sign in sign up
<<
Home , Apodiformes, Passerine

Letters •

Why so many kinds of ?

Raikow and Bledsoe (2000), in em- Slud 1976). It is unreasonable to assume the list of possible reasons for passerine bracing the null model of Slowinski that there is no underlying biological rea- success, but I would place more empha- and Guyer (1989a, 1989b), may perhaps son for this pattern and for the major sis on it than he did. be said literally to have added nothing turnover in avifaunas in the Northern It is difficult to discuss the nest-bufld- to the Kst of suggestions for why there are Hemisphere in favor of after ing capabilities of passerines without re- so many species of passerine birds. When the Oligocène. sorting to anthropomorphisms such as Raikow (1986) addressed this problem Reproductive adaptations presumably "clever" or "ingenious." Is it not mar- previously and could find no key mor- made the holometabolous insects velous, however, that a highly special- phological adaptations to explain the di- (Coleóptera, Díptera, Hymenoptera, ized aerial feeder such as a cliff versity of Passeriformes, the so-called Lepidoptera, and so on) the dominant ( pyrrhonota), with tiny, weak or perching birds, he despaired of organisms on earth. Likewise, it bill and feet, can fashion a complex nest and suggested that the problem may be appears that reproductive adaptations, out of gobs of mud fastened to a flat, only "an accident of classificatory his- not morphology, are responsible for the vertical surface? What adjective suffices tory," which brought on a storm of protest dominance of passerine birds over other to describe the nest of {Systematic Zoology 37: 68-76; 41: 242- orders of birds. (Orthotomus), which actually stitch the 247). Now, Raikow and Bledsoe have sub- Most arboreal nonpasserine birds are edges of two leaves together with plant stituted one form of nihilism for another. obligate cavity nesters that rely on holes fibers to provide a concealed cup in The nuH model formulation deals merely in trees, earthen banks, or termite nests. which to place their nest? The long wo- with numerical probability and lacks any This automatically restricts them to habi- ven bags of oropendulas and relatives explanatory value in this case, unless per- tats where those features are present and (Icteridae) and the various complex wo- haps it could be shown that no evolu- at the same time makes competition for ven structures made by weaverbirds tionary forces influenced patterns of nest sites severe. Cavities cannot be read- () are well-known examples of speciation in birds. ily concealed and numerous predators the complexity of passerine nest types. What is known about the dynamics exploit the vulnerability of cavity-dweUing The delicate cup nests of and timing of the radiation of the organisms (snakes, certain hawks, and so (Poliopttla), made of spider webs cov- Passeriformes does not support the suc- on). On the other hand, passerine birds, ered with lichens and fastened to a hor- cess of these birds being due to random both oscines and suboscines, have an ex- izontal limb, or the cup nest of the processes. Similar early and middle traordinary abñity to fashion nests out of Australian sitellas (Sitella), placed in an avifaunas from the United States a wide variety of materials, to place them acutely angled crotch and covered with (Wyoming), England, and Germany show in an equally wide variety of situations, minute, verticafly-oriented strips of bark, that by about 50 million ago, Europe and, when need be, to camouflage or con- are exquisite uses of camouflage among and North America were inhabited by a ceal them with unusual "resourcefijlness." passerines. diverse array of arboreal birds, some of That the ability of passerines to protect Baptista and Trafl (1992) emphasized very small size, but not one of which was their and young in this manner may the abflity of passerines to move into new a passerine (Olson 1989, Mayr 1998a, have led to their success as a group is not environments as a key factor in the suc- 1998b, 2000). Although passerines oc- a new idea, having been advanced clearly cess of the order. If there is any single at- cured in the early Eocene of and in some detail by Irwin (1962) in an tribute that would make this possible, it (Boles 1997), the earliest evidence of the otherwise obscure publication that has is the ability of passerines to adapt their order in the Northern Hemisphere is since been widely overlooked. CoUias breeding regimen to locally avaflable nest from the latest Oligocène (about 25 mil- (1997) added nest-buflding behavior to sites and nesting materials. The brain in lion years ago) of France (Mourer- the entire order appears to be "hard- Letters to the Editor Chauviré et al. 1996). Passerines do not wired" for nest-building inventiveness. BioScience become an important component of the Attn: Science Editor The next most successflfl order of birds in avifaunas of the Northern Hemisphere 1444 Eye St., NW Suite 200 terms of number of species is the until the (about 20 to 5 million Washington, DC 20005 (swifts and ), years ago), when they radiated explo- The Stan of BioScietice reserves the right to edit species which also have complex and di- sively. Passerines are now the numeri- letters for length or clarity without notifying verse nest-building habits. This is un- cally dominant group of birds in virtually the author. Letters are published as space likely to be coincidental. becomes available. all geographical areas of the globe (see

268 BioScience • April 2001 I Vol. 51 No. 4 » Letters

As a morphologist, I am quite willing . 2000. Tiny hoopoe-like birds from the To understand the value of null to concede that morphology is not going Middle Eocene of Messel (Germany). Auk model approaches, consider the ques- to provide the answer to the numerical 117:964-970. tion "How different do sister taxa need dominance of the Passeriformes. Mor- Mourer-Chaviré C, Hugueney M, Jonet P. 1996. to be in species richness to cause a phology WÜ1 stiU be useful, in tandem Paleogene avian localities of France. Pages search for a deterministic explanation with molecular studies, in providing 567-598 in Mlikovsky J. Tertiary avian locali- for the difference?" Would sister groups much-needed improvements in our un- ties of Europe. Acta Universitatis Caroliniae of, say, 100 and 110 species, respectively, derstanding of relationships within the Geológica 39: 519-846. differ sufficiently to warrant a search for Passeriformes, which, in contrast to the Olson SL. 1989. Aspects of global avifaunal an explanation like the one Olson pro- impression given by Raikow and Bledsoe, dynamics during the Cenozoic. Acta XIX poses for passerine diversity? Exacfly how is still very imperfect. At the same time, Congressus Internationalis Ornithologici 2: different must groups be in species rich- a detailed analysis of nest types could 2023-2029. ness before one would be willing to in- also contribute materially to a better phy- Raikow RI. 1986. Why are there so many kinds of voke a deterministic, as opposed to logenetic fi-amework and lead to fasci- passerine birds? Systematic Zoology 35: stochastic, cause for the difference? nating insights into divergence and 255-259. As noted in our article, and contrary to convergence in nest-building behavior. Raikow RI, Bledsoe AH. 2000. Phylogeny and evo- Olson's implication at the end of his first Another fruitful avenue of investigation lution of the passerine birds. BioScience 50: paragraph, we are not arguing here that would be descriptive and experimental 487^99. the number of species in a given clade has studies of the neuroanatomy and neuro- Slowinski JB, Guyer C. 1989a. Testing null models no antecedent causes. It of course must. physiology of nest-building behavior, in questions of evolutionary success. System- Instead, we merely ask by how much sis- along the lines of the studies of vocaliza- atic Zoology 38: 189-191. ter groups must differ in species rich- tions that have generated so much at- . 1989b. Testing the stochasticity of pat- ness to prompt a search for an tention in in recent years. It terns of organismal diversity: An improved explanation for the difference in the adap- is time for the analysis of passerine suc- null model. American Naturalist 134: tations possessed by the groups. We reit- cess to progress beyond disheartened sor- 907-921. erate that the null model approach is ties into semantics and probability Slud P. 1976. Geographic and climatic relation- designed precisely to answer the ques- statistics and enter the realm of enthusi- ships of avifaunas with special reference to tion of whether such a search is required. astic biological exploration. comparative distribution in the Neotropics. Gotelli and Graves (1996) identify five STORRS L. OLSON Smithsonian Contributions to Zoology 212: important features of null models. With Department of Zoology 1-149. respect to Olson's statement that the nuH National Museum of Natural History model we used "lacks any explanatory Smithsonian Institution Response from Raikow and value," we believe its most important fea- Washington, DC 20560 Bledsoe tures are (1) that it allows for the possi- Olson makes two main points in his bility that no deterministic mechanism is References cited letter. He criticizes our appKcation of operating to produce an observed result; Baptista LF, Trail PW. 1992. The role of song in a null model to the study of the diversity (2) that stochastic processes may be re- the of passerine diversity. Systemat- of passerine birds, and he offers the hy- sponsible for a result; and (3) that parsi- ic Biology 41: 242-247. pothesis that the great diversity of - mony dictates that we favor simple Boles WE. 1997. Fossü songbirds (Passeriformes) ine birds results from their nest-bunding explanations over complex explanations. from the early Eocene of Australia. Emu 97: capabilities. In presenting the latter as an If a properly constructed null model pre- 43-50. alternative to a null model explanation of dicts a pattern seen in nature, then we are Collias NE. 1997. On the origins and evolution of passerine diversity, Olson conflates two obligated under the principle of parsi- nest building by passerine birds. Condor 99: different questions. The first asks whether mony to favor the simple explanation 253-270. the passerines are significantly different that stochastic processes have produced Irwin MPS. 1962. The adaptive significance of from their sister group in number of the pattern. nest construction in the evolution and radia- species. The second asks whether such a We believe that, in a real sense, there is tion of the Passeres. Proceedings of the First difference may be attributed to a partic- explanatory power in such a result. We Federal Science Congress (Harare, Rhodesia): ular cause•in this instance, according to would argue that a random branching 217-219. Olson, the possession of a particular re- process is sufficient to explain the dif- Mayr G. 1998a. "Coraciiforme" und "piciforme" productive adaptation. We argue, as have ference in species richness between Klernvögel aus dem Mittel-Eozän der Grube others (e.g.. Bond and Opell 1998, Chan passerines and their sister group. This Messel (Hessen, Deutschland). Courier and Moore 1999), that before asking this stochastic explanation is indeed an ex- Forschungsinstitut Senckenberg 205: 1-101. second question, one should first demon- planation, contrary to what Olson im- . 1998b. A new family of Eocene zygo- strate that a group is indeed unusually plies. It is also important to note that dactyl birds. Senkenbergiana Lethaea 78: species rich. Null model approaches are nuH models have basic biological princi- 199-209. designed specificaUy for this purpose. ples at their core; they are far from being

April 2001 I Vol. 51 No. 4 • BioScience 269 Letters

a form of "nihilism." In the model we References cited pesticide use, taking into accoimt the dif- used, these principles are imbedded in the Bond JE, Opell BD. 1998. Testing adaptive radia- ferent value of birds to different observers. model's assumptions•among others, tion and key innovation hypotheses in spi- Our review (Pimentel et al. 1993) sug- that speciation events are essentially in- ders. Evolution 52: 403^14. gested that individual birds were valued stantaneous, that they are independent, Chan KMA, Moore BR. 1999. Accounting for at 40$ per by bird watchers, $216 per even that speciation occurs at all. Any of mode of speciation increases power and real- bird by hunters, and $800 per bird based the model's assumptions are open for ism of tests of phylogenetic asymmetry. on the costs of replacement. We esti- inspection of how violations of them American Naturahst 153: 332-346. mated the value of an adult bird, then, to would affect our interpretation of species Gotelli NJ, Graves OR. 1996. Null models in ecol- be approximately $30. Some reviewers richness in sister . But null models ogy Washington (DC): Smithsonian Institu- thought this amount too low, and others should not be dismissed out of hand. tion Press. too high. Data in the literature suggest Null model assessment is increasingly Pearson PN. 1998. Speciation and extinction that about 67 million birds are killed in recognized as a critical first step in analy- asymmetries in paleontological phylogenies: the field with pesticides, and this number ses of species richness. Bond and Opell Evidence for evolutionary progress. Paleobi- does not even include the young birds (1998; araneoid spiders), Pearson (1998; ology 24: 305-335. that die in the nest because their parents planktonic foraminifera, nannofossils, Purvis A, Nee S., Harvey PH. 1995. Macroevolu- are killed by pesticides or the yoimg birds and graptoloids), Purvis et al. (1995; pri- tionary inferences from primate phylogeny that were killed when they were fed mates), and Wollenberg et al. (1996; Proceedings of the Royal Society of London pesticide-contaminated insects. columbine plants, cranes, and Drosophila) Series B Biological Sciences 260: 329-333. Gatto and de Leo say that a contingent all have used a null model as a starting Slowinski JB, Guyer C. 1989. Testing null models valuation method•whereby respon- in questions of evolutionary success. System- dents to a questionnaire state how much point to investigate patterns of species atic Zoology 38: 189-191. they would be willing to pay for some en- richness. Sometimes null models are suf- WoUenberg K, Arnold J, Avise JC. 1996. Recogniz- vironmental resource•is the only pric- ficient to explain the diversity of sister ing the forest for the tree: Testing temporal ing technique "capable of providing an groups (e.g., Drosophila virilis species patterns of cladogenesis using a null model of estimate of existent values" (Gatto and de group; WoUenberg et al. 1996). In other stochastic diversification. Leo 2000, p. 348). In my view, however, instances, they are not (e.g., cercopithe- and Evolution 13: 833-849. it would be a mistake to use willingness cid primates; Purvis et al. 1995). R. J. RAIKOW to pay to assess the value of birds killed Regardless, all of these authors have A. H. BLEDSOE by pesticides. The general public simply started with a null model analysis, rec- Department of Biological Sciences does not know how many birds are killed ognizing, as did Slowinski and Guyer in University of Pittsburgh by pesticides, so any valuation based on their seminal paper (1989, p. 190), that "a Pittsburgh, PA 15260 the public's wiUingness to pay must be random branching pattern inherently suspect. produces sister taxa of disparate size" (their Pricing and Our papers (Pimentel et al. 1992, emphasis). Ecosystem Services 1993) give details and sources and de- We would be more enthusiastic had • agree with Marino Gatto and Giulio scribe how we arrived at our estimates Olson mounted serious potential objec- de Leo: It would be desirable to base es- of the value of species and ecosystems. tions to our analysis: that the assumptions timates of the benefits of biodiversity Scientists, decisionmakers, and the pub- of the null model might be incorrect, and ecosystem services (environmental lic are very much interested in these en- that violations of these assumptions assessment) on the knowledge of an in- vironmental and economic data, because might seriously undermine our result, terdisciplinary team that could "investi- this type of accounting gives them some that the Sibley-Ahlquist phylogeny might gate aU possible environmental, social, idea of the magnitude of a particular be flawed and hence that we are making and economic consequences of a pro- ecological and economic problem. an inappropriate comparison, to name a posed activity" (Gatto and de Leo 2000, Certainly, we all hope that better data few. But Olson mounts no such serious p. 353). However, this ideal-sounding ap- will be forthcoming, supplied through objections. Furthermore, we would be proach invites some practical questions. the type of interdisciplinary investiga- delighted to see a rigorous, compelling For example, how large and diverse tions suggested by the authors. But demonstration that the possession of an should such an interdisciplinary team until those investigations are achieved, adaptation like nest-building capabili- be? Also, is it possible for any team, let us use current data to inform scien- ties causally explains passerine diversity. however large and diverse, to investigate tists, decisionmakers, and the public of But that demonstration has not occurred. all the possible environmental, social, the magnitude of ecosystem problems. Far from consisting of "disheartened and economic consequences of an activ- DAVID PIMENTEL sorties into semantics and probability ity? College of Agriculture and statistics," our analysis, we hope, has Eight years ago, my colleagues and I as- Life Sciences introduced an important consideration sessed the environmental and economic Cornell University into the debate about passerine diversity. costs incurred for bird losses caused by Ithaca, NY 14853

270 BioScience • April 2001 I Vol. 51 No. 4 » Letters

References cited interdisciplinary team and the types of mortality. Because the econo- Gatto M, de Leo GA. 2000. Pricing biodiversity expertise it should encompass are cali- mists calculate that a good deal of the and ecosystem services: The never-ending brated on the basis of these findings. grant money must go toward prepar- story. BioScience 50: 347-355. According to Canter (1996, p. 50), "The ing, printing, and mailing the question- Pimentel D, Stachow U, Takacs DA, Brubaker number of members of an interdisci- naires•and toward paying HW, Dumas AR, Meaney JJ, O'Neil JAS, Onsi plinary team can vary from as few as 2 themselves•they have to behave in the DE, Corzilius DB. 1992. Conserving biologi- to perhaps as many as 8 or 10 individ- following way: They do not consult an cal diversity in agricultural/forestry systems. uals, depending on the size and com- agronomist, but make their own rough BioScience 42: 354-362. plexity of the study. Typically a team estimate of increased harvests based on Pimentel D, Acquay H, Biltonen M, Rice P, Silva comprises three to four members." previous similar case studies. They give M, Nelson J, Lipner V, Giordano S, Horowitz In any case, the question is not posed a little money to an ecotoxicologist to A, D'Amore M. 1993. Assessment of environ- properly in the context of our article. conduct a study and estimate the pesti- mental and economic costs of pesticide use. The right question would be "Given a cide effects on just one charismatic Pages 47-84 in Pimentel D, Lehman H, eds. limited budget to evaluate a proposed species of bird. Thus the questionnaire The Pesticide Question: Environment, Eco- activity, would it be better to spend the recipients learn nothing of the pesticide nomics and Ethics. New York: Chapman and money on an interdiscipHnary study that effects on mammals and other birds, Hall. explicitly incorporates multiple evalua- and the returns in terms of increased tion criteria or on a cost-benefit analysis crop production are not known pre- Response from Gatto and conducted by economic consultants who cisely. The economists collect the ques- De Leo: employ monetary pricing techniques?" tionnaires and, for each unit of pesticide Professor Pimentel's comments on our Suppose Professor Pimentel is granted sprayed per square kilometer, compute article provide us with a further op- $100,000 to assess whether, and in what the difference between benefits and portunity to clarify our viewpoints on amount, a new pesticide can be intro- costs. valuing biodiversity and ecosystem ser- duced into our environment. A first op- Even though the costs include only vices. In particular, he raises four points tion would be for him to use those dollars the noxious effects on one charismatic that we would like to address. to hire an interdisciplinary team con- species, not the many other animal First, he asks how large and diverse an sisting of for example, an ecotoxicologist, species that wiU undoubtedly be affected, interdisciplinary team should be to con- an agronomist, and an economist, who and the farmer's production is estimated duct an evaluation that meets the stan- will provide independent evaluations only imprecisely. Professor Pimentel dards set forth in our article•that is, an from three different perspectives: The would nonetheless calculate the amount evaluation that accounts for the full ecotoxicologist will estimate how differ- of pesticide that corresponds to the max- range of environmental, social, and eco- ent amounts of pesticide per unit of crop imum estimated net monetary benefit nomic consequences of an activity. The area (measured, e.g., as kg of pesticide per and communicate this figure to deci- answer, simply put, is that the size of an square km) wiU translate into different in- sionmakers. Honestly, we believe that it interdisciplinary team depends on the creases of mortality (measured, e.g., as would be better for the citizens and tax- importance of the proposed project, percentage mortality per ) in birds payers for Professor Pimentel to use the plan, or policy. The requirements for and mammals; the agronomist wiU sug- first option rather than the second• environmental impact assessments gest how these amounts translate into that is, he should hire an interdiscipli- (EIAs) usually vary according to ex- different increases of harvested biomass nary team and conduct a multicriteria pected impacts, which can be deter- (measured, e.g., as tons per square km); analysis. mined roughly by looking at the size of and the economist will figure how the Second, Professor Pimentel wonders the proposed project (as noted in farmer's budget (measured as thousands whether it is possible to investigate all European Directive 97/11/EC; European dollars per square km) is affected. Then the possible consequences of a proposed Union 1997) or by conducting a pre- Professor Pimentel can use multicriteria activity. We did not mean to imply that liminary study (as called for in US reg- analysis to assess the tradeoff between all the consequences could be investi- ulations; Council on Environmental the monetary net benefit to the farmer gated; what we meant was that the inter- Quality 1978). Then, if a comprehensive and the mortality of birds and mam- disciplinary team should do its best to study is deemed necessary, the so-called mals, an analysis that will help the deci- forget no significant consequences of a scoping process (Council on sionmaker and the citizens reach an proposed activity. How to ensure that Environmental Quality 1978, Canter informed conclusion about how many nothing significant is omitted in envi- 1996, European Union 1997) will iden- kilograms of pesticide per square kilo- ronmental impact assessments (EIAs) tify the critical social and environmen- meter can be allowed. has been the subject of much debate. The tal problems to be analyzed, depending A second option is to use the grant simple solution is to employ "checklists" on the vulnerability of the territory and money to hire economists to prepare for different categories of projects or the frequency, type, and magnitude of questionnaires asking some citizens the plans (Canter 1996, pp. 86 ff). Checklists the potential impacts. The size of the dollar value they attach to one unit of compiled on the basis of accumulated

April 2001 I Vol. 51 No. 4 • BioScience 271 Letters

experience have been extensively reported on the value of the existence of birds makers, and the public of the magnitude in, for example, the EIA Guidelines pub- (40$ or $216 or $800 per bird) is a beau- of ecosystems problems." But should we lished in 1995 by the European Union tiful demonstration of this fact. convey information on the magnitude of Centre for EIA in Manchester, United Fourth, until data collected through an ecosystem problem as a single num- Kingdom {www.art.man.ac.uk/eia/lfl2. the type of interdisciplinary investiga- ber (net monetary benefit) or as a more htm#lfl2). Of course, humans are fallible tions we suggest are available. Professor inclusive array of information? We be- and cannot forecast everything, but we Pimental comments, let us use current lieve that the monetary solution is sim- can try to do our best within reasonable data•presumably gathered through tra- plistically attractive but very misleading constraints set by time and resources. ditional accounting techniques•to give in most cases. One of the advantages of undertaking the public some idea of the magnitude this effort is that the possible conflicts of ecosystem problems. This point seems MARINO GATTO between environmentahsts and develop- to contradict Professor Pimentel's third Dipartimento di Elettronica e ers over debated questions will emerge point, that "it would be a mistake to use Informazione before a decision is taken, not afterward. willingness to pay to assess the value of Politécnico di Milano Thus less time and money are spent on birds killed by pesticides," or, in general Milano, Italy the whole decision process. terms, to assess the existence value of a Third, Pimentel says we claim that species. Why does Professor Pimentel GIULIO DE LEO contingent evaluation is the only tech- claim, at the end of his letter, that these Dipartimento di Scienze Ambientali nique capable of providing an estimate of accounting exercises are useful, while in Universita degli Studi di Parma existence values, and such techniques the previous paragraph he maintained Parma, Italy would be inappropriate for assessing, for that they are not? Moreover, it must be instance, the value of birds killed by pes- considered that even pricing techniques, References cited ticides. Let us point out that we did not if properly used (Portney 1994), require Canter LW. 1996. Environmental Impact Assess- say a contingent evaluation method is interdisciplinary investigation. In fact, re- ment. 2nd ed. London: McGraw-Hill. the only technique capable of providing spondents to questionnaires that are Council on Environmental Quality. 1978. Regula- an estimate of existence values. We said used to derive estimates of willingness to tions for implementing the provisions of the that among pricing techniques, the con- pay must be properly informed about National Environmental Policy Act. Federal tingent evaluation methods approach is the significant consequences of a pro- Register 43: 55978-56007. considered by economists to be the only posed activity. Such information can be European Union. 1997. European Council Direc- one capable of providing an estimate of provided only after an investigation that tive 97/11/EC of 3 March 1997 amending existence values, and we added that other explores the different social, environ- Directive 85/337/EEC on the assessment of problems undermined the effectiveness of mental, technical, and economic aspects the effects of certain public and private pro- any pricing technique. Indeed, a good of the problem. As we said above, pric- jects on the environment. Official Journal L deal of our paper is devoted to convinc- ing techniques and EIA-like techniques 073: 0005-0015 ing the reader that existence values can- must be compared ceteris paribus. There Portney PR. 1994. The contingent valuation not be consistently estimated via pricing is no doubt that we must "use current debate: Why economists should care. Journal techniques. Professor Pimentel's example data to inform scientists, decision- of Economic Perspectives 8: 3•17.

272 BioScience • April 2001 / Vol. 51 No. 4