun 2019 

The flowers are held on long thin stalks.  The flowers are also slightly bell shaped and this and the long stalks is normally what makes them stand out from the Dovesfoot Cranesbill that is the most common species. The plants are often taller and climb through other vegetation.  Long-stalked Cranesbill is quite rare in Cambridgeshire, so it is always nice to find.

3 June 2019 Pink petals normally slightly notched.
A key secondary feature is all the hairs on Long-stalked Cranesbill are pressed flat as shown above. These flattened hairs also cover the leaf stems and the leaves.

3 Jun 2019 Leaf with flattened white hairs.

Leaf shape is quite complex but a key feature shared with Cut-leaved Cranesbill is that the cuts often go to the base and the lobes have thin sections . The lobes tend to be slightly thicker than with Cut-leaved Cranesbill.

11th June 2019 Devils Ditch, Cambridgeshire
Close up of flower and leaf showing reddish stigma and pale anthers with blue line.

Long-stalked Cranesbill Flower.

3) Geranium dissectum        Cut-leaved Cranesbill

19May2019 Needingworth, Cambridgeshire

The standout feature, apart from the deep pink colour of the petals, is the very long sepal tip. The flower is similar to the Dovesfoot Cranesbill in that it has notched pink petals but the sepal tips are  much longer than the petals and I have not seen any plants that have any variation in this key feature.
Flowers have short stems. Occasionally the petals lack the notch.

Next example is a pale flower variation which is quite rare.

19 May 2019 Needingworth, Cambridgeshire.

In common with other cranesbills and storksbills, the pale versions of these flowers also have paler stigma and anther colour. In this case the stigma is off white and the anthers are pale pink. The normal deep pink flower has a dark pink stigma. Note this photo shows the five petal veins well, as it is easy to miss the two outer ones.

19 May 2019 Flower stem hairs
This photo shows the mainly glandular hairs with clear glands on the flower stem and the sepals mixed in with some non-glandular hairs. Note the aphid which has to battle against all this sticky stuff.

The flowers are on short stalks and the leaves are deeply cut and individual lobes are thin. Plants are often quite large but the flowers are tiny.


These three species are fairly straight forward to identify.  The leaf shape of the Round-leaved Cranesbill and pale versions being the possible variations.   Careful examination of the hairs will offer a clear ID feature in each case.

Peter Leonard
Rampton, Cambridgeshire
14th November 2019

Tuesday, 24 September 2019

Dovesfoot Cranesbill vs Small-flowered

Additional notes and photos on flowering Geraniums to support the ID Key features Blog.
Hedgerow Cranesbill

The ID Key features blog, August 2019 is a reduced set of features to distinguish between all the wild Geraniums found in the UK when in flower.
Additional features and photos are included here to give a more complete picture of the differences.

Three species are covered in this blog which share a common feature in that they do not have extension tips to the sepals.
The leaves are cut to 2/3rd towards base and are of a similar shape.
Dovesfoot Cranesbill ( Geranium molle)
Small-flowered Cranesbill ( Geranium pusillum)
Hedgerow Cranesbill ( Geranium pyrenaicum )

1) Dovesfoot Cranesbill ( Geranium molle)

Petal colour is often the first clue in separating G. Molle from G. pusillum, normally being a deep pink, however white forms are common with intermediate colours also common, including some with a pale purple tinge.
Small-flowered, (G pusillum) often have a blue tinge in the petals which are also often narrow with a gap between them.

Small-flowered G. pusillum in all the photos I have taken, have had a pale off-white stigma whereas G. molle have had a pink or reddish purple stigma.

Pale form but still has red/purple stigma.
Stigma colour is not mentioned in the literature as a method of identification which is surprising but perhaps occasionally some very pale forms of Dovesfoot Cranesbill may have a pale stigma. I have yet to find and photograph one.

Occasionally the petals may not have the notch, see the comments and photo at the end of the ID Key August blog.

The style splits into 5 curved stigma arms at the centre of the flower, with the stamen filaments holding the anthers arranged round the stigma in two rows. G. Molle has ten anthers whereas G. pusillum has only five of its ten stamen filaments having anthers. Anthers do fall off (after splitting to reveal pollen) so this feature may not be a clear cut in some flowers.

A pale pink version that is quite unusual. 24th June 2019, Cambridge.

Normal deep pink colour, Ten blue anthers present.

Each petal has five pale green veins from the green base.  Only five of the ten anthers showing probably because the first five have already fallen off having split. This can occur quite quickly and possibly most flowers are self-pollinating in these smaller geraniums.
Ten anthers present , 5 have split and have revealed their pollen and the other 5 are held lower into the flower and line up with the petal centre lines.   22nd Sept 2019 Rampton.
Note the overlapping petals have 5 veins which can be paler or darker than the pink of the petal and this may be dependent on the angle of the light. Some veins do split but not as much as the veins in the larger flowered Hedgerow Cranesbill ( G. pyrenaicum)  which shares quite a few features with G. Molle.  A significant difference is that the sepals on G. molle have long hairs whereas G. pyrenaicum has only short hairs ( many glandular ) .
Hedgerow Cranesbill (G. pyrenaicum) has flowers twice the size of  Dovesfoot Cranesbill (G.molle) and the sepals are shorter relative to the length of the petals. Size is always a dangerous parameter in plants as growing conditions can change some features more than others. Often flower size is largely maintained even when produced from tiny malnourished plants.

Hedgerow Cranesbill (G. pyrenaicum ) with short hairs on sepals. 

Side view showing tips of the sepals just short of the petal notches, approx 2/3rd of the lengths of the petals.
Side view, note long hairs plus short glandular hairs on sepals, a feature shared with G. pusillum.

Flower and leaf stems have 3 hair types, short glandular hairs, short simple hairs and very long hairs.
Mericarp at base of beak showing wrinkled groves and the impossibly small hairs which the photo does not show.
The seeds are held inside the mericarp case.  G. Molle.
Comparison of mericarp at same stage showing the lack of groves and abundant hairs for Geranium pusillum.
The both cases a sepals has been pulled back to show the mericarp.

2) Small-flowered Cranesbill ( Geranium pusillum)

Small flowers and no long hairs on flower and leaf stalks, separate pusillum from molle.  Leaf shape has been used in non flowering plants, as a identification feature but I found this too difficult. Much easier to use hair length.
 The leaves are split to 2/3rds towards base and are very similar to G. molle.   The flowers are small and the three veined petals have either no or very limited overlap. G Molle petals often overlap but there is variation with some having no overlap.  The sepals have long hairs plus short glandular hairs much like G. molle. 

G. pusillum, white 5 pointed stigma with 5 anthers only as the other 5 stamen lack anthers. Hard to see in this photo.
Side view showing sepals nearly as long as petals. Long hairs on sepals.
G. Pusillum showing pale stigma and five anthers. 13th June 2019, Rampton

G. pusillum, leaf stalks which have no long hairs, just short hairs (glandular and non-glandular).

Section of flower stem and sepal at base of flower showing hairs.

Flower stem hairs, a dense combination of straight simple hairs and glandular hairs , all short.

Typical leaf from a small plant growing in a gravel drive.
More complex leaf of G. pusillum growing in good soil.
Leaf shape is variable but certain features hold, like the 2/3rd cut towards base.

Hairy mericarps attached to the beak. G. pusillum, Rampton 24th Sept 2019

3) Hedgerow Cranesbill ( Geranium pyrenacum)

Geranium pyrenacum pale form
Side view of pale form of G. pyrenaicum 

Normal pink flower showing the inner 5  anthers having split and the outer 5 still intact.
Note the stigma has not opened. 

The photo above shows the Hedgerow Cranesbill follows other larger cranesbills in that the stigma is not open to receive the pollen from the split anthers at least at first, suggesting pollination is dependent on insects rather than self-pollination that may occur in the smaller species like G. molle.

First 5 anthers have fallen off and the second set of 5 have split open. Stigma has also opened with its 5 pale pink arms. The Hedgerow Cranesbill seems to have a more complex timing arrangement of pollen and stigma availability,  but from a identification view only, the stigma is pale pink and has thinner longer arms than G. molle.

Short hairs, mainly glandular on flower stalks.

Leaf stalk has 3 types of hairs, with short glandular and non-glandular plus long simple hairs.

G. pyrenaicum .  Leaf split into in this case 7 sections by cuts up to 2/3rds. 

Leaves can be split into from 5 to 9 lobes/sections depending on how deep the cuts go and this is open to different points of view, as you can't use the vein structure to give a clear result.

Seeds just before they get thrown from the mericaps.
4) Conclusion.
As usual, I use the excellent 'Harrap's Wild Flowers' as my constant companion in the field.
Additional information on Geraniums was also found in 'Hardy Geraniums' by Peter F. Yeo which has detailed introductory chapters on the family. It covers both wild and cultivated types.
I am fortunate that both G. Molle and G. pusillum both grow in my gravel drive here in Rampton.
I suppose the good news is that to date, none of these three species has hybridised so identification should be straight forward. It is just a matter of remembering the key features which is why I made up the ID Key blog in the first place. Please comment if you see any incorrect assumptions.

Peter Leonard
Rampton, Cambridge
24th September 2019