Rallus minutus Alcover, Pieper, Pereira & Rando, 2015, n. sp.

Rallus minutus

R. minutus n. sp. is one of the most distinctive species found. It is a dwarf rail, the smallest among the Macaronesian species, only slightly larger than Atlantisia rogersi, with robust leg bones (especially tarsometatarsus) and reduced wings. Its bill is relatively curved and the bill to total skull length ratio is slightly larger than in R. aquaticus. A long and curved bill in insular rails might be related to foraging by probing in soft soil, deep litter and mosses, as suggested elsewhere for Capellirallus karamu from New Zealand (Worthy and Holdaway 2002).

The bones of R. minutus n. sp. are generally similar to those of R. aquaticus, but differ considerably in size and proportions. All the bones are clearly shorter than their respective R. aquaticus counterparts. R. minutus n. sp. combines reduced wing bones (mainly carpometacarpus and ulna) with relatively short and massive leg bones (mainly tarsometatarsus). The humerus, ulna and carpometacarpus are 35%, 47% and 49% shorter than in R. aquaticus, while the femur, tibiotarsus and tarsometatarsus are 28%, 28% and 36% shorter.

R. minutus n. sp. has a shorter coracoid, scapula, humerus, ulna and carpometacarpus than in R. montivagorum n. sp. and R. aquaticus. The reduction of humerus length is of about 26% compared to the coracoid and 35% with respect to the scapula. The ulna is smaller than in R. lowei n. sp. (28%), R. montivagorum n. sp. (37%; U = -2.646; p = 0.003) and R. aquaticus (47%; U = -2.838; p <0.001). The carpometacarpus is 42% and 49% shorter than in R. montivagorum n. sp. and R. aquaticus respectively (the carpometacarpus of R. lowei n. sp. is not available). R.

minutus n. sp. also exhibits the most reduced carina sterni of all the Macaronesian rails. The carpometacarpus is much shorter relative to the humerus (Cmc/H=0.44) than in the rest known species of Rallus (including R. cyanocavi), which is an indication of extremely weak development of primary flight feathers.

The femur and the tibiotarsus are shorter than in R. lowei n. sp. (22% and 14%, respectively), R. adolfocaesaris n. sp. (23% and 10%), R. carvaoensis n. sp. (18% and 13%), R. montivagorum n. sp. (17% and 14%) and R. aquaticus (28% for both bones). The tarsometatarsus is robust (Figure 17). This bone is shorter than in R. lowei n. sp. (47% smaller), R. carvaoensis n. sp. (13%), R. montivagorum n. sp. (23%; p <0.005 in the three cases) and R. adolfocaesaris n. sp. (20%; U = -2.236; p = 0.036). In addition, the tarsometatarsus distal width is smaller than in R. lowei n. sp. (p <0.001) and R. aquaticus (p = 0.001), but similar to R. carvaoensis n. sp. (p = 0.871), R. montivagorun n. sp. (p = 0.678) and R. adolfocaesaris n. sp. (U = -1.167; p = 0.262).

The humerus-to- length to femur length ratio (0.84) and the humerus+ulna+carpometacarpus to femur length ratio (1.8:1) show that R. minutus n. sp., was a flightless species (Livezey 2003).

The Figure 21 shows the femur vs. tarsometatarsus relative lengths, in relation to total length of leg bones, in the Macaronesian taxa and other extant and extinct rails (Table 2). This plot shows that rails with a tarsometatarsus relative length above 27.5% survive, the only exception being the extinct R. montivagorum n. sp. from Pico (Figure 21). The PCA performed on the proportions of hind and forelimb bones (Table 2; Figure 22) produced two principal components explaining 88.6% of the total variance. 64.1% is explained by PC1, which shows a high positive weighting for ulna, carpometacarpus and tarsometatarsus, and a high negative weighting for humerus and femur proportions. PC2 explains 24.5% of the variance and shows a high positive weighting for tibiotarsus, and a moderate negative weighting for ulna and femur proportions. All the extinct species score zero or below on PC1 with the only exception of R. montivagorum n.sp (the only flying species among the Macaronesian rails).