Modelos para medir la Calidad del Servicio

Five broods o f zebra fin ch es were norm ally-raised by both parents u n til independence (NR group) and f iv e broods were reared w ith th eir mother only, th e ir fa th er being removed before the fourth chick hatched (FR group). Some fem ales were relu ctan t to stop brooding th e ir chicks once the male had been removed so th e ir chicks were hand fed u n til feeding was resumed. For the f i r s t ten days a ll the chicks were handled to control fo r human contact.

■ 1 • V :.ï' % ■ - 25 - . .-S ' . ■ ■■ ,v'. . i ■ ■ ■ _{4 -}S: -%

were removed and housed with two norm ally-raised zebra fin ch song; tu to r s, one with fawn plumage and one with w ild type (g rey ). A ll the

song tu tors had been used in previous stu d ies o f parental behaviour and were known to have normal song outputs. Each pair of song tu to rs was selec ted on the b a sis o f song stru ctu re, i . e . with songs as d iffe r e n t as p o ssib le, y et s t i l l rep resen tative o f normal zebra fin ch song (see Sossinka and Bohner 19% ). In the NR group one tu tor (t s ) had a high proportion o f song elem ents in common w ith the young birds' fath er whereas the other tu tor (TD) did not share any song elem ents w ith the fath er (see Table 3 .1 ).

:

A ll the birds were housed in wooden, w ire-fronted cages but a fter independence the tu to rs in the NR group were separated from each other and from the young males by w ire mesh screens which divided the cages in to three id e n tic a l portions. Wire mesh prevents physical encounters but v isu a l and vocal in tera c tio n i s maintained, thereby perm itting normal song learn in g to occur (se e Bales 1987b). '

Song Terminology

and

Zebra fin ch males normally sin g only one song phrase, which i s usually Repeated several tim es, during a bout o f sin gin g. Within each phrase there are a number o f d iffere n t elem ents (4 - 1,1 in th is study, mean = 8 ), u sually sung in a fix ed sequence. An element i s the sm allest continuous trace on a sonagram which i s tem porally d is tin c t from neighbouring song elem ents. O ccasionally two d iffere n t elements were combined and le a rn t as one. These fused elem ents were scored as two separate elem ents sin ce each occurred in d iv id u a lly in the songs o f other males with common song learn in g h isto ry .

-

- 26 -

4. ;-.

f ; : î }

The songs of both tu tors and the young m ales were recorded w ith a is Uher 4000 tape recorder a t 4 months o f age. At le a s t ten song bouts per bird were analysed on a sound-spectrograph (Kay D ig ita l Sonagraph ; g 7800). The song stru ctu re Of each pupil and h is tu to rs were compared by making tra ce s of ty p ica l song phrases and a ssessin g the s im ila r itie s in frequency pattern, modulation and length of song , elem ents. This assessm ent was done blind; tu to rs' and p u pils' songs 4

, ■ ' -4'

were compared before the behavioural observations had been analysed.

< ■ .

Elements were c la s s ifie d as eith er common to both tutor and pupil ' or as unrelated in stru ctu re. In most casés the c la s s ific a tio n was not d if f ic u lt . However, elem ents were not always copied accurately , and m ales varied considerably in th is resp ect. In Fig. 3^1 for ^f' example, element "d" in B258‘ s song was copied more accurately by P3 ^

y . / : ■ - t

than by P5. N onetheless, the context in which these occur and th e ir ” sim ila rity in a number of resp ects suggests they are based upon t h e ¥ tu tor element concerned, rather than improvised or based on other- )! elem ents. Other zebra fin ch workers independently made the assessment !

, ' ' ' ■ T

and a high degree of in ter-ob server r e lia b ilty was reach (90%). The ÿ amount of song a pupil learn t frcra h is tu to rs was calcu lated in two ways: the proportion of song elem ents a bird learn t from each tu tor 4? and the proportion of each tu to r 's song the bird le a r n t. The f i r s t value was calcu lated as th e number of d iffe r e n t elem ents lea rn t from a . 44; tutor divided by the number of d ifferen t.elem en ts in th e young m ale's song. The second measure was the number of elem ents in the tu to r 's song th at was le a rn t as a proportion of the number of elem ents each 4 tutor sang in a ty p ica l song phrase.

. .

'4#

- 27 -

Behavioural Observations. '

Behavioural observations were made for two 30-minute periods each day to monitor in te r a c tio n s between the two tu to rs and th eir p u pils ,4 for a l l ten clu tch es sin ce behavioural d iffere n c es . between tu to rs m i^ t govern the p u p il's s e le c tio n of song elem ents to learn . The birds were colour-ringed to f a c il it a t e in d ivid u al recogn ition , and the plumage type noted.

Physical encounters w ith a ll other birds in the cage were recorded continuously throughout each thirty-m in u te period. Clumping, preening, tutor aggression towards the young m ales (measured as the number of pecks and chases to each in d ivid u al) were monitored. Dominance rank between tu to rs was determined by the number of aggressive attack s d irected towards th e other tu tor. The song output for each tutor was calcu lated as the to ta l number of song phrases recorded during the periods of observation.

RESULTS. _■;r-

Tutor Choice.

The r e s u lts in Tables 3 .I I and 3 .I l l show th at nine out of ten males in the NR group and a l l eleven young m ales in the FR group learn t song elem ents from only one of the song tu to rs but there was considerable v a ria tio n in th e proportion of e l m ent s copied. Fig. 3.1 shows the sonagrams of p u p ils P3 and P5 and th e ir two song tu to rs in the FR group. Both of these pupils lea rn t song elem ents from B258 but none f r o m B259. . Fig. 3.2 shows the sonagrams of pu pils £50 and

P52 in the NR group. Both p u pils lea rn t only frcxn the tu tor whose song was sim ila r to th e ir fa th e r 's .

- 28 -

Choice C riteria.

To determine the c r ite r ia which a young male zebra fin ch uses in s e le c tin g a song tu tor I examined three p ossib le cues; v isu a l im printing on the mother, sib lin g s or s e lf ; vocal d ifferen ces b etw een ,: the two tu to rs and d iffere n c es in the degree of physical in tera c tio n the two tu to rs had w ith th e ir p u p ils.

(1 ). V i sual iropr in t i ng ; - Immelmann e t a l (1978) found that;x::: d ifferen ces in plumage colour between w ild type and w hite birds were ’ important in zebra fin ch mate ch oice. Males v is u a lly imprint on th e plumage of th e ir parents and s e le c t a mate of th e same type. Visual im printing might a lso occur between fawn and w ild typ e\,in d ivid u als. i

For song learn in g, however, I found no evidence of v isu a l im printing to the mother ( ^ t e s t , n s). S im ilarly, sib lin g plumage . colouration had no sig n ific a n t e ffe c t on tu tor choice Çj^est, ns) and there was no evidence of auto-im printing to plumage type ^(%e8t, n s).

(2 ), Vocal d iffere n c es between tu to r s ;- D ifferen ces between th e two tu tors in the number of d iffere n t elem ents in a song phrase and song output; measured in terms of the number of sin gin g bouts recorded in one hour's observations per day^ during the period of observations, ; were examined* Given th at only tu tors w ith normal song outputs were used in the experim ents, v a ria tio n w ith in th ese lim its did not a ffe c t a p u p il's choice of song tu tor (Wilcoxon t e s t , n s); th is confirms the ’

r e s u lts of Bohner (1983). x)

4' - ' .. '

. ■ ^ . H-i

The number of song elem ents in the two tu to r 's song phrases does not appear to be important in tutor ohoice for eith er group or for

both groups combined (Wilcoxon, 'n s). .

For the norm ally-raised young m ales, however, elem ent-type i s i - important (^ ^ 6 .4 0 0 , df=1, P<0.05). Nine out of ten males chose song elem ents common to TS and F; the other male sàng a hybrid song composed of two o rig in a l elem ents, two TD elem ents and two elem ents present in the songs of both TS and F. Only one bird sang an' element #' p ecu liar to the fa th e r 's song whereas f iv e birds sang one or more ;■ elem ents which were s p e c ific to TS (see Table 3 . I I I ) . This suggests th at most of the elem ents common to TS and F were learn t from t s a fte r

* ' ' - ' ' '

independence; th is agrees w ith Bales (1985b).

, .

(3 ). Physical in te r a c tio n s ;- Ten Cate e t al (1984) found th at ;

. ' ' . - - #

zebra fin ch es crossfostered by Bengalese fin ch es were more lik e ly to . prefer a Bengalese fin ch mate i f the fo ster-p a ren ts had been ) aggressive towards them. The r e s u lts presented here fo r the FR group show th at aggression i s a lso important in song learn in g. A ll eleven . p u p ils chose to learn from the tutor which was most aggressive towards them ( ^ 1 1 .0 , d f= l, P<0.001), ir r e sp e c tiv e of the ov era ll d ifferen ce in amount of aggression displayed by the two; tu tors 0 ^ e s t , ns) or in th e ir dominance ranking ^ |^ est, n s). In the NR group aggression was f. prevented by the w ire p a rtitio n .

Other physical in tera c tio n s, i . e . preening and clumping, were - examined but there was no evidence th at they are important in tu tor choice.

^ 30 - : - , ■ ? ’ '

#

' _f . . . ' _{-. ,x ^} ^ ^ _{^ y .' . '}

; "

)

■ ■ ■

.

'

:.-■■■

'

,

j

. '- , . ; . .. \y

A lthou^ the amount of aggression d irected towards the pupil i s importaht in tu tor choice there i s no sig n ific a n t correla tio n between the le v e l of tu tor aggression and the amount of song lea rn t (Wilcoxon,

... .... :&

n s). The la t t e r may be a consequence of ind ividu al d iffere n c es in ; :' learning cap acity. '