. Bonner zoologische Monographien. Zoology. JAHN, BIRD COMMUNITIES OF THE ECUADORIAN CHOCO capture/recapture data. The first and third divisions separated some open-country and forest-inhabiting species rather cleanly. However, divisions 2 and 4 divide several forest birds into distinct groups, with- out the reasons for the separations being clear. In the case of division 2, Myiornis atricapillus and Hylophi- lus decurtatus are subcanopy-dwelling birds (positive sign), whereas the rest mainly inhabit the forest undergrowth. Because the second division may not represent meaningful results, it s
. Bonner zoologische Monographien. Zoology. JAHN, BIRD COMMUNITIES OF THE ECUADORIAN CHOCO capture/recapture data. The first and third divisions separated some open-country and forest-inhabiting species rather cleanly. However, divisions 2 and 4 divide several forest birds into distinct groups, with- out the reasons for the separations being clear. In the case of division 2, Myiornis atricapillus and Hylophi- lus decurtatus are subcanopy-dwelling birds (positive sign), whereas the rest mainly inhabit the forest undergrowth. Because the second division may not represent meaningful results, it seems advisable to ignore the higher divisions. In contrast to TWINSPAN, the program IndVal „ (Dufrene & Legendre 1997) computes indicator groups based on predefined habitat categories. Con- sequently, the allocation of indicators to certain habitats is unambiguous. Moreover, IndVal also pre- sents the basis for the assignments, , for each habitat considered, the relationship between abun- dance and number of sampling units in which the corresponding species was recorded as well as a t-test allowing the significance of the results to be assessed (Tables 50 and 51). Although many of the indicator species determined by IndVal were also identified by TWINSPAN, with the latter approach major uncer- tainties remain whether the results are meaningful or not. The number of indicator taxa determined on the basis of the transect-mapping sample (n = 61; Table 50) was slightly greater than for the mist- netting data (n = 47; Table 49). Based on captures, IndVal listed 27 indicator species for open country, 8 for young successional forest, and 12 for medium- age forest. The corresponding numbers for transect- mapping records are 29, 10, and 22, THMIQU A ysF TGTHLPA* THTAMA THMYPA-^]^ TYTOCIA TPTHEP DECATR- TYMZGR> a EMSPCV THMZEX PTlMICI ATPTYDE A TPCTOL THGYLEA^ DEXIER FOHYPEA^ magp kPRSAAE TUTUDA PRLECO THMYAX 50 100 150 Axis 1 D-F/OC FIG. 46.
Size: 960px × 2604px
Photo credit: © Library Book Collection / Alamy / Afripics
License: Licensed
Model Released: No
Keywords: ., bookcentury1900, bookcollectionbiodiversity, booksubjectzoology
