I have found a quick way to count the samples using the 'Find' tool in Adobe Reader.
- E1b1b : 190 (= 8.3%)
-- E-V12 : 7
-- E-V13 : 92 (= 4%)
-- E-V22 : 6
-- E-M81 : 41 (= 1.8%)
-- E-M123 : 44 (= 1.9%)
- G2a : 99 (= 4.3%)
- I1 : 44 (= 1.9%)
- I2 : 115 (= 5%)
-- I2-P215 : 11 (= 0.5%)
-- I2-M223 : 37 (= 1.6%)
-- I2-P37.2 : 25 (P37.2+subclades = 2.9%)
--- I2-M26 : 38
--- I2-M423 : 4
- J1 : 55 (= 2.4%)
-- J1-P58 : 23 (= 1%)
- J2a : 169 (= 7.3%)
-- J2a-M67 : 39
-- J2a-M92 : 27
- J2b : 42 (= 1.8%)
- R1a : 32 (= 1.4%)
- R1b : 1525 (= 66.3%)
- R1b-M343 (incl. V88) : 22 (= 1%)
-- R1b-L23 : 1
--- R1b-U106 : 51 (U106+Z381 = 4.8%)
---- R1b-Z381 : 59
--- R1b-P312 : 371
---- R1b-L21 : 140 (= 6.1%)
---- R1b-U152 : 199 (= 8.7%)
---- R1b-Z195 : 132 (Z195+subclades = 28.2%)
----- R1b-SRY2627 : 222
----- R1b-Z220 : 199
------ R1b-Z278 : 74
------- R1b-M153 : 21
- T : 28 (= 1.2%)
TOTAL : 2299 samples
The paper says that they tested approximately 2500 samples. The data sheet only 55 pages with 42 samples per page, so it should be 2310 samples minus one line for the header. That leaves 10 samples unaccounted for. I must have forgotten some haplogroup(s). If someone finds the mistake, please let me know.
ANALYSIS
I have checked a bit the origin of the samples. A minority are immigrants from other regions of Spain.
What we notice at first sight is that I1, I2-M223, R1a, J1, J2a , E1b1b (esp. E-M123) and T have slightly higher frequencies than previously reported in smaller studies, while R1b which is considerably lower (82% => 66%).
Interestingly there isn't any Germanic R1b-106, I1 or I2-M223 in the Balearic samples (Mallorca, Menorca). I1 and R1b-S106 appears to be most common around Barcelona and central Catalonia. Haplogroup I1 is just above 5% in coastal Catalonia.
R1b-U152 is evenly distributed in all regions. At over 8%, it is by far the highest regional frequency reported to date in the Iberian peninsula. Since the Mediterranean coast of Spain was never known to be Celtic speaking, it looks like the Romans played a bigger role in spreading U152 in Iberia than the Celts.
There is a hotspot of G2a around Lleida (24 of of 223 samples, or 10.8%), Central Catalonia (15 out of 234 samples, or 6.4%) and around Barcelona, but there is very little of it in the Valencian region. Lleida is the inland Pyrenees region, which would have served as a refuge for the Neolithic population, like most mountainous parts of Europe.
There is a hotspot of J1 in Girona (both M267 and P58), where it makes up 8.3% of the population (18 of 219 samples). But otherwise J1 is well distributed in most regions except Lleida and Mallorca which only have one sample.
There is a hotspot of E-M123 in Castelló (16 out of 144 samples, or 11.1%), but that haplogroup is absent from the Balearic samples.
Most of the T1a samples are concentrated around Central Catalonia, Camp de Tarragona and especially Penedès (7 out of 164 samples, or 4.3%), but T1a is also found in Valencia, Mallorca, Barcelona and Girona. None in Lleide or Pireneu.
J1-P58, E-M123 and T1a are all potentially of Jewish or Arabic origin, although the Greeks and Romans could also have contributed in Catalonia.