SHORT COMMUNICATIONS

The Condor 90~705-708 0 The Cooper Ornithological Society 1988

ABILITY OF TWO SPECIES OF OAK WOODLAND BIRDS TO SUBSIST ON ACORNS’

WALTER D. KOENIG AND M. KATY HECK~ HastingsReservation and Museum of VertebrateZoology, Universityof California, Carmel Valley, CA 93924

Key words: Acorn Woodpecker;acorns: oaks; Scrub outdoor flight aviaries (2 m x 4 m x 3 m) containing Jay; tannins. several dead tree limbs (and a roost box for the wood- peckers)for the duration of the study, which took place Acorns are an extremely important food resourcefor December 1985 to March 1986. The woodpeckerscon- bird populationsin temperate North America (Martin sisted of a first-year male and two adult females; the et al. 1951). However, relatively little is known about jays included three first-year birds of unknown sex. the ability of animals to digest and maintain them- (First-year Scrub Jays at Hastings Reservation spend selveson acorns.Nowhere is this ignorancemore bla- as high a proportion of their time foraging for and tant than in California, where 15 speciesof oaks (Quer- handling acorns as do adults [W. J. Carmen, unpubl. cus spp.) dominate millions of hectares of land area data], and thus the lack of any adults in this sample is (Griffin and Critchfield 1972, Griffin 1977), and serve unlikely to have been misleading.) No attempt was (via their fruit) as a primary food resourcefor at least made to control the diet of birds prior to the study; all 16 speciesof birds (Martin et al. 1951). birds were fed a mixture of high-protein dog food and Of the birds dependent on acorns in foothill wood- acorns for several weeks before trials started. land regions of California, two of the most character- Prior to each trial, cageswere cleaned and all traces istic are the Acorn Woodpecker(Melanerpesformicivo- of edible items removed. Birds were weighed daily and rus) and Scrub Jay (Aphelocoma coerulescens).The given water and acorns ad libitum during trials. Be- former is well known for its habit of caching large cause of acorn storageby experimental birds in loca- numbers of acorns in specially modified storagetrees tions which resulted in loss of acorns from the cages, or granaries (MacRoberts and MacRoberts 1976). it was not possibleto measurethe massof acornseaten Acorns comprise over 50% of this woodpecker’s total during trials. Individual trials were of variable length, diet (Beal 191 1), and are critical to overwinter survival but generally lasted 13 to 15 days or until a bird lost and subsequentspring breeding (Koenig and Mumme approximately 10% of its body mass. Four trials were 1987). Acorns also comprise over 50% of the fall and terminated prematurely becauseof mass loss by the winter diet of Scrub Jays (Beal 1910; W. J. Carmen, subject,while an additional four were stopped after 4 unpubl. data). to 8 days for unrelated reasons.A total of 25 trials (12 Despite the dependenceof these specieson acorns, for Acorn Woodpeckers; 13 for Scrub Jays) were per- it is unknown to what extent, if any, they are able to formed. In order to avoid pseudoreplication(Hurlbert circumvent the effectsof tannins, a key secondaryme- 1984), trials on the sameindividual for the same acorn tabolite present in acorns. Tannins bind with proteins specieswere averagedprior to analysis. (Martin and Martin 1982) and are generally accepted Three speciesof acornswere used in trials: coastlive as defense compounds against herbivory (Bate-Smith oak (Quercus agrifolia; five trials for Acorn Wood- 1973, Tempel 1981). This study was conductedin or- peckersand seven for ScrubJays), valley oak (Q. loba- der to obtain preliminary data on the ability of Acorn ta; five trials for Acorn Woodpeckersand six for Scrub Woodpeckersand ScrubJays to overcome tannins and Jays), and canyon live oak (Q. chrysolepis;two trials to maintain their body mass exclusively on acorns. for Acorn Woodpeckersonly). Theseacorn species were chosen for their abundance at Hastings Reservation METHODS and becauseeach is a member of a different subgenus Three Acorn Woodpeckersand three Scrub Jays were (Ervthrobalanusl“red” or “black” oaks. hereafter des- captured at Hastings Reservation, Monterey County, ignated as BO], euercus[“white” oaks,hereafter WO], California, during autumn 1985 and maintained in and Protobalanm [“intermediate” oaks], respectively; Tucker 1980). Acorns were collected in autumn 1985 and frozen until used.Composition of the speciesused ’ Received 16 September 1987. Final acceptance29 here, determined from analysis of acorns collected at February, 1988. Hastings Reservation, are presented in Table 1. 2 Present address: Department of Biological Sci- Except for total available carbohydrates,which were ences, Purdue University, West Lafayette, IN 47907. virtually identical in the three species,acorns differed

[7051 106 SHORT COMMUNICATIONS

TABLE 1. Composition of acorns collected at Hastings Reservation.’

Percent Tannins available Ellerpv COtl- Hydra- Percent Percent carbo- co&it Species Subgenus densed lyzable lipids protein hydrates &J/P)

Coast live oak (Quercusugrifolia) Erythrobalanus 1.27 19.00 24.3 7.1 13.2 13.23 Canyon live oak (Q. chrysolepis) Protobalanus 0.84 8.29 16.7 3.9 12.6 11.39 Valley oak (Q. lobata) Quercus 0.89 9.96 5.6 5.5 13.6 5.53 a Crudeprotein wasmeasured by the micro-Kjeldahl method,nonstructural carboh dratesb digestionin 0.2 M NH$O, (Smith et al. 1964), and lipids by extractionin SkellySolveF. Condensedtannins were determined by a mod1.x ed acxdl2 ed vanillin method(Broadhurst and Jones1978) and hydrolyzabletannins by a modified iodate technique(Bate-Smith 1977). For additional details on tannin assays,see Faeth (1986).

considerablyin composition(Ofcarcik and Burns 1971). RESULTS Protein content was relatively low in all three species, The mean change in body mass experienced by the but was highest in Q. agrzfilia. Lipid content varied subjectsduring the trials are presentedin Table 2; mean the most, being over four times higher in Q. ugrifoliu values for each day are plotted in Figure 1. Acorn than in Q. lobata, with Q. chrysolepisintermediate. Woodpeckers (Table 2 and Fig. 1) maintained their Both condensed and hydrolyzable tannins were also body massfor the duration of the trials when fed either highestin Q. agrifolia, and were approximately equiv- Q. lobata or Q. chrysolepisacorns; woodpeckerseven alent in Q. lobata and Q. chrysolepis.Largely because gained mass on the former. In contrast, woodpeckers of higher lipid content, the estimated energetic value lost an average of 0.32 g day-l when fed Q. ugrifoliu of Q. ugrifoliu acorns (per gram) is over twice that of acorns. However, none of the five trials in which Q. Q. lobata acorns. However, this difference comes at ugrifolia acorns were fed to Acorn Woodpeckers had the cost of 43% more condensed and 91% more hy- to be terminated prior to the 14-day maximum du- drolyzable tannins (Table 1). ration of our trials. We testedfor differencesin treatmentsby an analysis of covariance(ANCOVA) in which the speciesof acorn was treated asa factor and day ofthe trial asa covariate. Resultsof this analysisindicate that the ability of Acorn Woodpeckersto maintain their body mass varied sig- nificantly with acorn species(F2,,os = 148.3, P < 0.00 1). In contrastwith these results, ScrubJays lost signif- icant body mass when fed either Q. lobata or Q. agri- folia acorns (Table 2, Fig. 1). The ANCOVA analysis showed no significant difference in their ability to maintain body mass on these two acorn species(F,,79 = 0.8, P > 0.25). (The apparent rise in body mass among ScrubJays fed Q. agrifoliu acornsbetween days 9 and 10 occursbecause several trials were terminated at day 9, leaving a reducedsample on subsequentdays.) Similar ANCOVAs comparingthe two speciesof birds (again controlling for the day of the trial) suggestthat Acorn Woodpeckerswere better able than Scrub Jays to subsiston both Q. agrifolia (F,,,, = 4.7, P < 0.05) and especially Q. lobata (F,,79 = 177.0, P < 0.001) acorns.

-8 DISCUSSION Different speciesof acorns vary significantly in com- -10 0 2 4 6 8 10 12 14 position (Ofcarcik and Bums 1971, Short and Epps 1976). Differences in lipid and tannin content, partic- Day ularly evident betweenthe WO and BO subgenera,may FIGURE 1. Mean (*SE) mass change of Acorn be dominant factors influencing preference and di- Woodpeckers (top), and Scrub Jays (bottom) fed ex- gestibilityof acorns(but seeBriggs and Smith, in press). clusively on acorns of the Quercusspecies indicated However, there is currently no consensusas to whether for the number of days given on the x-axis. Each day animals usuallyprefer WO acorns,with their relatively is considered separately for each trial, but multiple low tannins and lipids, or BO acorns, with their high trials in the same category using the same individual lipids and tannins. bird are averaged.Premature termination of sometrials The majority of work in this area has been done on results in reduced sample sizes after day 4. squirrels. Short (1976) for example, found that fox SHORT COMMUNICATIONS 101

TABLE 2. Average mass changeper day (in grams) for two oak woodland speciesfed exclusively acornsa

Coast live oak (Quercus agrrfilia) Canyon live oak (Q. chrydepis) Valley oak (Q. lobam)

Acorn Woodpecker -0.32 ? 0.46*** -0.01 + 0.03 +0.19 f 0.38*** Scrub Jay -0.30 + 0.12* - -0.39 f 0.06***

sData are the slopes (*SE) and significance of the regressions of change in body mass from start of trial on day of trial. Each day is considered separately for each trial, but multiple trials in the same category using the same individual bird are averaged. * = P < 0.05; *** = P < 0.001; other P > 0.05.

squirrels(Sciurus niger) preferred acorns from the rel- prefer Q. ugrifolia acorns over alternative, low-tannin atively low-tannin, low-lipid WO species.In contrast, speciesat Hastings Reservation and elsewherein Cal- Smith and Follmer (1972) found that digestive effi- ifornia (N. Pratini, unpubl. data). The resolution of ciency and preferencewas greaterfor Shumard oak (Q. this paradox may lie at least in part in differencesbe- shumardiz) acorns, a BO species, than for white tween natural conditionsand those offeredby our feed- oak (Q. a&a) acorns for both fox and grey (Sciurus ing trials. In our experiments, food was unlimited, and carolinensis)squirrels. More recently, Smallwood and birds were therefore probably able to compensatefor Peters (1986) demonstrated that adding tannins to the decreasedenergetic value of Q. lobata acorns by processedacorn meal decreased,and adding lipids in- simply eatingmore. However, in the wild, reproductive creased,acceptability of experimental food balls to grey successof Acorn Woodpeckersis often limited by the squirrels. size of the storage facilities in which they can cache There are as yet no comparable studiesusing birds. acorns (Koenig and Mumme 1987). Under these con- However, several studies,including Perrins (1976) on ditions, greater energetic content might be of much BlueTits (Paruscaeruleus). Marauardt and Ward (1979) greaterimportance to the birds than lower tannin levels, on domestic chickens(G&us d&nesticus),and Koenig especiallyif birds need only rarely depend exclusively (unpubl.) on Acorn Woodpeckers,indicate that tannins on acornsfor sustenanceduring the winter. In addition, adversely affect growth rate or digestibility of acorns the antibacterial properties of tannins (Grant and by birds. This negative influence is presumably a con- McMurtry 1978) may make tannin-rich acorns such sequenceof the ability of tannins to bind with proteins, as Q. agrifolia particularly well-suited to the long-term thereby making the protein unavailable to seed pred- storageneeds of Acorn Woodpeckers. ators (Bate-Smith 1973, Tempel 1981, Martin and In contrastwith the Acorn Woodpeckers,Scrub Jays Martin 1982). were apparently not able to sustain themselves indef- In the trials performed here, both Acorn Wood- initely on either species of acorn, a result probably, peckers and Scrub Jays lost body mass when fed ex- although not conclusively, attributable to a lower tol- clusively high-tannin, high-lipid Q. agrifolia acorns, erance for tannins. Scrub Jaysalso store acorns exten- while Acorn Woodpeckers, but not Scrub Jays, were sively (Grinnell 1936), but do so in the ground where able to maintain body massfor 2 weekson low-tannin, storagespace is not limited. These two differencessug- low-lipid Q. lob&a acorns (Table 2, Fig. 1). Acorn gest that Scrub Jays should prefer low-tannin acorns Woodpeckerswere alsoable to subsiston Q. chrysolepis more than do Acorn Woodpeckers,and that ScrubJays acorns,which have levels of tannins comparable to Q. should be more dependent on alternative, low-tannin lob&z, lipids intermediate to the other two acorn food resourcesthan are Acorn Woodpeckers. species, and the lowest protein concentration of the However, at least three considerationsargue for cau- three species(Table 1). Thus, tannins are most likely tion before drawing conclusionsabout the ecological to have been responsiblefor the significant difference influence of tannins on Scrub Jays in the wild. First, in the ability of Acorn Woodpeckersto maintain their although storagespace for Scrub Jays may not be lim- body mass on these speciesof acorns. ited, the bonus of the higher energetic content of a Increasedlipids did not appearto increasethe ability recovered BO acorn may still be more important than ofeitber speciesto subsiston acorns.Scrub Jays showed the lower tannin levels of a WO acorn. Second,tannins no significant difference in their ability to maintain may leach out of acornsinto the soil following storage, themselves on either Q. lob&a or Q. agrifolia acorns thus reducing the initial difference in tannin concen- despite their four-fold differencein lipid content, while trations found between acorn species.Third, the ten- Acorn Woodpecker performance was inversely corre- dency of WO acorns to sprout soon after falling to the lated with lipid content of the acornsused in the trials. ground, whereas BO acorns tend to remain dormant More precise conclusionsof the independent conse- for much of the fall and winter (USDA 1974), may quencesof lipid and tannin content on the ability of also make BO species,such as Q. agrifolia, more at- the speciesto maintain body mass would have to be tractive to ground-cachingspecies such as Scrub Jays made usingexperimental food items suchas those cre- (Smallwood and Peters 1986). ated by Smallwood and Peters (1986). Thus, the ecological significance of our results, in One prediction from these results is that Acorn terms of acorn preference and use by either Acorn Woodpeckersshould choose acorns with relatively low Woodpeckers or Scrub Jays, is unclear. Additional tannin content. Surprisingly, there is some evidence studies,both in the field on acorn use and in the lab- that this is not the case,and that Acorn Woodpeckers oratory on acorn preference and digestibility, will be 708 SHORT COMMUNICATIONS necessaryin order to understandthe coadaptationsof Woodpecker. Princeton Univ. Press, Princeton, these birds to the oaks and acorns on which they de- NJ. pend. MACROBERT~,M. H., AND B. R. MACROBERTS.1976. Social organization and behavior of the acorn We thank Bill Carmen for his help in capturingjays, woodpecker in central coastal California. Omi- Barbara Oskroba (Colorado State University) for ana- thol. Monogr. No. 2 1. American Ornithologists’ lyzing acorns,and especiallyStan Faeth (Arizona State Union, Washington, DC. University) for generouslyperforming the tannin anal- MARQUARDT,R. R., AND A. T. WARD. 1979. Chick yses. The manuscript was greatly improved by the performance as affectedby autoclavetreatment of comments of John Briggsand Peter Smallwood. Sup- tannin-containing and tannin-free cultivars of fa- port was provided by NSF grant BSR84-10809. babeans.Can. J. Anim. Sci. 59:781-789. LITERATURE CITED MARTIN, A. C., H. S. ZIM, AND A. L. NELSON. 1951. American wildlife and vlants. Dover. New York. BATE-SMITH,E. C. 1973. Haemanalysis of tannins: MARTIN, J. S., AND M. M: MARTIN. 1982. Tannin the concept of relative astringency. Phytochem- assaysin ecologicalstudies: lack of correlation be- istry 12:907-912. tween phenolics, proanthocyanidinsand protein- BATE-SMITH,E. C. 1977. Astringent tannins of Acer precipitating constituentsin mature foliage of six species.Phytochemistry 16:142 l-l 426. oak species.Oecologia 54:205-211. BEAL,F.E.L. 1910. Birds of California in relation to OFCARCIK,R. P., AND E. E. BURNS. 1971. Chemical the fruit industry. Part II. U.S. Dep. Agric. Biol. and physicalproperties of selectedacorns. J. Food Surv. Bull. 34. Sci. 36~576-578. BEAL,F.E.L. 1911. Food of the woodpeckersof the PERRINS,C. M. 1976. Possible effects of qualitative United States. U.S. Dep. Agric. Biol. Surv. Bull. changesin the insect diet of avian predators. Ibis 37. 118:580-584. BRIGGS,J. M., AND K. SMITH. In press. Influence of SHORT,H. L. 1976. Composition and squirrel use of habitat on acorn preferencesof white-footed mice acorns of black and white oak groups. J. Wildl. (Peromyscusleucopus). J. Mammal. Manage. 40:479483. BROADHURST,R. B., ANDW. T. JONES.1978. Analysis SHORT,H. L., AND E. A. EPPS,JR. 1976. Nutrient of condensed tannins using acidified vanillin. J. quality and digestibility of seeds and fruits from Sci. Food Agric. 291788-794. southern forests. J. Wildl. Manage. 40:283-289. FAETH, S. H. 1986. Indirect interactions between SMALLWOOD,P. D., AND W. D. PETERS. 1986. Grey temporally separatedherbivores mediated by the squirrel food preferences:the effectsof tannin and host plant. Ecology 671479-494. fat concentration. Ecology 67:168-175. GRANT,W. D., ANDC. M. MCMURTRY. 1978. Effects SMITH,C. C., AND D. FOLLMER. 1972. Food prefer- of condensedtannins on the growth of micro-or- ences of squirrels. Ecology 53:82-9 1. ganisms.v. 427-430. In M. W. Loutit and J.A.R. SMITH,D.. G. N. PAULSEN.AND C. A. RAGUSE. 1964. Miles [eds.], Microbial Ecology. Springer, New Extraction of total available carbohydratesfrom York. grass and legume tissue. Plant Physiol. 39:960- GRIFRN, J. R. 1977. Oak woodland, p. 383415. In 962. M. G. Barbour and J. Major [eds.], Terrestrial TEMPEL,A. S. 1981. Field studiesof the relationship vegetation of California. John Wiley and Sons, between herbivore damage and tannin concentra- New York. tion in bracken (Pteridium aquilinum Kuhm). GRIFFIN,J. R., AND W. B. CRITCHFIELD.1972. The Oecologia 51:97-106. distribution of forest treesin California. U.S. Dep. TUCKER,J. M. 1980. Taxonomy of California oaks, Agric. For. Serv. Res. Pap. PSW-82. p. 19-29. In T. R. Plumb [tech.coordinator], Sym- GRINNELL,J. 1936. Up-hill planters. Condor 38:80- posium on the ecology, management, and utili- 82. zation of California oaks. Pacific SouthwestForest HURLBERT,S. H. 1984. Pseudoreplication and the and Range Exp. Station Gen. Tech. Rep. PSW44. design of ecological field experiments. Ecol. UNITEDSTATES DEPARTMENT OF AGRICULTURE. 1974. Monogr. 54:187-211. Seedsof woody plants in the United States. U.S. KOENIG,W. D., ANDR. L. MUMME. 1987. Population Dep. Agric., Agric. Handbook 450. ecology of the cooperatively breeding Acorn