Phylogenetic Analysis of Avian Parental Care

The Auk 109(4):828-846, 1992

PHYLOGENETIC ANALYSIS OF AVIAN PARENTAL CARE

MARY C. MCKITRICK Museumof Zoologyand Departmentof Biology,University of Michigan, Ann Arbor,Michigan 48109, USA

ABsTRACT.--Phylogeneticpatterns can be usedto generatedetailed historical hypotheses about the evolution of charactersystems, including parental-carebehavior. Phylogenetic analysisof 60 taxa usingparental-care data (15 characters)in combinationwith anatomical data (69 characters)shows that biparental care is primitive for birds, and that biparental incubationarose from an ancestralcondition in which neither parent incubated.Components of parentalcare may be decoupled,such that incubationand feedingof nestlingsis biparental, but postfledgingcare is uniparental;the patternof decouplingvaries across taxa. Behavioral plasticity,environmentally induced variation, and quantitativevariation in charactersthat are codedqualitatively are someof the problemsinherent in suchan analysis.Nevertheless, historicalanalysis offers a goodvehicle for obtainingrigorous hypotheses about the coevo- lution of avian charactersystems. Received 5 July 1991,accepted 29 October1991.

THEUSE of phylogeneticpattern to elucidate phylogeny(one of the "topthree heinous 'crimes questionsabout other evolutionaryand ecolog- againstphylogenetics' "; Brooksand McLennan ical patternsor processeshas becomeincreas- 1991). Looking at numerousspecies from the ingly commonin the last 15 years.Several rig- same taxonomic family actually may be quite orous attemptsto incorporatebehavioral data valid, dependingon phylogeneticrelationships into phylogeneticanalyses or to comparesuch amongthose species. What mattersabove all is datawith existingphylogenies have beenmade the number of times the trait in questionarose, in the lastfive years.Clutton-Brock and Harvey rather than how many genera or families show (1977) recognized the pitfalls of comparing the trait; this can only be determined by a de- closely related speciesto study the origin of tailed phylogeny of the organismsand traits. ecologicalphenomena, and this has been fur- The trait could well have arisen numerous times ther emphasizedby Clutton-Brockand Harvey within the family, in which caseit is not only (1979), Ridley (1978, 1983), Harvey and Mace meaningful but necessaryto include members (1982), Felsenstein(1985), Dobson (1985), Pagel of the samefamily in the study. With such in- and Harvey (1988), H/Sglund(1989), and Bj/Srk- formation, one can ask appropriate questions lund (1990). Coddington (1988) outlined a about the adaptive significanceof that trait. It method for using cladistichypotheses of rela- is not profitable to begin the investigationby tionship to test hypothesesof adaptation,and determiningwhether one shouldcount genera, this kind of approachhas been fruitfully ap- families or orders in which the trait occurs. plied in a numberof recentstudies (Prum 1990, Rather, one should begin with a detailed phy- B. N. Danforth unpubl. data,R. J.Smith unpubl. logeny and determine whether the trait arose data). once or many times, and the pattern of origin A historicalapproach is a profitable one, as and loss. Adaptive explanationscan only be it attemptsto find the correcthierarchical level tested with convergent cases,not with homol- at which hypothesesof adaptationare mean- ogousones. In other words,two speciesshow- ingful. It is not enough simply to dismiss"ad- ing the trait can only serve as two data points aptations"as artifactsof phylogeneticpropin- in supportof an adaptiveexplanation of origin quity; rather, one mustdissect out the effectsof if they did not inherit the trait from a common phylogeny so that adaptive explanationsmay ancestor. be viewed in the appropriatecontext. Clutton- McLennan et al. (1988) lamented what they Brockand Harvey's (1979) concernabout bias- perceivedas a declinein the use of behavioral ing investigationsof ecologicalphenomena by data to constructphylogenetic hypothesesin examining closely related specieswas appro- the last three decades,citing work by Lorenz priate. However, they may have been some- (1941) in which this was done successfullyfor what tyrannizedby taxonomyin their view of ducks(Anatidae) and work by later authorsin the problem--that is, equatingtaxonomy with which phylogeny was ignored. A noncladistic 828 October1992] PhylogenyofAvian Parental Care 829 attempt to extend Lorenz's work was made by Cracraft's (1981) classification, which lacks Johnsgard(1961), for example.McLennan et al. characterdata and explicit hypothesesof phy- (1988) were remarkablysuccessful in generat- logeneticrelationships, is of limited utility for ing phylogenies for gasterosteidfishes using the purposeof understandingthe evolutionof data on reproductivebehavior, and argued that character systems(e.g. behavior). Without a behavioral data are extremely useful for con- phylogeny,van Rhijn was limited to making structing phylogenies. hypothesesbased on statistical probabilities The evolution of parental care was an early rather than historical evidence. focus of the phylogenetic approach. Ridley In this study, I use characterdata from the (1978) reviewed the circumstancesof paternal hindlimb musculature (from McKitrick 1991) in care in a broad array of taxa to consider the combinationwith published behavioral data to effectsof fertilization mode, territoriality, fe- test van Rhijn's hypothesisthat male-only pa- male choice and other factors on the evolution rental careis primitive in birds,and to generate, of paternal care.He recognizedthe possibility in effect,phylogenies of avian parental-carebe- of phylogeneticbias in formulatinghypotheses havior. In doing so, I also was able to test some about the evolution of the phenomenon. Git- of the hypothesesof Silver et al. (1985) about tleman (1981) used phylogeny to test predic- the causalnature of associationsamong certain tions about probabilities of transition among aspectsof parental care. In considering the typesof parentalcare in bony fishes(i.e. among question of what aspectsof parental care are states of uniparental [male], uniparental [fe- primitive for birds,however, it shouldbe borne male], biparental, and no parental care). He in mind that there is nothing magical about foundthat the mostcommon evolutionary tran- statementsregarding the primitive or derived sition was from no parental care to uniparental nature of these character states. A character state care by males, but pointed out possiblestatis- that is derived for birds is primitive within birds. tical problems with the conclusion that this A statemay be derivedfor birds,but showearly transition had the highest likelihood. transitions to a different state, such that the ma- Van Rhijn (1984, 1985, 1990) proposed that jority of birds show a different state from the pure male parental care is the primitive con- one that aroseat the baseof the avian lineage. dition in birds. His hypothesisis basedupon These terms are relative, and the hierarchical the premise that "monogamouspaternal care level to which they refer shouldalways be con- systemscan easily evolve towards all recent sidered. mating systemsin birds" (van Rhijn 1984:103), Phylogeniesare hypothesesabout character whereasbiparental caresystems cannot. This is evolution and, as such, they allow the refine- in contrastwith the suggestionby Emlen and ment of ecologicalquestions such as, why does Oring (1977:220)that "complete male parental speciesX exhibit biparental care?With a phy- care is most likely to develop in groups with logeny, one can go beyondthe proximatean- ... a phylogenetic history of shared incuba- swer to this question,namely, "Becauseits an- tion." If van Rhijn's hypothesisis correct,then cestors exhibited biparental care," and seek complete male care would probably have ultimate answersby examiningthe ecological evolved from a condition where incubation was parametersthat may have contributed to the lacking,as in crocodiliansand possiblynonavi- origin of biparentalcare in the variouslineages an dinosaurs;van Rhijn, however, limited his in which it arose.My study is not an attempt discussion to birds. to poseand answer suchquestions, but rather Van Rhijn (1990) attempted to support his to begin to constructa historical framework view primarily by using Cracraft's(1981) clas- within which suchquestions may be explored. sificationof birdsand Strauch's(1978) phylog- The hindlimb musculature of birds has been eny of the order Charadriiformes (shorebirds usedeffectively to generatecladistic hypotheses and allies). Testing van Rhijn's intriguing hy- for restrictedlineages (e.g. Old World subos- pothesis was somewhat hampered at the time cines;Raikow 1987), and has recently been an- by the unavailabilityof cladisticanalyses of avi- alyzed for a large and diversegroup of birds an character data, with which the behavioral (McKitrick 1991). The latter data set was chosen data could be either comparedor combined to to complementthe behavioraldata summarized generatea detailed hypothesisabout the evo- here becauseit is the most taxonomicallycom- lution of parental care in birds. Furthermore, prehensiveset of characterdata available for 830 MiddyC. MCKIT•CK [Auk,Vol. 109 birds. No assumptionsare made here about the number of trees and thereby reduce the number of independenceof the hindlimbmusculature and assumptionsthat were necessaryfor formulatinghy- parental-carebehavior. potheses about character evolution, I examined each of thesegroups of "same"taxa and arbitrarily elim- METHODS inated all but one of the taxa in each group. This resultedin eliminationof 17 taxa,leaving a total of I used published data on 66 charactersfrom arian 60. hindlimb musculature(for details,see McKitrick 1991), A 75%majority-rule consensus of the final treeswas and combined these with 15 behavioral characters obtained(Fig. 1);such a treeshows all groupingsthat taken from the literature. The behaviors were chosen are present in 75% of all trees. It was chosenover the becausethey summarizeparental care in birds and strict-consensustree becausethe latter collapsesall becausethey seemedamenable to relatively unam- groupings that do not occur in 100% of the trees. I biguousinterpretation and coding.All charactersare considerthe majority-rule tree to be a more repre- describedbriefly in Appendix 1. sentativeand informative summaryof relationships. Completeor interpretablebehavioral data were not Becausethe consensustree is a summary, rather alwaysavailable for the sametaxa used by McKitrick than an actual hypothesisof charactertransitions, I (1991). Therefore, in 13 casesa speciesclassified in used one of the actual shortest-lengthtrees (Fig. 2; the same genus was substituted(see Appendix 2). designated Tree 1) to illustrate these transitions. Of Behavioral characters were obtained for 76 arian taxa. 58 nodesin the consensustree (Fig. 1), only four An ancestor based on crocodilians was used as the (6.9%) were unstable; this indicates that all of the outgroup."Dummy" charactersfor paleognath(rat- fundamental(actual) trees were generallyvery sim- ites and tinamous), neognath (all other birds), and ilar. For this reason,using one tree to representall arian monophylywere includedas well. Thesehelped of the treesis a usefulway to illustratepossible char- to ensure that the outgroup would always remain acter transitions. Three optimization routines were outsideof the ingroup (for justification,see McKitrick applied to Tree 1 (Acctran, Deltran, and MinF; see 1991). Furthermore,the analyseswere constrained below). Theseoptimizations are different methodsof (using a "opological Constraints"option in program assigningcharacter states to the interior nodes(hy- used;see below) suchthat paleognathswere the sister potheticalancestors) of the treesgenerated by PAUP. groupto neognaths.Thus, the paleognathswere des- For example, in Figure 2 the terminal taxa Rheaand ignated as the sistergroup to neognaths,and the An- Crypturellusare linked at node 61. For somecharacters, cestorwas includedas the sistergroup to all the birds. there is a transition of character states between node This structure is the basis for outgroup comparison 62 and node 61, and between node 61 and Rhea. The (Ralkow 1982,Maddison et al. 1984) and permits one three optimizationroutines always will yield trees to test hypothesesof synapomorphy;if a character with the same overall number of transitions(steps), state is synapomorphousfor the ingroup, then the but the transitions between nodes 62 and 61, and two outgroups(in this casethe "sister group" and between node 61 and Rhea will not be the same for outgroup) should have the alternative characterstate. each optimization. The Acctran (Accelerated Trans- The combineddata set was analyzed using PAUP formation)algorithm is basedon the assumptionthat 3.0s(Phylogenetic Analysis Using Parsimony;Swof- reversals (0 • 1 • 0) are more common than inde- ford 1990) on a Macintosh IIfx computer.The hind- pendent origins of a characterstate (1 • 0 • 1), while limb charactersalso were analyzedseparately to de- Deltran (Delayed Transformation)assumes the op- termine whether the behavioral charactersactually posite (i.e. independent origins are more common were makinga differencein the resultingtree topol- than reversals). The MinF routine maximizes the ogy. The Heuristic algorithm was used,as it finds the numberof autapomorphies(derived states unique to minimum length trees and is recommendedfor data that terminal taxon) wherever possible.Therefore, sets with more than 20 taxa. The shortest trees were the algorithmsmake different assumptions about evo- found usingthe RandomAddition option becausethe lutionary processes. order in which the taxaare read by the programcan Severalof the charactersused in this analysisare make a significantdifference in whether the shortest relatively unenlighteningbecause little information tree is found. With this option, the taxa are read in is available about them in the literature. Nevertheless, random order 10 different times, and for each repli- I included such characters to call attention to the lack cation I allowed a maximum of 50 trees to be gen- of information.One suchexample is mateguarding erated. I then took one of the resulting shortesttrees by males(character 72). In mostcases, the information and usedthis as a "seed"to completethe analysis(i.e. is simply unavailable(see Birkhead 1979);in others, from thisseed all possibleminimum-length trees were the author mayhave observedmate guarding but did sought).Initial analysisproduced a numberof groups not mention it, or was not explicit about whether it that were consideredby PAUP to be "the same";in occurs.For example, Mock (1979) noted that males other words, these groups contained three or more are very stronglyterritorial during the egg-laying taxa at the same node. In order to reduce the final stage in Ardea herodias;however, he did not refer to October1992] PhylogenyofAvian Parental Care 831

Ancestor Ancestor Rhea Rhea •-E• crypturellus t;2.['61 C Crypturellus • Nothoprocta -- Nothoprocta -- Dapt•n Daption

Oceanires Oceanires Fregotta -- Oceanodroma -- Ocean•droma Pygoscelis Eudyptula Eudyptula Gayla Pcdiceps -- Chen r8c GaviaChen

Copphus ] • Cepphus • Ptychoramphu$ • Cororhinca ß76 •'- FraterculaCororhinca __• Rynchops r• Rynchops 8o 79-[• Stoma L LarusStercoranus 118 Grus Fulica

Gymnogyps -8? E CoragypsGymnogyps Mycteria [-91• Ciconia • Plegadis .96½5 ½ ButoridesFlorida Acc•iter 113 • ..•? .• Acc•iter

103 -- Eulampls [111J 107J ,lo4r--Colaptos

/I ' I I '-• Co•,us 06%1 I r-- coccyzus I ' { G•coccyx ] I10•,n•---- Geococcyx

[ E o•xu•a

Fig. 1. A 75%majority-rule consensus of 63 trees Fig. 2. Tree 1 of 63 basedon hindlimb muscleand based on hindlimb muscle and behavioral data; all behavioral data. Internal nodes are numbered. groups occurred in 100% of these trees unless otherwise indicated. however, a specieshad female-only care in most re- spects,but malescontributed defense. In these cases, thisbehavior as mate guarding and, therefore,I coded character75 wascoded as biparental, while the other that characterwith "?" (missing)for that species. care-relatedcharacters were coded as female-only. A growing number of apparently monogamous Character 76 (distinguishablereversed sexual di- specieshave been shown to exhibit multiple parent- morphism)was coded in a nontraditionalmanner for agein their broods(for a review, seeMcKitrick 1990). owls. Although owls are consideredto exhibit re- Unfortunately, little information on actualparentage versedsexual dimorphism (Mueller 1986),I did not is available for most of the primarily nonpasserine so code the two owls included in this analysis.The speciesincluded in the present analysis.Extra-pair ratio of male wing length to femalewing length is copulationshave been reported in several of these 0.975in Otusasio and 0.952in Bubovirginianus (Mueller species(see Gladstone 1979, Ford 1983).Because the 1986:table 1); these numbers are so close to 1.000 that speciesexhibiting the behaviorappear to be primarily classifyingthese species as showing reversedsexual monogamous,I have coded species showing multiple dimorphism seemed unwarranted. parentageor extra-pair copulationsas "0" (monoga- Character84 (posthatchingdevelopment) was coded mous) for character 70. basedon the classificationsystem of Winklet and Wal- Character75 (defenseof young/parentalcare) is in ters (1983), but their three categoriesof precociality effecta summaryof parental carein birds. Generally, were lumpedinto one, as were their two categories if a specieswas biparental for other aspectsof careit of altriciality.Taxa whose classification was specula- wasbiparental for character75 aswell. Occasionally, tive were codedas having that speculativecondition. 832 MARYC. MCKITRICK [Auk, Vol. 109

Ancestor Rhea hypothesizedcharacter transitions described in F CryptttrellusAppendix 3 are basedon this tree. For the analysesbased on hindlimb data only, -- Ocean•droma 161 trees of 247 stepswere found (consistency Fregetta index excluding uninformative characterswas Eudyptula 0.357). A 75% majority-rule consensusof these Gayla trees is shown in Figure 3.

-- then Alca DISCUSSION Brachyramphus ] • UriaCepphus L-• ' Ptychoramphus The consistencyindices are higher than ex- • Cetorhinca • Fratercula pected for this number of taxa (consistencyin- • StemaRynchop$ dex of 0.30 expectedfor 60 taxa; the expected - value decreaseswith increasingnumbers of taxa Ardea basedon empirical analysisby Sandersonand Butorioles FIodda Donoghue 1989). The level of homoplasy, therefore, is not high for a data set of this size. • Coragyps • Gymnogyps However, many nontraditional groupsappear Grus Fulica in these analysescompared with the topology r•__[• MycteriaLepfopt#os of the trees obtainedby McKitrick (1991) using r-• I Ciconia • Phoeniconwus hindlimb muscle data for a larger number of taxa. For example,whereas auks are monophy- • -89%CathartesDendragapus letic and gulls and terns are as well, the two Falco groups do not cluster together. Auks (Fig. 2, • ccpiternode 77) cluster with procellariiforms, pen- Chaetura guins, loonsand grebes(node 71). The results Eulampis of the analysisof morphologyplus behavior are Tyrannus very similar to thosefrom the analysisof mor- Chordeiles phologyalone, differing primarily in the place- Coccyzus Geococcyx ment of grouse,ducks, herons, and the hoatzin. CrotOphaga Zenaida This suggeststhat neither dataset is biasingthe Opisthocomus resultsunduly, and that both data setscontain Fig. 3. A 75%maiority-rule consensus tree of 161 a comparabledegree of historicalinformation. treesbased on hindlimb muscledata only; all groups I suggestthat the present results can be re- occurred in 100% of these trees unless otherwise in- gardedas a startingpoint, an exerciseillustrat- dicated. ing how comparativephylogenetic data may be employedto understandthe evolutionof char- SeeAppendix 1 for a completelist of charactersused acter suites(such as behaviors involved in pa- in the analysis.The characterdata are shown in Ap- rental care). As more data becomeavailable, the pendix 2. resultsof such analyseswill be more represen- tative of the true phylogeny and of character evolution as well. RESULTS The resultsof suchanalyses may be appraised I found 63 shortest-lengthtrees of 339 steps from two perspectives.The trees may be con- based on muscle and behavioral data; the con- sidered as hypothesesabout the phylogenetic sistencyindex (ci) excluding uninformative relationshipsamong the taxa under consider- characters was 0.353. Uninformative characters ation. The trees also may be consideredas hy- are those that show only one transition at a pothesesabout the evolution of the characters terminal taxon (i.e. they are autapomorphies). themselves,although some authorswould dis- A 75% majority-rule consensustree, showing agree(e.g. Brooksand McLennan 1991:63,1991: the percentageof the 63 trees in which each 141).These hypotheses are constrainedby one's group appears,is depictedin Figure 1. Figure starting assumptions,namely, that ratites are 2 shows the first of the 63 trees (Tree 1); the the sistergroup to neognathsand crocodilians October1992] PhylogenyofArian Parental Care 833 are the sistergroup to birds.Appendix 3 shows Accordingto Acctran,female-only care (state 0) the hypothesizedchanges in the behavioral isprimitive for birds,with biparentalcare (state charactersusing three optimization routines. 1) arising near the baseof the avian lineage,at The hypothesesare basedon Tree 1 of 63. A node 113 (neognathsminus the duck-grouse few of these are described below. group).There is onetransition to male-onlycare, For character70 (mating system),monogamy in ratites,and one reversalto female-only care, (state0) is primitive within birds (as has been in Eulampis.With Deltran, no-careis primitive generallyassumed) when Acctranis used;Del- for birds, with a transition to biparental care at tran and MinF show polygyny (state1) at the node 113, one to male-only care in ratites,and baseof the avian lineage with a transition to two to female-only care, in the duck-grouse monogamyoccurring near the basalnode (at groupand in Eulampis.With MinF, biparental node 117, with only ratitesexcluded from this care is primitive, and male-onlycare arisesin clade). Acctran shows three independent ori- ratites, while female-only care arises in the gins of polygyny,while MinF showsone. All duck-grousegroup and in Eulampis. optimizationsshow two independentorigins of For character81 (postfledgingcare), all op- promiscuity(state 3). timizationsyield identicalhypotheses. Biparen- Character75 (defenseof young/parentalcare) tal care (state 1) is primitive within birds. In is of particularinterest, as it summarizesthe addition, there is one transition to female-only roles of the sexesin caring for the young. All care (at the duck-grouseclade), two to male- three optimizationsindicate that someform of only care(at nodes62, 73 and in Zenaida),and biparentalcare (defense of young) is primitive three to no-care (at node 78 and in Florida and for birds. Chaetura). Character transitions for character 78 (incu- For character82 (group breeding), state 1 of bation) are complex. According to Acctran, a character82 has arisen unambiguouslythree transition from no incubation (state 3; croco- times (in Chaetura,Corvus, and Crotophaga). dilians do not incubate) to female-only incu- For character84 (posthatchingdevelopment), bation (state 0) occurs at the base of the avian all transitionsare unambiguous.Precociality is lineage (between nodes 117 and 113). Female- primitive (state0), and altriciality (state1) arose only incubation occursin the lineage at node in the large clade (node 113) that containsall 116, the duck-galliform clade. Biparental in- birds exceptducks, grouse, the hoatzin, and rat- cubation (state 1) is a synapomorphy(shared- ites. Altriciality was subsequentlylost three derived character state) for the remainder of the times: at node 85, and in Phoeniconaiusand Chor- neognaths(see Methods) and, therefore, prim- deiles. itive within this group. Male-only incubation Silver et al. (1985) used canonical-correlation (state 2) arises once, at the base of the ratite analysisto determinewhich ecologicaland life- lineage (node 62). In all, female-only incuba- history parameterswere the best predictorsof tion arises four times. Deltran represents no- various aspectsof avian paternal care. Their re- incubationas being the primitive condition for suits indicated that mode of posthatching de- birds and, thus, female-only incubation is de- velopment, mating system,some habitat char- rived for the duck-galliform clade rather than acteristics,and relative clutch massexplain the primitive. Biparental incubation is a synapo- mostvariance in paternal-careactivities. I used morphy for the remainder of the neognaths, the phylogenetic framework of the present and male-only incubation is a synapomorphy analysisto test the hypothesisof Silver et al. for the ratites.Female-only care arises a total of (1985)that the evolutionof paternalcare is caus- six times. Finally, MinF representsbiparental ally correlatedwith the occurrenceof altricial- incubation as a synapomorphy for birds and, ity. thus, it is primitive within that lineage.Again, The more data that becomeavailable for phy- female-onlyincubation is derived for the duck- logenetic analysis,the more ways we can invent galliform lineage,and for five other lineagesas for studying characterevolution; unfortunate- well. Male-only incubationis derived for rat- ly, the invention of algorithmsfor testing hy- ites. pothesesabout characterevolution necessarily For character80 (posthatchingcare), the three lagsbehind. At present,no algorithmsare avail- optimizationseach lead to differenthypotheses. able for testing significanceof associationbe- 834 MARYC. MCKITRICK [Auk,Vol. 109

I examinedthe distribution of statechanges in character80 (posthatchingcare) with respectto the occurrenceof an independent character(see below), and receded the character as presence or absenceof paternal care. For example, "biparental care" became "paternal care present"; "female-only care" became "paternal care absent." The origin of female-only or no parental care subsequentto the origin of biparental or male-only care was interpreted as a loss of paternal care. Figure 4 shows the state changes for character80 before receding;Figure 5 shows the receded transitions for this character. Of the predictorsstudied by Silveret al. (1985), posthatchingdevelopment and mating system were the only ones analyzed in the present study.Of thesetwo, posthatchingdevelopment wasthe only one for which hypothesizedchar- actertransitions were identical for all three ep- timizatiens on Tree 1. I traced the occurrence Colapres of the derived state of this character,namely altriciality, onto that tree (Figs. 4 and 5). Its occurrence is referred to as the black area of the tree. To testthe hypothesisof Silver et al. (1985), I wished to determine the probability of obtaining as many or more gains in the black and Fig. 4. Tree 1 showing occurrenceof altriciality as many or fewer lossesin the black simply by (heavy black lines) and statetransitions for character chance. In other words, does the black area at- 80 (posthatchingcare of young). tract gains and repel losses?Should the occurrenceof altricialitylead to the origin of paternal tween multistate characters. Maddison (1990) care in character 80? The observed pattern of presented an algorithm to test associationbe- gainsand lossesin the black areawas zero gains tween binary characters;that is, to determine and one loss (Fig. 5), and the probability of the likelihood of an observednumber of gains obtaining this pattern by chance is 0.814. In and lossesin a "dependent" variable occurring other words, the pattern is not significant and by chancewith respectto the occurrenceof an altriciality doesnot appearto lead to the origin independent variable. This test is designednot of paternal care.The implicit assumptionof Sil- to study correlation of change in one variable ver et al. (1985) was that female-only care is with changein another, as other methodshave primitive for birds, whereas in fact biparental done (e.g. Felsenstein1985), but rather to test care is primitive (and widespread;Fig. 4). Evi- the significanceof the origin of the dependent dently, factorsother than altriciality were the variable anywhere subsequentto the origin(s) selectiveagent for the origin of paternal care of the independent variable on the tree. In oth- in birds. er words, it teststhe null hypothesisthat gains Emlen and Oring (1977) hypothesized that and lossesof the dependent characterare ran- "complete male parental care" (which ! inter- domly distributed on the tree with respectto pret to mean male-only care) arosein taxa with the state of the independent character. This a shared history of incubation. To examine this makesthe test ideal for studying evolutionary hypothesis,I mapped onto Tree 1 the occur- causalityin correlationsamong character states. rence of shared incubation (character78), as in- Maddisen (1990) suggestedways to recede dicated by Acctran optimization (Fig. 6), and multistate charactersas binary charactersfor the origin of male-only care for character 81 purposesof this test,and he further pointed out (pestfledgingcare). The map shows an early that the testis probablymore conservativewhen origin of shared incubation within birds, with the characters tested were used to build the trees. four out of five originsof male-only carearising October1992] PhylogenyofAvian Parental Care 835

A•stor Rhea NothO•octa

Cha•ura

CHARACTER 80 RECODED

Fig. 5. Tree 1 showing character 80 recoded. Fig. 6. Tree 1 showing origin and loss of shared incubationand origin and lossof male-only care for character81 (postfledgingcare). in taxawith a history of sharedincubation. One origin of male-only care occurred in a group (ratites)with no history of parental incubation. pothesesbased on phylogeneticanalysis of Thus,Emlen and Oring's (1977)hypothesis ap- characterdata and then examinethe possible pearsto be largely supported.However, Mad- evolutionarypathways revealed by the data. dison's(1990) test indicatesthat the probability Furthermore,separating parental-care behavior of obtaining four or more gains in the black into component parts is more useful than ex- area is 0.757. This is partly due to the fact that amining parental care as a unit, as separation shared incubationis so widely distributed on revealsthe decouplingof someaspects of care. the tree. A higher representationof members For example,care of nestlingsmay be biparen- of Passeriformesin the tree might well change tal, while care of fledglingsis uniparental,as this result, as many passerinesdo not have in some auks. shared incubation. Drawbacksof the phylogeneticanalysis.--Sev- Basedon the datapresented here, van Rhijn's eral problemsare inherent in a study of this (1990) hypothesisthat male-only parental care kind. Immelmann (1974) noted that certain is primitive within birds is incorrect. I believe kinds of behavioral characters are not "taxo- that van Rhijn's approachhas merit in that it nomically"useful because of their plasticityin attemptsto place the study of behavioral evo- responseto environmental variables. However, lution within a phylogeneticframework, where hypothesesof convergencecan only be made it belongs.However, his proposalsregarding within the contextof a phylogenetichypothe- the direction of evolution are ad hoc in that sis;in otherwords, one cannot determine a priori they are ahistorical,are made outside the con- whether a given characterwill be phylogenet- text of an explicit phylogenetichypothesis, and icallyinformative. The meanconsistency index are basedon somenotion of probability and the was not significantly different for the muscle irreversibility of certain evolutionary path- characters(0.542) and the behavioral characters ways. It is more meaningful to formulate hy- (0.420; Kruskal-Wallis, k = 1.095, df = 1, P = 836 MARYC. MCKITRICK [Auk, Vol. 109

0.295), suggestingthat the two kinds of char- for future researchon avian mating systems. actershave equal potential to be phylogeneti- One of thesewas to ask why malesin monog- cally informative. amoussystems contribute so much parentalcare. Throughphylogenetic analysis one may learn My analysismakes it possibleto begin to try to more about the tendencies of certain characters answersuch a question.For example,if we know to originate in many taxa. Even if a character how many times male-only postfiedging care exhibits homoplasy,however, it still may be arosefrom biparentalcare, and how many times informative at some lower taxonomic level. If, the absenceof postfiedgingcare arose,we can however, these environmental variables cause make comparisonsof ecologicalvariation across intraindividual variation in parental-care be- taxaand form useful hypothesesabout the con- havior, which they undoubtedlydo (Emlen and ditions leading to the evolution of suchpatterns Oring ! 977), then the classificationof behaviors of care. usedin the presentstudy is probablysomewhat Although my study is to date the mosttaxo- inaccurate. nomically comprehensiveattempt to analyze The classificationof behaviormay be complex phylogeneticallyavian parental-caredata, it is (Shields 1984), and the classificationof mating neverthelesslimited in its scopeby the avail- systemsin particular can be cumbersome(Or- ability of information. Of great interest, for ex- ing 1982, Mock 1983). Consequently,coding ample, would be a cladisticanalysis of the Cha- charactersmay fail to characterizethe behaviors radriiformes, a group within which parental- adequately, even though much attention was care behavior varies widely. At present there paid to this in the present analysis. Further- are no comparabledata on the limb musculature more, even though both sexesmay perform a for the majority of this group, in particular the certain carebehavior, there may be quantitative sandpipers(Scolopacidae), so very few repre- or qualitative differencesin the care provided sentatives of the order are included here. Mor- by each (Winkler 1987); this is not reflected in phologicalwork on that group is in progress the presentqualitative codingsystem. There may (McKitrick unpubl. data, P. Chu unpubl. data) be intraspecific geographic variation in paren- and, eventually, these data will be combined tal care (e.g. Agelaiusphoeniceus; Payne 1969, with behavioraldata for a detailedphylogenetic 1979) that is not accountedfor in the present analysis.Also, few passerinespecies were in- coding scheme. cluded in the current analysis,despite the in- For at least two reasons, I have attempted to teresting variation in their parental-carebehav- be conservative in my interpretation of hy- ior, because their hindlimb musculature is pothesesyielded by the phylogeneticanalyses relatively uninformative phylogenetically(e.g. in this study. One reasonis that some of the Raikow 1978, McKitrick ! 985). For the charadri- phylogenetic relationshipssuggested by these iforms and passetines,the behavioral data base analysesdiffer markedly from all others pro- is richer than the morphologicalone. Collecting posed in the literature, including those based the original data is a painstaking process;col- on the same data but with more taxa included lating the data from the published literature is (McKitrick 1991). If the relationships as pre- fraught with difficulties, such as the danger of sented here are incorrect, as at least some un- making errorsin interpreting the work of many doubtedly are, then the hypothesesabout char- different researcherswho have brought differ- acter transformations are also incorrect. g second ent methods, assumptions,and biasesto their reason, related to the first, is that the addition work. Furthermore, a morphologistis likely at or deletion of characters and taxa has an im- times to misinterpret behavioral studies. portant effect on tree topology and, as more Brooks and McLennan (1991:342) noted that information becomes available for the taxa ex- "We have discoveredthat there is a paucity of amined in this study or subsetsof these taxa, rigorous phylogenetic hypothesesfrom which more meaningful statements about character we can begin historical ecologicalstudies .... " evolution will be possible. This is an extreme understatement, and in ex- A phylogenetic approachcan yield hypoth- pressing it the authors might well have men- esesabout the pattern of origin of behavioral tioned alsothe paucity of comparativedata sets characters, so that the appropriate kinds of for the same group of taxa. What I have pre- questionscan be asked about these patterns. sented here is an attempt to begin to remedy Mock (1985) suggestedseveral lines of inquiry this for birds. I am optimistic that future work- October1992] PhylogenyofAvian Parental Care 837 ers will continue to gather anatomical, behav- the Common Murre Uria aalge.Anim. Behav.33: ioral and other "phenotypic" data (including 608-619. molecular) for a wide variety of taxa. It is my BIRKHEAD,T. R., AND C. M. LESSELLS.1988. Copula- hope that these workers will gather such data tion behaviour of the Osprey Pandionhaliaetus. Anim. Behav. 36:1672-1682. with the goal of using them for analyseslike BJ•RKLUND,M. 1990. A phylogeneticinterpretation the present one, analysesthat will be greater of sexualdimorphism in body sizeand ornament both in scopeand in detail. Through such work in relation to mating systemin birds. J. Evol. Biol. we have an excellent chanceof formulating rig- 3:171-183. oroushypotheses about phylogenetic relation- BOSWALL,J., ANDD. MAClvER. 1975. The Magellanic ships among birds and about the evolution of Penguin Spheniscusmagellanicus. Pages 271-305 in complexavian charactersystems. The biology of penguins (B. Stonehouse,Ed.). University Park Press,Baltimore. BROOKS,D. R., •3'•DD. A. MCLENN•3'•. 1991. Phylog- ACKNOWLEDGMENTS eny, ecology, and behavior: A researchprogram I thank Philip Chu and William Fink for much in comparative biology. Univ. Chicago Press, helpful discussionof phylogeneticanalysis, and Phil Chicago. Myers for advice on statistics.Steve Beissinger,Phil BROWN,L. H., AND A. ROOT. 1971. The breeding Chu, Raymond Pierotti, Gary Schnell, Sarah Sloane, behaviour of the LesserFlamingo Phoeniconaius minor. Ibis 113:147-172. David Winklet, and an anonymous reviewer criti- cized earlier drafts of the manuscriptand provided BRUNING,D. F. 1974. Social structure and reproduc- many useful suggestions.Wayne Maddison gener- tive behavior in the Greater Rhea. Living Bird 13:251-294. ously assistedwith application of his correlation test. Robert Storer provided information on grebes,Ray- BURGER,J. 1981. Sexual differences in parental ac- mond Pierottiprovided information on gullsand auks, tivities of breeding Black Skimmers. Am. Nat. 117:975-984. and Cynthia Simsprovided referenceson Corvus.This work was supportedby grant BSR-9006208from the BURGER,J., L. M. MILLER, AND D.C. HAHN. 1978. National Science Foundation. Behavior and sex roles of nesting Anhingas at San Bias, Mexico. Wilson Bull. 90:359-375. BURNS,F.L. 1900. A monograph of the Flicker (Co- LITERATURE CITED laptesauratus). Wilson Bull. 12:1-82. AINLEY, D. G., R. E. LERESCHE,AND W. J. L. SLADEN. BUTLER,R. G., AND S. JANES-BUTLER.1983. Sexual differences in the behavior of adult Great Black- 1983. Breeding biology of the Ad•lie Penguin. Univ. California Press,Berkeley. backed Gulls (Larusmarinus) during the pre- and AINLEY, D. G., AND g. P. SCHLATTER. 1972. Chick- post-hatchingperiods. Auk 100:63-75. raising ability in Ad•lie Penguins.Auk 89:559- CLUTTON-BROCK,T. H., AND P. H. HARVEY. 1977. Pri- 566. mate ecology and social organization. J. Zool. ALLEN,A.A. 1924. A contributionto the life history (Lond.) 183:1-33. and economic status of the Screech Owl (Otus CLUTTON-BROCK, T. H., AND P. H. HARVEY. 1979. asio). Auk 41:1-16. Comparison and adaptation. Proc. R. Soc. Lond. Set. B Biol. Sci. 205:547-565. ASHCROFT,R. 1979. Survival ratesand breeding biology of Puffins on Skomer Island, Wales. Ornis CODDINGTON,J. g. 1988. Cladistic tests of adapta- Scand. 10:100-110. tional hypotheses.Cladistics 4:3-22. BAYNARD,O.E. 1913. Home life of the GlossyIbis COLES,V. 1944. Nesting of the Turkey Vulture in (Plegadisautumnolis Linn.). Wilson Bull. 25:103- Ohio caves. Auk 61:219-228. 117. COOKE, F., M. A. BOUSFIELD,AND A. SADURA. 1981. BECK,J. R., AND D. W. BROWN. 1971. The breeding Mate change and reproductive successin the biologyof the Black-belliedStorm-Petrel Fregetta Lesser Snow Goose. Condor 83:322-327. tropica.Ibis 113:73-90. CRACRAET,J. 1981. Toward a phylogenetic classifi- BENT, A. C. 1940. Life histories of North American cation of the Recent birds of the world (class cuckoos,goatsuckers, hummingbirds and their Aves). Auk 98:681-714. allies. U.S. Natl. Mus. Bull. 176. CRACRA•T,J. 1986. The origin and early diversifi- BERMAN, S.L. 1984. The hindlimb musculature of cation of birds. Paleobiology 12:383-399. the White-fronted Amazon (Amazonaalbifrons, CRACRAFT,J. AND D. P. MINDELL. 1989. The early Psittaciformes). Auk 101:74-92. history of modern birds: A comparisonof mo- BIRKHEAD,T.R. 1979. Mate-guarding in the Magpie lecularand morphologicalevidence. Pages 389- Pica pica. Anim. Behav. 27:866-874. 403 in The hierarchy of life (B. Fernholm, K. Bre- BIRKHEAD,T. R., S. D. JOHNSON,AND D. N. NETTLESHIP. met and H. J6rnvall, Eds.). Elsevier SciencePub- 1985. Extra-pair matings and mate guarding in lishers, Amsterdam. 838 MARYC. McKITRICK [Auk, Vol. 109

DALKE, P. D., A. S. LEOPOLD, AND D. L. SPENCER. 1946. HARRIS,M.P. 1966. Breeding biology of the Manx The ecologyand managementof the Wild Turkey ShearwaterPuffinus puffinus. Ibis 108:17-33. in Missouri. Conserv. Comm., Federal Aid-Wild- HARRIS,M.P., AND T. R. BIRKHEAD.1985. Breeding life Prog., Tech. Bull. No. 1. ecologyof the Atlantic Alcidae.Pages 156-204 in DERKSON,D.V. 1977. A quantitative analysisof the The Atlantic Alcidae (D. N. Nettleship and T. R. incubationbehavior of the Ad•lie Penguin.Auk Birkhead, Eds.). Academic Press, London. 94:552-566. HARVEY,P. H., ANDG. M. MACE. 1982. Comparisons DEXTER,R. W. 1952. Extra-parental cooperation in between taxa and adaptive trends: Problems of the nestingof Chimney Swifts. Wilson Bull. 64: methodology.Pages 343-361 in Current prob- 133-139. lems in sociobiology(King's College Sociobiol- DEXTER,R. W. 1961. Further studieson nesting of ogy Group, Eds.). Cambridge Univ. Press,Cam- the Common Nighthawk. Bird-Banding32:79- bridge. 85. HATCH,S. A. 1987. Adult survival and productivity DOBSON,F. S. 1985. The use of phylogeny in be- of Northern Fulmars in Alaska. Condor 89:685- havior and ecology.Evolution 39:1384-1388. 696. DRENT,R.H. 1965. Breedingbiology of the Pigeon HATCH, S.A. 1990. Individual variation in behavior Guillemot, Cepphuscolumba. Ardea 53:99-160. and breedingsuccess of Northern Fulmars.Auk ELLISON,L.N. 1973. Seasonalsocialorganizationand 107:750-755. movements of Spruce Grouse. Condor 75:375- HATCHWELL,B.J. 1988. Intraspecificvariation in ex- 385. tra-pair copulationand matedefence in Common EMLEN,S. T., ANDL. W. ORING. 1977. Ecology,sexual Guillemots Uria aalge.Behaviour 107:157-185. selection,and the evolution of mating systems. HAVERSCHMIDT,F. 1949. The life of the White Stork. Science 197:215-223. E. J. Brill, Leiden. ERPaNGTON,P. L. 1932. Studies on the behavior of HODGEs,A.F. 1969. A time lapsestudy of Kittiwake the Great Horned Owl. Wilson Bull. 44:212-220. incubationrhythms. Ibis 111:442-443. FELSENSTEIN,J. 1985. Phylogeniesand the compar- HOGLUND,J. 1989. Size and plumage dimorphism ative method. Am. Nat. 125:1-15. in lek-breeding birds: A comparativeanalysis. FINNEY,G. H., AND F. COOKE. 1978. Reproductive Am. Nat. 134:72-87. habits in the Snow Goose: The influence of fe- HUNT, R.H. 1975. Maternal behavior in the More- male age. Condor 80:147-158. let's crocodile, Crocodylusmoreleti. Copeia 1975: FISCHER,R. B. 1958. The breeding biology of the 763-764. Chimney Swift, Chaeturapelagica (Linnaeus). N.Y. IMBER,M. J. 1987. Breedingecology and conserva- State Mus. Sci. Serv. Bull. 368:1-141. tion of the BlackPetrel (Procellariaparkinsoni). No- FISHER,H. I. 1971. The Laysan Albatross:Its incu- tornis 34:19-39. bation, hatching,and associatedbehaviors. Liv- IMMELMANN, K. 1974. The use of behavioural char- ing Bird 10:19-78. acters in avian classification.Page 206 in Pro- FITCH, H. S., F. SWENSON,AND D. F. TILLOTSON. 1946. ceedings16th InternationalOrnithological Con- Behavior and food habits of the Red-tailed Hawk. gress(H. J.Frith and J.H. Calaby,Eds.). Canberra, Condor 48:205-237. 1974. Australian Academy of Science,Canberra. FOLSE,L. J.,JR., AND K. A. ARNOLD.1978. Population JOHNSGARD,P. A. 1961. The taxonomy of the Anat- ecologyof Roadrunners(Geococcyx californianus) idae--A behaviouralanalysis. Ibis 103a:71-85. in South Texas. Southwest. Nat. 23:1-28. JOHNSON,R. A. 1941. Nesting behavior of the At- FORD,N.L. 1983. Variation in mate fidelity in mo- lantic Murre. Auk 58:153-163. nogamousbirds. Curr. Ornithol. 1:329-356. JOYNER,D.E. 1977. Behaviorof Ruddy Duck broods GEORGE,J. C., ANDA. J. BURGER.1966. Avian myol- in Utah. Auk 94:343-349. ogy. AcademicPress, New York. KAHL,M.P. 1966. A contributionto the ecologyand GITTLEMAN,J. L. 1981. The phylogeny of parental care in fishes. Anita. Behav. 29:936-941. reproductivebiology of the MarabouStork (Lep- GLADFELTER,H. L., AND R. S. McBuP,NEY. 1971. Mat- toptiloscrumeniferos) in East Africa. J. Zool. (Lond.) 148:289-311. ing activityof RuffedGrouse. Auk 88:176-177. GLADSTONE,D.E. 1979. Promiscuityin monogamous KAHL,M.P. 1972. Comparativeethology of the Cico- colonial birds. Am. Nat. 114:545-557. niidae.The woodstorks(genera Mycteria and Ibis). Ibis 114:15-29. GooD, E. E. 1952. The life history of the American Crow (Corvusbrachyrhynchos Brehm). Ph.D. dis- KEITH, L. B. 1977. Great Horned Owl. Can. Wildl. sertation, Ohio State Univ., Columbus. Serv. Dep. Supply & Services,Ottawa. GOODBURN,S.F. 1984. Mate guardingin the Mallard KILHAM, L. 1966. Reproductive behavior of Hairy Anasplatyrhynchos. Ornis Scand.15:261-265. Woodpeckers.I. Pair formation and courtship. GREEN,R. 1976. Breedingbehaviour of OspreysPan- Wilson Bull. 78:251-265. dion haliaetus in Scotland. Ibis 118:475-490. KILHAM,L. 1968. Reproductivebehavior of Hairy October1992] PhylogenyofAvian Parental Care 839

Woodpeckers.II. Nestingand habitat.Wilson Bull. comparativeethology and phylogenetic system- 80:286-305. atics:A casestudy using gasterosteidfishes. Can. KILHAM,L. 1969. Reproductive behavior of Hairy J. Zool. 66:2177-2190. Woodpeckers.III. Agonisticbehavior in relation MOCK,D.W. 1979. Displayand repertoireshifts and to courtship and territory. Wilson Bull. 81:169- "extra-marital" courtship in herons. Behaviour 183. 69:57-71. LACK,D. 1968. Ecologicaladaptations for breeding MOCK,D. W. 1983. On the study of avian mating in birds. Methuen & Co., London. systems.Pages 55-84 in Perspectivesin ornithol- LANCASTER,D.A. 1964. Life history of the Boucard ogy (A. H. Brush and G. A. Clark, Eds.). Cam- Tinamou in British Honduras Part II: Breeding bridge Univ. Press,Cambridge. biology. Condor 66:253-276. MOCK,D.W. 1985. An introduction to the neglected LANK, D. B., P. MINEAU, R. F. ROCKWELL,AND F. COOKE. mating system. Pages 1-10 in Avian monogamy 1989. Intraspecificnest parasitism and extra-pair (P. A. Gowaty and D. W. Mock, Eds.). Ornithol. copulation in LesserSnow Geese.Anim. Behav. Monogr. No. 37. 37:74-89. MOLLER,A. P. 1987. Copulation behaviour in the LEBRET,T. 1961. The pair formation in the annual Goshawk,Accipiter gentills. Anim. Behav.35:755- cycleof the Mallard (Anasplatyrhynchos L.). Ardea 763. 49:97-158. MosnY, H. S., AND C. O. HANDLEY. 1942. The Wild LEFEBVRE,E.A. 1977. LaysanAlbatross breeding be- Turkey in Virginia. Its habits and management. havior. Auk 94:270-274. EducationalDivision, Virginia Comm. Gameand LEWIS, R.A. 1985. Do Blue Grouse form leks? Auk Inland Fisheries, Richmond. 102:180-184. MUELLER,H.C. 1986. The evolution of reversed sex- LORENZ,K. 1941. VergleichendeBewegungstudien ual dimorphismin owls: An empirical analysis an Anatinen. J. Ornithol. 89:194-294. of possibleselective factors. Wilson Bull. 98:387- MADDISON,W. P. 1990. A model for testingthe cor- 406. related evolution of two binary characters:Are NELSON,J. B. 1966. The breedingbiology of the Gan- gains or lossesconcentrated on certain branches net, Sula bassana,on the Bass Rock, Scotland. Ibis of a phylogenetictree? Evolution 44:539-557. 108:584-624. MADDISON,W. P., M. J. DONOGHUE,AND D. R. MAD- NESBITT,S.A. 1989. The significanceof mate lossin DISON. 1984. Outgroup analysisand parsimony. Florida Sandhill Cranes. Wilson Bull. 101:648- Syst. Zool. 33:83-103. 651. MAHER,W.J. 1974. Ecologyof Pomarine, Parasitic NOBLE-ROLLIN,D. 1975. Daily behaviourof the White and Long-tailedjaegers in northern Alaska. Pa- Stork. Dawn Song and All Day (World Bird Re- cific Coast Avifauna, No. 37. searchStation, Glanton, Northumberland) 2:23- MAUNDER,J. E., AND W. THRELFALL.1972. The breed- 28. ing biology of the Black-legged Kittiwake in O'DONALD,P. 1983. The Arctic Skua,a studyof the Newfoundland. Auk 89:789-816. ecologyand evolutionof a seabird.Cambridge MCDONALD,S. D. 1970. The breeding behavior of Univ. Press,Cambridge. the Rock Ptarmigan.Living Bird 9:195-238. OHMART,R. D. 1973. Observationsof the breeding MCINTYRE,J.W. 1988. The Common Loon. Spirit of northern lakes. Univ. Minnesota Press, Minne- adaptationsof the Roadrunner. Condor 75:140- 149. apolis. MCKINNEY,F. 1986. Mate guarding and the confi- ORING,L. 1982. Avian mating systems.Pages 1-92 dence of multiple paternity in waterfowl. Pages in Avian biology,vol. 6 (D. S. Farher,J. R. King, and K. C. Parkes, Eds.). Academic Press, New 411-419 in Acta XIX CongressusInter•ationalis York. Ornithologici (H. Ouellet, Ed.). Ottawa, Ontario, 1986. National Museum of Natural Science, Ot- PAGEL, M.D., AND P. H. HARVEY. 1988. Recent de- tawa. velopmentsin the analysisof comparativedata. McKITRICK,M.C. 1985. Pelvicmyology of the king- Q. Rev. Biol. 63:413-440. birds and their allies (Aves: Tyrannidae). Ann. PALMER,R. S. (Ed.). 1962. Handbook of North Amer- Carnegie Mus. 54:275-317. ican birds, vol. 1. Yale Univ. Press, New Haven. MCKITRICK, M. C. 1990. Genetic evidence for mul- PALMER,R. S. (Ed.). 1988a. Handbook of North tiple parentagein EasternKingbirds (Tyrannus American birds, vol. 4. Yale Univ. Press, New tyrannus).Behav. Ecol. Sociobiol. 26:149-155. Haven. MCKITRICK,M.C. 1991. Phylogeneticanalysis of avi- PALMER, R. S. (Ed.). 1988b. Handbook of North an hind limb musculature. Misc. Publ. Mus. Zool. American birds, vol. 5. Yale Univ. Press, New Univ. Mich., No. 179. Haven. MCLENNAN, D. A., D. R. BROOKS,AND J. D. MCPHAIL. PAYNE,R. B. 1969. Breedingseasons and reproduc- 1988. The benefits of communication between tive physiology of Tricolored Blackbirds and 840 MAgYC. McKrrRIC•< [Auk,Vol. 109

Redwinged Blackbirds.Univ. Calif. Publ. Zool. SANDERSON,M. J., AND M. J. DONOGHUE. 1989. Pat- 90:1-115. terns of variation in levels of homoplasy.Evo- PAYNE, R. B. 1979. Sexual selection and intersexual lution 43:1781-1795. differencesin varianceof breedingsuccess. Am. SEALy,S. G. 1974. Breedingphenology and clutch Nat. 114:447-452. size in the Marbled Murrelet. Auk 91:10-23. PEARSON,A. K., AND O. P. PEARSON. 1955. Natural SHERMAN,A. 1911. Nest life of the Screech Owl. Auk history and breeding behavior of the tinamou, 28:155-168. Nothoproctaornata. Auk 72:113-127. SHIELDS,W. M. 1984. Barn Swallow mobbing: Self PERDECK,A.C. 1963. The early reproductivebehav- defence,collateral kin defence,group defence, iour of the Arctic Skua,Stercorarius parasiticus (L.). or parental care?Anim. Behav.32:132-148. Ardea 51:1-15. SIEGFRIED,W. R. 1976. Socialorganization in Ruddy PITELKA,F. A., P.O. TOMICH, AND G. W. TREICHEL. and Maccoa ducks. Auk 93:560-570. 1955. Breedingbehavior of jaegersand owlsnear SILVER, R., H. ANDREWS,AND G. F. BALL. 1985. Pa- Barrow, Alaska. Condor 57:3-18. rental carein an ecologicalperspective: A quan- PLUMS,W. J. 1965. Observations on the breeding titative analysisof avian subfamilies.Am. Zool. biology of the Razorbill. Br. Birds 58:449-456. 25:823-840. POOLE,A. 1985. Courtship feeding and Osprey re- SPENCER,O. R. 1943. Nesting habits of the Black- production.Auk 102:479-492. billed Cuckoo. Wilson Bull. 55:11-22. POOLEY,A. C. 1977. Nest opening responseof the STEWART,P.A. 1974. A nesting of BlackVultures. Nile crocodileCrocodylus niloticus. J. Zool. (Lond.) Auk 91:595-600. 182:17-26. STONEHOUSE,B.,AND S. STONEHOUSE.1963. The frig- PRATT,H. 1970. Breedingbiology of GreatBlue Her- ate bird Fregataaquila of AscensionIsland. Ibis ons and Common Egrets in central California. 103b:409-422. Condor 72:407-416. STRAHL,S.D. 1988. The socialorganization and be- PRUM,R.O. 1990. Phylogeneticanalysis of the evo- haviour of the Hoatzin Opisthocomushoazin in lution of display behavior in the NeDtropical central Venezuela. Ibis 130:483-502. manakins(Aves: Pipridae). Ethology 84:202-231. STRAUCH,J.G. 1978. The phylogenyof the Charadri- QUINN, J. S. 1990. Sexual size dimorphism and pa- iformes(Aves): A new estimateusing the method rental care patterns in a monomorphic and di- of charactercompatibility analysis.Trans. Zool. morphic larid. Auk 107:260-274. Soc. Lond. 34:263-345. RAIKOW,R.J. 1978. Appendicular royologyand re- SUTHERLAND,C. A. 1963. Notes on the behavior of lationshipsof the New World nine-primariedos- Common Nighthawks in Florida. Living Bird cines (Aves: Passeriformes).Bull. Carnegie Mus. 2:31-39. Nat. Hist. 7:1-43. SWOFFORD,D.L. 1990. PAUP: Phylogeneticanalysis R•IKOW,R.J. 1982. Monophyly of the Passeriformes: using parsimony, version 3.0s. Illinois Natural Test of a phylogenetichypothesis. Auk 99:431- 445. History Survey, Champaign. THORESON,A.C. 1964. The breedingbehavior of the RAIKOW,R.J. 1987. Hindlimb myology and evolu- Cassin Auklet. Condor 66:456-476. tion of the Old World suboscinepasserine birds (Acanthisittidae,Pittidae, Philepittidae, Eury- THORESON,A.C. 1969. Observationson the breeding laimidae). Ornithol. Monogr. No. 41. behaviourof the diving petrel Pelecanoidesu. uri- natrix (Gmelin). Notornis 16:241-260. REILLY,P. N., AND P. BALMFORD.1975. A breeding study of the Little Penguin Eudyptulaminor in VANRHIIN, J.G. 1984. Phylogeneticalconstraints in Australia.Pages 161-187 in The biology of pen- the evolutionof parentalcare strategies in birds. Neth. J. Zool. 34:103-122. guins(B. Stonehouse, Ed.). University Park Press, Baltimore. VAN RHIIN, J.G. 1985. A scenario for the evolution RIDLEY,M. 1978. Paternal care. Anim. Behav. 26:904- of socialorganization in RuffsPhilomachus pugnax 932. and other charadriiformspecies. Ardea 73.25-37. RIDLEY,M. 1983. The explanation of organic diver- VANRHIIN, J.G. 1990. Unidirectionality in the phy- sity.The comparativemethod and adaptations for 1ogenyof socialorganization, with specialref- mating. ClarendonPress, Oxford. erence to birds. Behaviour 115:153-174. RODGERS,J. A., JR. 1980. Little Blue Heron breeding VEHRENCAMP,S., B. S. BOWEN,AND R. R. KOFORD. 1986. behavior. Auk 97:371-384. Breedingroles and pairing patternswithin com- RYAN,M. R., ANDJ. J. DINSMORE. 1979. A quantitative munal groupsof Groove-billedAnis. Anim. Be- studyof the behaviorof breedingAmerican Coots. hav. 34:347-366. Auk 96:704-713. WALKINSHAW,L.H. 1965. One hundred thirty-three SAGAR,P.M. 1979. Breeding of the Cape Pigeon Michigan Sandhill Crane nests.Jack-Pine War- (Daptioncapense) at the SnaresIslands. Notornis bler 43:137-143. 26:23-26. WARHAM,J. 1975. The crestedpenguins. Pages 189- October1992] PhylogenyofAvian Parental Care 841

269 in The biology of penguins(B. Stonehouse, WittiNs, D. A., AND R. D. MORRIS. 1987. Parental Ed.). University Park Press,Baltimore. care of the Common Tern Sterna hirundo. Ibis 129: WARnAM, J., B. R. KEELS, AND G. J. W•LSON. 1977. 533-540. Breedingof the Mottled Petrel. Auk 94:1-17. W•LE•, R. H. 1974. Evolution of socialorganization WEIMERSKIRSCH,H., P. JOUVENTIN,AND J. C. STAHL. and life-history patternsamong grouse.Q. Rev. 1986. Comparativeecology of the six albatross Biol. 49:201-227. speciesbreeding on the Crozet Islands.Ibis 128: W•LLOVC•B¾,E. J., AND T. J. CADE. 1964. Breeding 195-213. behavior of the American Kestrel (Sparrow WELLER,M.W. 1958. Observations on the incubation Hawk). Living Bird 3:75-96. behavior of a Common Nighthawk. Auk 75:48- WILSON,U. W., ANDD. A. MANVWAL.1986. Breeding 59. biologyof the RhinocerosAuklet in Washington. WERSCHKUL,D.F. 1982a. Parental investment: Influ- Condor 88:143-155. enceof nest-guardingby male Little BlueHerons WINICLER,D. W. 1987. A generalmodel for parental Florida caerulea. Ibis 124:343-346. care. Am. Nat. 130:526-543. WERSC•ICUL,D. F. 1982b. Nesting ecologyof Little W•NICL•R,D. W., AND J. R. WALTERS.1983. The de- Blue Heron: Promiscuous behavior. Condor 84: terminationof clutchsize in precocialbirds. Curt. 381-384. Ornithol. 1:33-68. WESTMORELAND,D., L. B. BEST,AND D. E. BLOCKSTEIN. WITTENBERGER,J. F. 1978. The evolution of mating 1986. Multiple-broodingas a reproductivestrat- systemsin grouse.Condor 80:126-137. egy: Time conservingadaptations in Mourning WOLF,L. L., AND J. S. WOLF. 1971. Nesting of the Doves. Auk 103:196-203. Purple-throated Carib Hummingbird. Ibis 113: WIGGINS,D. A., AND R. D. MORRIS. 1986. Criteria for 306-315. female choice of mates: Courtship feeding and WooDs,R.S. 1960. Notes on the nestingRoadrun- paternal care in the Common Tern. Am. Nat. 128: ner. Condor 62:483-484. 126-129.

APPENDIX1. Charactersused in the analysis. APPENDIX 1. Continued.

1. M. iliotibialis mediaIls: (1) present; (0) absent. 2. M. obturatoriuslateralis, pars dorsalis:(0) present;(1) M. iliotibialis lateralis,pars preacetabularis:(0) pres- absent. 23. M. obturatorius mediaIls, two heads of ent; (1) absent;(2) vestigial.3. M. iliotibialislateralis, origin: (0) absent;(1) present.24. M. obturatoriusme- pars acetabularis:(0) present;(1) absent;(2) aponeu- diaIls, number of tendons of insertion: (0) one tendon; rotic; (A) 0 + 1.4. M. iliotibialis lateralis,pars postace- (1) two tendons;(2) three tendons.25. M. obturatorius tabularis: (0) present; (1) absent; (A) 0 + 1. 5. M. mediaIls,enlarged in width: (0) no; (1) yes. 26. Min. iliotrochantericus caudalis: (1) reduced; (0) unre- obturatorius medialis and obturatorius lateralis, distal duced.6. M. iliotrochantericuscranialis, strongly fused fusion: (0) yes; (1) no. 27. M. iliofemoralis internus: with M. iliotrochantericus caudalis: (1) fused; (0) un- (0) present;(1) absent.28. M. iliofemoralisinternus: fused. 7. M. iliofemoralis externus: (0) present; (1) (0) "typical";(1) "unusuallyshort and broad."29. M. absent. 8. M. femorotibialis externus, distal head: (0) arablens:(0) present;(1) absent.30. M. arablens,ex- present;(1) absent.9. M. femorotibialisinternus, lon- tent of origin: (0) limited to pectineal process;(1) gitudinal division:(1) present;(0) absent.10. M. ilio- extendingfrom pectinealprocess to pubis;(2) one fibularis, ansa iliofibularis forms a single ligament: origin from pectinealprocess and one from pubis.31. (1) present;(0) absent.11. M. iliofibularis, ansailiofib- M. arablens,longitudinal division: (0) absent;(1) pres- ularisarms elongated: (1) present;(0) absent;(2) mod- ent. 32. M. gastrocnemiuspars lateralis: (0) single;(1) erately elongate.12. M. flexor cruris lateralis, pars double;(A) 0 + 1.33. M. gastrocnemiuspars mediaIls, accessoria:(0) present; (1) absent; (A) 0 + 1. 13. M. pateliar band: (0) present;(1) absent;(A) 0 + 1. 34. flexor cruris lateralis,pars pelvica:(0) present;(1) ab- M. gastrocnemiuspars mediaIls, number of heads:(0) sent. 14. M. flexor cruris lateralis,pars accessoria:(1) one head; (1) two heads;(A) 0 + 1. 35. M. gastroc- reduced; (0) not reduced; (?) absent. 15. M. caudofe- nemius, fourth head: (0) absent; (1) present. 36. M. moralis:(0) present;(1) absent;(2) poorly developed. gastrocnemius,tendon of insertioncontributes to os- 16. M. iliofemoralis:(0) present;(1) absent;(2) poorly siftcation of the hypotarsus:(0) no; (1) yes. 37. M. developed;(A) 0 + 1. 17. M. flexor cruris mediaIls, tibialis cranialis, number of tendons of insertion: (0) two distinctparts: (1) present;(0) absent.18. M. flexor bifurcated tendon; (1) one tendon. 38. M. extensor cruris medialis and M. flexor cruris lateralis, tendons: digitorumlongus, number of headsof origin:(0) one (0) fused; (1) unfused. 19. M. pubo-ischio-femoralis, head (from tibia); (1) two heads(from tibia and fibula); division into pars cranialis and pars caudalis:(1) di- (2) two heads (from tibia and femur). 39. M. extensor vided; (0) not divided. 20. M. pubo-ischio-femoralis, digitorum longus, hallucal tendon: (0) absent;(1) muscular slip: (0) absent; (1) present. 21. M. pubo- present.40. M. fibularislongus: (0) present;(1) poorly ischio-femoralis, pars profundus divided into two developed;(2) absent;(B) 0 + 2. 41. M. fibularis lon- parts:(0) undivided; (1) divided; (2) intermediate. 22. gus,tibial head:(0) present;(1) fibularhead only; (2) 8•2 M_•¾ C. McKrrmcl< [Auk, Vol. 109

APPENDIX 1. Continued. APPENDIX 1. Continued.

arising from underlying musclesand from tibia. 42. brevis digiti IV: (0) present;(1) vestigial; (2) absent; M. fibularislongus, branch to FPD3: (0) present;(1) (C) 1 + 2. 66. M. lumbricalis: (0) absent; (1) present; absent.43. M. fibularis brevis: (0) present; (1) weak; (2) weak or vestigial. 67. Feathers:(1) present; (0) (2) absent. 44. M. flexor perforans et perforatus digiti absent (to reflect monophyly of birds and thereby III, vinculum: (0) present; (1) absent. 45. M. flexor excludeAncestor). 68. Neognath monophyly("dum- perforanset perforatusdigiti II, relationshipto M. my variable"):(1) reflectsmonophyly of the ingroup flexor perforans et perforatusdigiti III: (0) doesnot (for actual synapomorphies,see Cracraft 1986, Cra- overlapand concealFPPD3; (1) doesoverlap. 46. M. craft and Mindell 1989); (0) all other taxa. 69. Paleog- flexor perforanset perforatusdigiti II, number of nath monophyly ("dummy variable"): (1) reflects heads:(0) one head; (1) intermediate; (2) two; (3) three. monophylyof the sistergroup of neognaths(for ac- 47. M. flexor perforanset perforatusdigiti II, origin tual synapomorphiessee Cracraft 1986, Cracraft and from ansa iliofibularis: (0) absent;(1) present.48. M. Mindell 1989); (0) all other taxa. 70. Mating system: flexor perforatus digiti II, position: (0) deeply situ- (0) monogamy;(1) polygyny;(2) polyandry;(3) prom- ated; (1) superficial. 49. M. plantaris: (0) present; (1) iscuity; (A) 1 + 2; (B) 0 + 1; (C) 0 + 3. 71. Duration absent; (A) 0 + 1. 50. M. plantaris: (0) "typical"; (1) of pair bond: (0) seasonal;(1) brief; (2) longer than very powerfully developed.51. M. flexor hallucislon- "brief" but less than one nesting effort; (3) longer gus,branch to hallux: (0) present;(1) lacking or weak. than one season(ranging from acrosstwo seasonsto 52. M. flexor hallucislongus and M. flexor digitorum life); (4) pair bond may be renewed annually but dis- longus, type of flexor arrangement:see George and solvesbetween seasons. 72. Mate guardingby males: Berger(1966:447) for descriptionof TypesI-VIII, and (0) absent;(1) present.73. Resourcedefense polygyny Betman (1984) for description of Type X (coded 9 (harem):(0) absent;(1) present;(A) 0 + 1.74. Lekking: herein); modificationfound in hummingbirdsis des- (0) absent;(1) present.75. Defenseof young/parental ignated Type 0. 53. M. flexor hallucislongus, number care: (0) biparental; (I) uniparental (female); (2) un- of heads: (0) one head; (I) two heads; (2) three heads. iparental (male); (3) double clutches;(4) no parental 54. M. flexor digitorum longus, number of heads:(0) care; (A) 0 + 1. 76. Distinguishable reverse sexual two heads; (1) three heads. 55. M. flexor digitorum dimorphism of any kind: (0) absent; (1) present. 77. longus, size: (0) "typical"; (1) very powerful; (2) in- Males feed females during nest care: (0) no; (1) yes. termediate.56. M. flexor digitorum longus, location: 78. Incubation:(0) by female; (1) by both parents;(2) (0) deeply situated;(1) superficially situated. 57. M. by male only; (3) by neither parent. 79. Nest-building popliteus:(0) present;(1) absent.58. M. flexor hallucis or nest-siteselection: (0) by female only; (1) by both brevis: (0) present;(I) vestigial;(2) absent;(B) 0 + 2. parents;(2) by male only. 80. Posthatching(prefledg- 59. M. flexor hallucis brevis, number of tendons of ing) care (feeding, as distinct from defense,character insertion: (0) one; (1) two. 60. M. extensor hallucis 75): (0) female only; (1) both parents;(2) male only; longus:(0) present;(1) absent.61. M. extensorhallucis (3) neither parent. 81. Postfledgingcare: (0) female longus,number of heads:(0) two heads;(1) one head; only; (1) both parents;(2) male only; (3) neither. 82. (A) 0 + 1.62. M. extensorhallucis longus, accessory: Group breeding: (0) absent;(1) present.83. Colonial (0) absent; (1) present. 63. M. abductor digiti II: (0) nesting: (0) absent;(1) present. 84. Posthatchingde- present;(1) absent;(2) vestigial.64. M. adductordigiti velopment: (0) precocial; (I) altricial. II: (0) present; (1) weak; (2) absent. 65. M. extensor October1992] PhylogenyofAvian Parental Care 843

•oo•oo•oo•oo•oo•oo•oo•oo•••

••oo•oo•oo•oo•oo•oo•oo•oo•

•.• ...... •oo•oo•ooõoo•

o• • 84:4: MARYC. McKITRICK [Auk,Vol. 109

.•.•.• •

oo•øø •OOO g•S•ggggggo• •.o

OOO oo •OOoog OO g OO ggggg•g gggggggggg•ggggøøoo googoogoogggggggoo oo oo

OO ggggg•gg• •oo gøøgøøgøøgggg•oo oo oo

ggg•g•ggggg•"" oo oo oo oo g•ø•ggggggggggøøOO OO OO googooOO OO•OOoogoogggg OO OO •gøøg••g•g•o•oo OO oooooooooooooooo OOO •oooooooooooo•o OOO •OOOOOOOOOOOO• oooooooooooooooo OOO OO••OOOOOOO oooooooooooooooo OOO OOOOOOOOOOOOOOOO •ooooooooooooo OOO OOOOOOOOOOOOOOOO oooooooooooooooo OOO OOOOOOOOOOOOO•O

•oooooooooooo OOO OOOOOOOOOOOOO•O

•O••OOOOOOO •ooooo•oooooooo•ooo•oooooooooooo OOO OOO OOOOO•OOOOOOO•O •ooooooooooooooooooooooooooooo OOO OOOOOOOOOOOOOOOO OOO OOOOOOOOOO•OOO

oog •g•ggg•øø •oo•oo•oogoogoog• oo oo g•o• oo oogggggg•o• oo •o• OO • OO g • o•.•øøg••øøOO OO googoogoogoogoo oo oo oo .... g ggo•googgggggggooOO OO OO OO googoogoogoog•goo oo oo oo oo ggøøgøøggggggggøø .... o•.g oo g oo g g ggøøgøøggggggggoooo oo oo googoogoogoogoogoo oo oo oo oo g•oo•o•• o•ggggg•goo oo OO •o• •o• oo oo oo ggøøgøø•gggggggø•' OO •ø•øø•øøg•øø• OO OO ggoog•OO oogggggggg-•' O• • •øøgøøgoo oo .... • oo ggoogoogggggggg•ooo oo oo OO oo• googoogoo•oo oo • •o•ooooooo OOO oooooooooooooooo •oooooooooooooooooooooooooooooo•o OOO oooooooooo•ooo OOO •ooooooo••oo•ooooooooooooooooo OOO oooooooooooooooo •oooooooooooooooo OOO oooooooooooooooo OO OO gøøgøøgøøoo oo oogoo øøgg• g ggøø•ggggggggggøøOO OO OO gøøgøøgøøg•g•OO OO OO OO •ggggggg•ggø•øøgggggggggggøø oo •o•goo•g•oo oo oo o•g•o• oo oog OO g OO gøøg•'gOO • oo OO oo• oo •oo oog•og oo OO oo oo oog oo goo • g•øøggggggggggggg•.•...... gggggggo• OO oooooooooo•ooo OO OO oo oo oo• oo • oo • g•øø •oggggggggøø O• •øø•øø•øøgøøgøøg googooggggggggooOO OO OO OO googoo OO OOoo OO goo•OO OOOOOOOOOOOOOOOOøøggggggggøø OOO OOOOOOOOOOOOOOOO October1992] PhylogenyofAvian Parental Care 845

AI•I•ENOIX3. Character-changelists for behavioral characters(70-84) basedon Tree 1 (Fig. 2). The "ci" refersto consistencyindex. Eachchange is one step in length. Arrows with double lines indicate tran- sitionsthat are the samefor all optimizations(ALL). Arrows with single lines indicate transitions that are not the samefor all optimizations(e.g. ACC- TRAN). The "wt" refers to "within-terminal taxon." See Methods for explanation of optimization routines.

Character 70 (ci = 0.667). ALL: node 102 0 • 3 Eulampis(a change from state 0 to state 3 between node 102 and Eularnpis);node 115 0 • 3 Dendragapus. ACCTRAN: node 118 0 • 1 Ancestor; node 62 0 • 1 node 61; Rhea1 • 12 = A (wt); Crypturellus1 • 12 = A (wt); node 114 0 • 1 Anas;Oxyura 0 • 03 = C (wt); Opisthocomus0 • 01 = B (wt). DELTRAN: Rhea1 • 12 = A (wt); Crypturellus1 • 12 = A (wt); node 62 1 • 0 Nothoprocta;node 117 1 • 0 node 113; node 114 1 • 00xyura; Oxyura 0 • 03 = C (wt); Opisthocomus1 • 01 = B (wt). MINF: Rhea1 • 12 = A (wt); Crypturellus 1 • 12 = A (wt); node 62 1 • 0 Nothoprocta;node 118 1 • 0 node 117; node 114 0 • 1 Anas;Oxyura 0 • 03 = C (wt); Opisthocomus0 • 01 = B (wt). Character 71 (ci = 0.222). ALL: node 61 2 • 0 Rhea;

0•0000 node 64 4 • 3 Pterodroma;node 68 4 • 0 Podiceps; node 69 4 • 3 Chen;node 74 4 • 3 Cepphus;node 84

0•.•0•. 0 • 3 Grus;node 115 2 • 1 Dendragapus.ACCTRAN: 0•0000 node 117 2 • 0 node 113; node 83 0 • 4 node 78; node 88 0 • 4 node 87; node 90 0 • 4 Ciconia; node O0 103 0 • 1 node 102; node 102 1 • 3 node 101; node 106 0 • 3 node 104; node 111 0 • 3 node 110; node 109 3 • 4 Crotophaga;node 112 0 • 1 Zenaida;Zenaida 000000 1 • 12 = A (wt). DELTRAN: node 113 2 • 0 node 93; 000000 node 83 0 • 4 node 78; node 88 0 • 4 node 87; node 000000 90 0 • 4 Ciconia; node 111 2 • 0 node 108; node 102 •.000• 0 • 3 node 101; node 102 0 • 1 Eularnpis;node 104 0 • 3 Picoides;node 109 2 • 3 Geococcyx;node 109 2 • 000•00 4 Crotophaga;Zenaida 2 • 12 = A (wt). MINF: node •0000 0000• 117 2 • 0 node 113; node 83 0 • 4 node 78; node 88 000000 000000 0 • 4 node 87; node 90 0 • 4 Ciconia; node 102 0 • 3 node 101; node 102 0 • 1 Eularnpis;node 104 0 • 3 Picoides;node 109 0 • 3 Geococcyx;node 109 0 • 4 Crotophaga;node 112 0 • 1 Zenaida;Zenaida 1 • 12 =

oo A (wt). oo oo Character 72 (ci = 0.143). ALL: node 102 1 • 0 Eularnpis;node 106 1 • 0 node 104; node 114 0 • 1 Anas. ACCTRAN: node 62 0 • 1 node 61; node 85 0 • 1 node 84; node 80 1 • 0 Rissa; node 113 0 • 1 .ooEE node 112. DELTRAN and MINF: node 61 0 • 1 Rhea; node83 0 • 1 node 78; node 79 0 • 1 Rynchops;node 112 0 • 1 node 111. gooooo 000000 Character 73 (ci = 1.000). ALL: node 62 0 • 1 node 00•00 61.

0000• • Character 74 (uninformative). Character 75 (ci = 0.600). ALL: node 118 0 • 2 node 62; node 102 0 • 1 Eularnpis.ACCTRAN: node 117 0 0•.0000 • • 1 node 116; node 114 1 • 00xyura; Opisthocomus •0•00 000000 • 1 • 12 = A (wt). DELTRAN and MINF: node 114 0 •

000000 1 Anas;node 116 0 • 1 node 115; Opisthocomus1 • •000•.• 12 = A (wt).

000000 '• Character 76 (ci = 0.500). ALL: node 82 0 • 1 Ster- •00•0 • corarius 1; node 101 0 • 1 node 99. 0•00 • 000000 Character 77 (ci = 0.250). ALL: node 81 0 • 1 node 000000 846 MARYC. MCKITRICK [Auk,Vol. 109

APPENDIX 3. Continued. APPENDIX 4. Continued.

80. ACCTRAN: node 93 0 • 1 node 92; node 108 0 • (Thoreson 1969). 16. Pygoscelisadeliae (Ainley and 1 node 107; node 102 1 • 0 Eulampis.DELTRAN and Schlater1972, Derkson 1977,Ainley et al. 1983).17. MINF: node 89 0 • 1 Leptoptilus;node 102 0 • 1 node Eudyptes,all spp. (Warham 1975). 18. Eudyptulaminor 101; node 105 0 • 1 Corvus. (Reilly and Balmford1975). 19. Spheniscusmagellanicus Character 78 (ci = 0.375). ALL: node 69 1 ==•0 Chen; (Boswalland MacIver 1975).20. Gaviaimmer (Mcintyre node 97 1 ==•0 Accipiter;node 105 1 ==•0 Tyrannus. 1988).21. Podicepsnigricollis (R. W. Storer,pers. comm.). ACCTRAN: node 118 0 • 3 Ancestor; node 118 0 • 22. Fregataaquila (Stonehouse and Stonehouse1963). 2 node 62; node 117 0 • 1 node 113; node 103 1 • 0 23. Phalacrocoraxauritus (Palmer 1962).24. Anhingaan- node 102; node 101 0 • 1 node 99. DELTRAN: node hinga(Burger et al. 1978). 25. Sula bassanus(Nelson 118 3 • 2 node 62; node 117 3 • 1 node 113; node 1966). 26. Ardea herodias(Pratt 1970, Mock 1979). 27. 100 1 • 00tus; node 102 1 • 0 Eulampis;node 117 3 Butoridesstriatus (Palmer 1962). 28. Florida caerulea • 0 node 116. MINF: node 118 1 • 3 Ancestor; node (Rodgers1980, Werschku11982a, b). 29. Mycteriaamer- 118 1 • 2 node 62; node 100 1 • 00tus; node 102 1 icana(Palmer 1962, Kah11972). 30. Ciconiaciconia (Hav- • 0 Eulampis;node 117 1 • 0 node 116. erschmidt 1949, Noble-Rollin 1975). 31. Leptoptilus Character 79 (ci = 0.286). ALL: node 117 0 • 1 node crumeniferus(Kahl 1966).32. Plegadisautumnolis (Bay- 113;node 89 1 • 2 Mycteria;node 96 1 ==•0 Chordeiles; nard 1913, Palmer 1962). 33. Phoeniconaiusminor (Brown node 102 1 • 0 Eulampis.ACCTRAN: node 118 0 • 2 and Root 1971). 34. Cathartesaura (Coles 1944, Palmer node 62; node 71 1 • 2 node 66; node 106 1 • 0 node 1988a).35. Coragypsatratus (Stewart 1974). 36. Gym- 105. DELTRAN: node 62 0 • 2 node 61; node 65 1 • nogypscalifornicus (Palmer 1988a).37. Accipitergentilis 20ceanodroma;node 105 1 • 0 Tyrannus.MINF: node (Moller 1987,Palmer 1988a). 38. Buteojamaicensis (Fitch 62 0 • 2 node 61; node 66 1 • 2 node 65; node 105 et al. 1946, Palmer 1988b). 39. Pandionhaliaetus (Green 1 • 0 Tyrannus. 1976,Poole 1985,Birkhead and Lessells1988). 40. Falco Character 80 (ci = 0.750). ALL: node 102 1 • 0 sparverius(Willoughby and Cade 1964,Palmer 1988b). Eulampis.ACCTRAN: node 118 0 • 3 Ancestor;node 41. Chencaerulescens (Finney and Cooke 1978, Cooke 118 0 • 2 node 62; node 117 0 • 1 node 113. DEL- et al. 1981, Lank et al. 1989). 42. Anasplatyrhynchos TRAN: node 118 3 • 2 node 62; node 117 3 • 1 node (Lebret 1961, Goodburn 1984, McKinney 1986). 43. 113; node 117 3 • 0 node 116. MINF: node 118 1 • Oxyurajamaicensis (Siegfried 1976, Joyner 1977). 44. 3 Ancestor; node 118 1 • 2 node 62; node 117 1 • 0 Dendragapusobscurus (Wiley 1974,Wittenberger 1978, node 116. Lewis 1985).45. Lagopusmutus (McDonald 1970).46. Character 81 (ci = 0.333). ALL: node 118 1 • 2 node Canachitescanadensis (Ellison 1973). 47. Bonasaumbellus 62; node 83 1 ==• 3 node 78; node 70 3 ==• 1 node 69; (Gladfelterand McBurney 1971,Wittenberger 1978). node 74 3 • 2 node 73; node 94 1 ==• 3 Florida; node 48.Meleagris gallopavo (Mosby and Handley1942, Dalke 96 1 • 2 Chordeiles;node 103 1 • 3 Chaetura; node et al. 1946). 49. Opisthocomushoazin (Strahl 1988). 50. 112 1 ==• 2 Zenaida; node 117 1 ==• 0 node 116. Grus canadensis(Walkinshaw 1965, Nesbitt 1989). 51. Character 82 (ci = 0.250). ALL: node 103 0 • 1 Fulicaamericana (Ryan and Dinsmore 1979).52. Ryn- Chaetura; node 105 0 ==• 1 Corvus; node 109 0 ==• 1 chopsniger (Burger 1981,Quinn 1990).53. Stercorarius Crotophaga;node 115 ==•Opisthocomus. pomarinus(Pitelka et al. 1955,Maher 1974).54. Ster- Character 83 (ci = 0.200). ALL: node 85 0 ==•1 node corariusparasiticus (Perdeck 1963, O'Donald 1983).55. 84; node 69 1 • 0 node 68; node 93 0 • I node 92; Larusmarinus (Butler and Janes-Butler1983). 56. Rissa node 95 0 • 1 node 94; node 109 0 • 1 Crotophaga. tridactyla(Hodges 1969, Maunder and Threlfall 1972). Character 84 (ci = 0.250). ALL: node 117 0 ==•1 node 57. Sternahirundo (Wiggins and Morris 1986, 1987). 113; node 88 1 • 0 node 85; node 91 1 • 0 Phoeni- 58. Alca torda(Plumb 1965, Harris and Birkhead 1985). conaius;node 96 1 • 0 Chordeiles. 59. Uria aalge(Johnson 1941, Birkheadet al. 1985, Hatchwell 1988).60. Cepphuscolumba (Drent 1965).61. Brachyramphusmarmoratum (Sealy 1974). 62. Ptycho- ramphusaleuticus (Thoreson 1964). 63. Cerorhincamono- APPENDIX 4. References used for each taxon. cerata(Wilson and Manuwal 1986).64. Fraterculaarctica (Ashcroft 1979). 65. Zenaidamacroura (Westmoreland 1. Ancestor,Crocodylus (Hunt 1975, Pooley 1977). 2. et al. 1986).66. Coccyzus erythropthalmus (Spencer, 1943). Rheaamericana (Bruning 1974).3. Crypturellusboucardi 67. Geococcyxcalifornianus (Woods 1960, Ohmart 1973, (Lancaster1964). 4. Nothoproctaornata (Pearson and Folse and Arnold 1978). 68. Crotophagasulcirostris Pearson 1955). 5. Diomedeaimmutabilis (Fisher 1971, (Vehrencampet al. 1986).69. Otusasio (Sherman 1911, Lefebvre 1977).6. Phoebetriafusca (Weimerskirsch et Allen 1924).70. Bubovirginianus (Errington 1932,Keith al. 1986).7. Daptioncapense (Sagar 1979). 8. Fulmarus 1977). 71. Chordeilesminor (Bent 1940, Weller 1958, glacialis(Hatch 1987,1990). 9. Procellariaparkinsoni (Im- Dexter 1961, Sutherland 1963). 72. Chaeturapelagica ber 1987). 10. Puffinuspuffinus (Palmer 1962, Harris (Dexter1952, Fischer 1958). 73. Eulampisjugularis (Wolf 1966, Lack 1968). 11. Pterodromainexpectata (Warham and Wolf 1971).74. Colapresauratus (Burns 1900).75. et al. 1977). 12. Oceanitesoceanites (Palmer 1962). 13. Picoidesvillosus (Kilham 1966, 1968, 1969).76. Tyrannus Fregettatropica (Beck and Brown 1971). 14. Oceanod- tyrannus(McKitrick 1990, pers. observ.).77. Corvus romaleucorrhoa (Palmer 1962). 15. Pelecanoidesurinatrix brachyrhynchos(Good 1952).