1

Morphological similarities between Spanish pigeon breeds

2
3

Pere M. PARS-CASANOVA*

Dept. of Animal Production; University of Lleida; Av. Alcalde Rovira Roure, 191;

25198-Lleida (Catalunya, Spain)

Telephone: +34973706460

Fax number: +34973702874

E-mail address: peremiquelp@prodan.udl.cat

Abstract: Relationships among 30 pigeon breeds were studied using data from 31

morphological characteristics. The relationships between breeds obtained from the study

were rather congruent with the used classification fanciers groups. The average

Euclidean distance between pigeon breeds was of 0.930.16 STD. The obtained results

were very compatible with those described in literature for each classificatory group.

Three large, perfectly definite clusters could be observed in this tree. A cluster was

formed by Balearic non-pouter breeds and the Catalan Bare Pigeon. Another cluster

occupied an intermediate position and is formed by thief pouters. They shared some

remarkable peculiar features that are morphologically very different from the other

10

breeds. A third more diversified cluster was close to Turkish Takla Tumbler which was

11

used as outgroup.

12
13

Key words: cladogram, Columba livia, doves, morphology

14
15

Introduction

16

The great diversity of pigeon breeds that exists in the world, and attempts at systematic

17

classification of these breeds into closely related groups are widely known (1). But

18

studies of filiations or possible genetic relationships among the different pigeon breeds

19

differ substantially, because not enough archaeological or historical documents are

20

provided to allow one to reconstruct clearly the diversification of pigeon from its origin.

21

According to Sambraus (2) there are ten groups of pigeons: form, tubenose, hen, crop,

22

colour, drummer, structure, little gulls, tumblers and display doves. Pigeon breeds differ

23

in body shape and structure, colour and markings of the plumage. But few studies from

24

morphological characters in pigeon breeds have been made (3, 4). Pigeon fanciers with

their own favourite breeds or particular interests have provided partial histories in books

and magazines dedicated to these birds; they are often selective and perhaps inaccurate,

thus, attempting to cobble together a classification of local breeds requires relying

largely on scientific methods. Morphological characters can provide very useful

information to complete other investigations about genetic relationships of domestic and

all other breeds in general as well as it is extremely important anthropologically.

Statistical techniques such as multivariate analyses and the application of numerical

taxonomy to the data derived from morphological characters allow a different treatment

of the information generated. This is shown in the works of Jordana et al. (5).

10

This paper presents a study from qualitative and quantitative data analyses, using

11

statistical methods and available computing packages specially designed for such

12

analyses. The research is centred in Spanish pigeon breeds although a Turkish breed has

13

been included.

14
15

Materials and Methods

16

Breeds studied

17

Twenty-nine Spanish pigeon breeds were studied: Borina (BBO), Balearic Pouter

18

(Gavatxut, BBP), Canary Pouter (CAP), Sevilian Colillano Pouter (COP), Gaditano

19

Pouter (also called Isleo and Jerezano, GAP), Granadian Pouter (GRP), Jan

20

Pouter (JIP), Sevilian Thief Pouter (LSP), Marchenero Pouter (MAP), Marteo Pouter

21

(MTP), Moroncelo Pouter (MEP), Nape Pigeon (Quebrado Murciano, MUP), Spanish

22

Owl Pouter (Rafeo, SOP), Veleo Pouter (VEP), Spanish Owl (Xorrera or Colom

23

dEnreixat, SOW), Escampadissa Roller (BES), Valencian Frill (VFI), Flamenca

24

Runt (Colom dUll, FLR), Giant Mallorquina Runt (Mallorquina de Casta Grossa,

BGR), Balearic Homer (Nas de Xot, BHO), Valencian New Pouter (VNP), Strawberry

Eye (Ull de Maduixa, STE), Balearic Esbart Roller (Pinta Balear or dEsbart,

BER), Fish Eye Roller (Refilador or Ull de Peix, REF), Catalonian Tumbler (CAT),

Rumped Pouter (Buchn Morrillero, RRP), Alteo Pouter (ALP), Spanish Barb

(Flamenquilla, FLA) and Catalan Bare or Naked-Neck Pigeon (Coll Pelat, NNP).

Takla Tumbler (TUT) was included as outgroup. This Turkish tumbler is originated

from Central Asia and in Turkey can be considered as the most popular pigeon breed.

Qualitative and quantitative analyses

A total of 31 morphological characters were studied in ideal specimen of each of the

10

studied breeds. The characteristics were those described habitually in a breed

11

description (6) and were considered as being informative (colours have not been

12

considered). The state of each character for each breed was obtained from official racial

13

standards recognized by the Real Federacin Espaola de Colombicultura or from

14

descriptions by Levy (4), Mackrott (7) and Schille (8). Numbers were assigned to each

15

state of the different character in an arbitrary manner. These numbers did not represent

16

any specific weight. The number of states for each character was established depending

17

upon the number of distinguishable phenotypic classes. Coded states moved in the range

18

1 to 4 and unique states in multistate characters were avoided. Characters and their state

19

are shown in Table 1.

20

For qualitative analysis, discrete characters (F, G, J, U and AA) were recoded into a

21

series of two or three-state characters, denoting absence, presence or both of the

22

characters. Continuous characters may be split into a small number of classes, each

23

representing one of the states of the each character in the data matrix. The original

24

matrix of morphological resemblances is shown in Table 2. Their analysis was based on

the parsimony principle, and the criterion was to find the tree (cladogram) that required

the least number of changes (9). The method used was Fitch parsimony. The used

heuristic algorithm was Subtree Pruning and Regrafting (SPR) (10).

For quantitative analysis, qualitative data were transformed and processed into a matrix

of similarity coming from indices pairwise distance formulae (11, 12). The Euclidean

distance (13) was calculated as a measure of this distance resemblance, under the

assumption of independence between considered characters. With these values a cluster

tree was elaborated using the Ward method. Ward's method is a hierarchical method

designed to optimize the minimum variance within clusters. The method searches

10

objects that can be grouped together while minimizing the increase in error sum of

11

squares.

12

All analyses were made using the PAST computing package (Paleontological Statistics

13

Software Package for Education and Data Analysis).

14
15

Results

16

Qualitative analysis

17

The cladogram resulting from the application of the Fitch parsimony method to the

18

morphological traits is shown in Figure 1. Fitch parsimony evaluated 12,700 trees. The

19

values in the tree indicated the number of replicates from the bootstrap analysis (loosely,

20

the width of the confidence interval). Fitch parsimony needed 200 steps (total length of

21

the tree) to rearrange the characters to obtain the minimum parsimonious tree. It is

22

imperative to note that the obtained cladogram must be viewed merely as a group tree

23

concerning morphological relationships among the studied breeds more than a

24

phylogenetic tree. Their similarities between groups must not be used to group like with

like in descending order of specificity. The Consistency Index (CI) of 0.31 indicated

that characters fit the obtained cladogram rather imperfectly. Figure 2 represents the

strict-rule consensus tree formed after one hundred bootstrap replicates. The parsimony

tree and the strict-rule consensus tree were very similar, which is frequently the case.

Quantitative analysis

The average value of Euclidean distances between breeds was of 0.930.16 STD, with

extreme values of 0.35 between the Borina vs Escampadissa Roller pair, and 1.42 for

the Spanish Owl Pouter vs Escampadissa Roller. That average distance had a different

magnitude to that obtained between other domestic animal species; for instance, ovine

10

had Euclidean distance of 1.150.21 STD (5), which indicates that different degree of

11

morphological difference exists between the breeds of other species. In Figure 3 it

12

appears the dendrogram produced by hierarchical cluster analysis of the data, using

13

Ward's Method and the Euclidean distance measure. Again Balearic breeds and pouters

14

were clearly clustered (with the uncertain position of Moroncelo Pouter (MEP) again

15

but with the Canary Pouter (CAP) included). Carunculated breeds formed a little cluster

16

as it was done by Catalan fliers and Takla Tumbler.

17
18

Discussion

19

Obtained values are slightly different from that obtained in a sheep research (CI= 0.368;

20

RI=0.614) (5) which in fact included 44 characters. However, such CI comparisons only

21

makes sense when comparing levels of homoplasy exhibited by cladograms concerning

22

somewhat similar number of characters referring to a similar type of data set in a same

23

biological group. Indeed, there is no reason to think that sheep and pigeon breeds,

24

exhibit exactly the same levels of homoplasy.

Although low significance levels indicate that there is little confidence in this

arrangement, our obtained results are very compatible with those described for each

classificatory group. Three large, perfectly definite clusters can be observed in this tree.

Cluster A is formed by Balearic non-pouter breeds [Borina (BBO), Escampadissa

Roller (BES), Giant Mallorquina Runt (BGR), Balearic Esbart Roller (BER) and

Balearic Homer (BHO)] that are all from Balearic Islands. The Catalan Bare Pigeon

(NNP) which is included in this cluster is a tumbler with a striking appearance that was

spread all around the Iberian Peninsula. Cluster B occupies an intermediate position and

is formed by thief pouters [Rumped Pouter (RRP), Veleo Pouter (VEP), Balearic

10

Pouter (BBP), Granadian Pouter (GRP), Gaditano Pouter (GAP), Marchenero Pouter

11

(MAP), Sevilian Thief Pouter (LSP), Spanish Owl Pouter (SOP), Jan Pouter (JIP),

12

Sevilian Colillano Pouter (COP), Marteo Pouter (MTP), Nape Pigeon (MUP), Alteo

13

Pouter (ALP), Valencian New Pouter (VNP) and Moroncelo Pouter (MEP)]. The

14

pouters share some remarkable peculiar features that are morphologically very different

15

from the other breeds -these characters concern, naturally, those related with globe and

16

global weight and justify their clustering-, so they appear clearly clustered. Moroncelo

17

Pouter (MEP) is the most separated breed in this group and this can be explained by the

18

fact that the breeds have a small crop that is never hunged. And more diversified cluster

19

C is close to the outgroup. Some of them [Valencian Frill (VFI), Fish Eye Roller (REF)

20

and Catalonian Tumbler (CAT)] are good fliers (Sambraus group IX). The Canary

21

Pouter (CAP) is included.

22

At an individual breed level, morphological analysis perfectly assigned each population

23

into their functional group. In cluster C: the Flamenca Runt (FLR), Spanish Barb (FLA)

24

and Strawberry Eye (STE) are included together forming a carunculated subgroup

(Sambraus group VII). The Spanish Owl (SOW) is an exhibition pigeon that shares

some rather peculiar morphological characters with the Valencian Frill (VFI) and that do

not appear in the rest of the breeds (pants, feather-legged, breast frill). Moreover it is

said that Valencian Frill is an offshoot of the ancient Tunisian Owl (4). Fish Eye Roller

(REF) appears closest to Takla Tumbler (TUT) and that is an interesting case, because

the Fish Eye Roller is probably related to Pappatacci (Spanish Primitive Jacobin, not

considered here). And the Pappatacci breed has its origins in Greece (from 1302-1313

the Catalonians ruled Athens, by conquest) (4). Perhaps the Fish Eye Roller is the most

oriental Spanish pigeon breed? The rest of pigeons of cluster C belong to ancient breeds

10

and it is said they have been brought to the Iberian Peninsula from the Barbary Coast

11

(North Africa) (4). The Canary Pouter (CAP) presents remarkable peculiar features that

12

are morphologically very different from the other thief pouters, perhaps due to the

13

influence of Rock pigeon from Canary Islands (Columba livia canariensis) that is said to

14

have had influence on its origin.

15

It must be remembered that morphological characters have been subjected to a great

16

pressure of artificial selection. Artificial selection has had evolutionary strength, and

17

would have been the major influence on the process of breed differentiation. The

18

cladogram would show that the Balearic breeds had maintained little influence from

19

other representatives of continental Spanish breeds. It seems naive not to believe that

20

insular breeds are influenced more easily with neighbour breeds than continental ones.

21

Ancient breeders did breed, and they did pass on information to other breeds. This

22

transmission created what it is referred to as breed traditions. As states Jerolmack (14):

23

culture is inscribed in animals through the process of domestication in ways that, while

context specific and somewhat fluid, are also cumulative and grounded in the biology of

the animal.

Breeds with the same functional purpose have thus more similarities in character states

for an evident reason of functionality. For instance, in thief pouters, while each breed

has its own morphological features, breeders selected for ways of flying ways, rendering

their morphological similarities clearly associated to an adaptation to the flying ability.

Or structure pigeons that are the less related ones as they have been selected for external

appearance in order to obtain aesthetically attractive birds, but not in terms of functional

traits.

10
11

Conclusion

12

It must be pointed out that the results obtained in our study attempted to show only the

13

degree of relationship of morphological similarity among some current Spanish pigeon

14

breeds, which may or may not be indicative of the true evolutionary history of their

15

populations. It ought to be considered that morphological characters have been

16

subjected to artificial selection over a long period of time and also, there has been

17

genetic migration among some of these populations. For studies of evolutionary

18

divergence, neutral genes, with a high rate of polymorphism and no relationship with

19

the fitness of the individuals, would be more appropriate, especially for alleles that

20

appeared by mutation and fixation or lost by drift, specifically the markers of DNA,

21

minisatellites and microsatellites. Nevertheless, analysis of morphological characters

22

indicated the relationships of similarity, at this level, between current Spanish pigeons.

23

10

References

1. Bodio, S.: Aloft: A Meditation on Pigeons and Pigeon-Flying. Lyons & Burford

Publishers. New York, 1990.

2. Sambraus, H.H.: A Colour Atlas of Livestock Breeds. Wolfe Publishing Ltd,

Germany, 1992.

3. Patent, D.H.: Pigeons. Clarion Books. New York, 1997.

4. Levy, W.M.: Encyclopedia of Pigeon Breeds. Sumter SC. Levi Publishing Co.

Sumter. South Carolina, 1965.

5. Jordana, J., Manteca, X., Ribo, O.: Comparative analysis of morphological and

10

behavioral characters in the domestic dog and their importance in the reconstruction of

11

phylogenetic relationships in canids. Genet. Mol. Biol., 1999; 22 (1).

12

6. Roelfsema, J.: Conservation of the Gelderse Slenk. Major Thesis. WUR, 2007.

13

7. Mackrott, H.: Palomas de Raza. Omega. Barcelona, 1997.

14

8. Schille, H.-J.: Gua de las Palomas de Raza. Arte Avcola. Valls, 2005.

15

9. Hammer, ., Harper, D.A.T., Ryan, P. D.: PAST. Paleontological Statistics Software

16

Package for Education and Data Analysis. Palaeontologia Electronica, 2001; 4 (1). On

17

line: http://palaeo-electronica.org/2001_1/past/issue1_01.html.

18

10. Kitching, I.J., Forey, P.L., Humphries, C.J., Williams, D.M.: Cladistics. Oxford

19

University Press, 1998.

20

11. Naylor, G., Kraus, F.: The Relationship between s and m and the Retention Index.

21

Syst. Biol., 1995; 44: 559-562.

22

12. Farris, J.S.: The retention index and the rescaled consistency index. Cladistics, 1989;

23

5: 417-419.

11

13. Sneath, P.H.A., Sokal, R.R.: Numerical Taxonomy. Ed. W.H. Freeman. San

Francisco, 1973.

14. Jerolmack, C.: Animal archaeology: Domestic pigeons and the nature-culture

dialectic. Qualitative Sociology Review III, 2007; (1): 74-95.

5
6

Table 1. Characters and their states, used for the construction of the morphological

resemblance matrix.
(A) Weight
1. Elipometrical (< 250 g)
2. Eumetrical (250-350 g)
3. Subhipermetrical (> 350 g)
(B) Size
1. Very small
2. Small
3. Medium
4. Big
5. Very big
(C) Chest
1. Medium
2. Wide
3. Very wide
(D) Legs length
1. Short
2. Medium
3. Long
(E) Legs thickness
1. Fine
2. Medium
3. Thick
(F) Leg frills? (pants)
1. Yes
2. No
3. Grouse-legged
(G) Clean legged?
1. Yes

(P) Chin wattles?

1. Without
2. Fine
3. Well developed
(Q) Eye ceres?
1. Fine
2. Developed
3. Well developed
(R) Globe seize
1. Little
2. Medium
3. Well developed
(S) Globe form
1. Slight
2. Round
3. Oval (pear shaped)
(T) Hunging of globe
1. Not hunged
2. Pendulous (low crop)
3. Hanging (high crop)
(U) Vertical crease in the globe?
1. Yes
2. No
3. Yes or no
(W) Neck length
1. Short
2. Medium
3. Long

12

2. No
3. Yes or no
(H) Relative head size
1. Small
2. Medium
3. Big
(I) Head shape
1. Rounded
2. Squared
3. Longish or almond shape (ram head)
4. Dice shaped (owl head)
(J) Crest?
1. Yes or no (plain or peak-crested)
2. No (plain headed)
(K) Beak bulkiness
1. Fine
2. Medium
3. Thick
(L) Beak length
1. Very short
2. Short
3. Medium
4. Long
(M) Beak form
1. Straight
2. Slightly curved
3. Curved
4. Very curved (owlish)
(N) Nose wattles?
1. Little
2. Slightly developed
3. Developed
(O) Aspect of nose wattles
1. Fine
2. Rough
1
2

(X) Neck thickness

1. Fine
2. Medium
3. Thick
(Z) Back
1. Flat
2. Slightly arched
3. Roached or arched
(AA) Garland?
1. No
2. Yes
(AB) Tail length
1. Short
2. Medium
3. Long
(AC) Tail form
1. Flat-tailed
2. Arched (tile-tailed)
(AD) Rump width
1. Medium
2. Broad
(AE) Point of primaries
1. Extending to tip of tail
2. Not extending to tip of tail
3. Longer than tip of tail
(AF) Position of wings upon the tail
1. Upon the tail
2. Below the tail
3. Upon or below the tail
(AG) Station
1. Horizontal (low carriage)
2. Vertical (upright)

13

1
2

Table 2. Morphological resemblance matrix. Abbreviations in the text.

Breeds/
character A B C D E F G H I J K L M N O P Q R S T U W X Z AA AB AC AD AE AF AG
ALP
BBO
BBP
BER
BES
BGR
BHO
CAP
CAT
COP
FLA
FLR
GAP
GRP
JIP
LSP
MAP
MEP
MTP
MUP
NNP
REF
RRP
SOP
SOW
STE
TUT
VEP
VFI
VNP

3
4

2
2
3
2
2
3
2
2
2
3
2
3
2
3
2
3
2
2
3
2
2
2
3
3
2
2
3
3
1
2

3
3
3
3
3
5
3
3
1
3
2
4
3
4
3
3
2
2
3
2
2
3
3
3
2
4
2
3
1
3

1
1
2
2
1
3
1
1
2
2
2
2
2
2
2
2
1
1
1
1
1
3
1
2
1
2
2
1
2
1

1
1
1
2
1
1
3
1
2
2
1
1
2
3
2
2
1
2
3
3
2
2
3
2
1
1
2
3
2
2

312
222
112
222
222
222
222
212
222
212
212
232
112
212
212
212
112
212
312
312
222
222
112
312
111
313
214
112
232
312

2
1
2
1
2
1
2
2
1
2
3
3
2
3
3
2
2
2
2
1
2
2
2
3
1
3
2
1
1
2

123
122
123
322
121
122
322
323
121
121
223
123
321
423
323
323
121
322
121
121
122
211
121
423
221
113
112
121
221
122

33
31
22
22
31
31
33
22
32
32
22
24
22
32
42
32
22
32
23
31
31
41
32
14
22
12
32
32
22
31

2
1
2
1
1
1
1
3
1
1
3
3
3
3
3
3
1
2
2
2
1
1
1
3
1
3
1
1
1
2

1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

11
11
21
21
11
21
33
21
21
21
13
23
21
22
31
31
21
21
21
21
21
11
21
22
11
23
11
21
11
22

3
1
2
1
1
1
1
2
1
3
1
1
3
2
2
3
3
1
2
3
1
1
1
3
1
1
1
2
1
3

322
112
232
112
112
112
112
311
112
332
112
112
231
231
323
321
231
212
222
322
112
112
131
321
112
112
112
231
112
323

2
1
1
1
1
1
2
2
3
3
1
1
2
3
3
3
1
3
3
3
2
3
3
1
2
1
2
3
2
2

2
3
3
2
2
3
3
2
2
3
2
2
2
3
3
2
3
1
2
1
2
2
1
3
3
2
2
2
2
3

31
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
31
11
11
21
12
11
11
31
31
11
11
11
11
11

2
3
2
3
3
3
2
1
2
2
2
2
1
2
2
2
1
2
2
1
2
2
2
1
2
2
2
2
2
2

1
1
2
1
1
1
1
1
1
1
1
1
2
2
1
1
2
2
1
1
1
1
2
1
1
1
1
2
1
1

1
1
1
1
1
2
1
1
1
1
1
1
2
1
1
2
2
1
2
2
1
1
1
2
1
1
1
2
1
1

1
1
2
2
1
1
2
1
1
3
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
1
2
1
1

1
2
1
2
3
2
1
1
1
1
1
1
1
2
2
1
1
1
1
1
1
1
2
1
1
1
3
1
1
1

2
1
2
1
1
1
2
1
2
2
2
1
2
2
2
2
1
2
2
1
2
2
2
2
2
2
2
1
2
1

14

Figure 1. Qualitative analysis of morphological data. Cladogram resulting from the

application of Fitch parsimony method and SPR heuristic algorithm. Abbreviations in

the text.

4
5
6

Figure 2. Strict-rule consensus tree from morphological data. Abbreviations in the text.

15

Figure 3. Dendrogram produced by hierarchical cluster analysis of the data, using

Ward's Method and the euclidean distance measure. Abbreviations in the text.