Determinants of the Vertical Distributions of Woodpecker Nest Cavities in the Sahuaro Cactus

The Condor90~791-801 0 The Cooper Ornithological Society 1988

DETERMINANTS OF THE VERTICAL DISTRIBUTIONS OF WOODPECKER NEST CAVITIES IN THE SAHUARO CACTUS

JOSEPH R. MCAULIFFE Department of BiologicalSciences, University of Nevada, Las Vegas,NV 89154

PAUL HENDRICKS Department of Zoology, WashingtonState University,Pullman, WA 99164

Abstract. In the SonoranDesert, Gilded (Northern) Flickers(Colaptesauratuschrysoides) and Gila Woodpeckers (Melanerpes uropygialis)excavate nest cavities within the giant sahuaro cactus. Vertical distributions of nest cavities excavated by the two woodpecker speciesdiffer in height and breadth. Flicker cavities are usually restricted to within 3 m of the stem apex. Nest cavities excavatedby Gila Woodpeckersare found over a considerably broader vertical rangeof heights;their averageheight is also lower. Competitive preemption of potential excavation sitescannot fully explain this differential utilization of vertical space. Rather, the initial positioning of nest cavities is constrained by cactus anatomy. Different nest-cavity sizes and wood-chiseling capabilities of the two woodpecker specieslimit the vertical range over which nest cavities can be excavated. In addition, an excavation by a flicker often results in death of the sahuaro.With this mortality, a long-running record of flicker nest-cavity heights is not preserved. In sahuaroscontaining only Gila Woodpecker cavities, the original heights of old excavations are preserved as the sahuarocontinues to grow, thus contributing to the greater vertical range of observed cavity heights. Key words: Carnegieagigantea; Colaptes auratus; Gila Woodpecker;Gilded Flicker: Me- lanerpesuropygialis; resource partitioning: sahuaro;Sonoran Desert; woodpeckers.

INTRODUCTION pecker nest-cavity excavation can result in death of sahuaros. In the Sonoran Desert, Gila Woodpeckers (Mel- The purpose of our paper is threefold: (1) to anerpes uropygialis) and Gilded (Northern) present comparative data on the vertical heights Flickers (Colaptes aura&s chrysoides)excavate of nest cavities excavated in sahuaros by Gila nest cavities within the giant sahuaro cactus, Woodpeckers and Gilded Flickers, (2) to deter- Carnegiea gigantea. Patterns of distribution of mine morphological characteristicsof sahuaros nest cavities within stemshave only recently been that may influence the location of nest-cavity reported (Inouye et al. 198 1, Korol and Hutto excavations, and (3) to examine the impacts of 1984). The above two studies were primarily nest-cavity excavation on sahuaros. concernedwith compassorientation of nest-cavity entrance holes. The vertical positioning of MATERIALS AND METHODS nest cavities has not been studied in detail. Information on nest-cavity locations was col- Aside from the anecdotal reports of Gilman lected in Organ Pipe Cactus Monument (ORPI), (19 15) and Brandt (195 I), patterns of nest-cavity Pima County, Arizona, in March 1983. The study excavation within internal tissuesof the sahuaro area was located on an alluvial fan approximately have not been described. Constraints imposed 1 km west of the mouth of Alamo Canyon on by internal anatomy may provide an important the west side of the Ajo Mountains (32.1”N, key to understanding patterns such as the height 112.8”W). distributions of nest cavities. Furthermore, the The Gilded Flicker is the larger of the two effects of tissue removal during nest-cavity ex- woodpecker species;the speciesresponsible for cavation on survival of sahuaroshave not been excavation of a nest cavity could easily be de- examined in depth, although Steenbergh and termined by the size of the cavity entrance hole Lowe (1976,1977,1983)commentedthatwood- (Gilman 19 15, Brandt 195 1) (Table 1). The presence of wear around the perimeter of the hole indicated repeated entry to the cavity and was I Received 11 January 1988. Final acceptance 28 used to distinguish actual nesting cavities from May 1988. shallow holes.

[7911 792 JOSEPHR. McAULIFFE AND PAUL HENDRICKS

casementsof nesting cavities (hereafter descrip- tively referred to as nest-cavity “boots”) can be found on the ground long after sahuaros containing them have died and the soft tissueshave decayed. We determined the dimensions and structure of nest cavities excavated by each woodpeckerspecies from measurements of these nest-cavity boots (Fig. 1). Occasionally, boots were partially destroyed or decayed, permitting only a partial set of measurements.Between 1983 and 1987 we collectedand measuredcavity boots whenever we found them at sitesthroughout the Sonoran Desert. Details of the internal anatomy of sahuaros were determined from cross sections of seven individuals that were toppled in storms in 1986 and 1987. Total height was measured and cross FIGURE 1. Diagramof a nest-cavityboot excavated by a Gilded Flicker. Descriptionsof dimensionsA-F sections were cut with a saw at distances of 25, arelisted in Table 1. The letterG refersto thediameter 50, 100, 200, and 300 cm from the apex of the of the severedwoody skeleton;H refersto the pro- main stem. Measurements taken at crosssections portion of individual woody skeletalrods that were included (1) average diameter, (2) the maximum severed. and minimum thicknessof the outer cortex measured from the outside surface of the internal For sahuaroscontaining entrance holes to nest skeletal rods to the ridges of the external pleats, cavities within the main stem, the nest-hole type (3) the diameter of the ring of woody skeletal (small for Gila Woodpeckers or large for Gilded rods, and (4) the average radial thickness of in- Flickers) was identified. Heights of both the en- dividual skeletal rods. trance hole and the sahuaro were measured by In June 1987, the possiblecontribution of nest- standing 25 m from the base of the sahuaro and cavity excavation by flickers to sahuaro mortal- measuring the apparent height of the entrance ity was examined at a site located on an alluvial hole and the stem apex on a vertical scale held fan on the south side of the Tortolita Mountains, 1 m from the observer’seye. The apparent heights approximately 40 km northwest of Tucson, Pima and actual heights represent sides of similar tri- County, Arizona (32.5”N, 111. low). The occur- angles. Apparent heights were converted to ac- rence of flicker nest-cavity boots in the remains tual heights by multiplying by 25, the ratio of of dead sahuaroswas compared with the occur- the distance from the eye to the cactusand from rence of flicker nest cavities in live sahuaros. the eye to the base of the scale. Only the remains of dead sahuarosthat still pos- In responseto excavation by a woodpecker, a sessedremnants ofthe epidermis were measured. sahuaro seals off the cavity with hard, suberin- Sahuarosin more advanced statesof decay were impregnated layers of cork. These durable en- not included. This insured that a more decay-

TABLE 1. Dimensionsof nest-cavityboots originally excavated by Gila Woodpeckersand Gilded Flickers. Lettercodes designating various measurements are picturedin Figure 1. All linear measurementsare in centimeters.

Chla Woodpeckers Gilded Flickers Code Description n R SD n R SD A. Verticalcavity length 13 26.5 5.83 31 36.2 6.56 B. Chamberdepth 13 9.2 1.58 34 13.9 2.10 C. Chamberwidth 13 13.3 3.49 37 15.7 2.31 D. Horizontaldepth 9 12.2 3.03 29 19.2 2.82 E. Verticalentrance diameter 6 5.1 0.45 20 6.8 0.92 F. Horizontalentrance diameter 15 5.5 0.76 42 8.6 1.01 WOODPECKER NEST-CAVITY HEIGHTS IN SAHUAROS 793 resistant cavity boot would be found with the skeletal remains if the cactus actually contained a nest cavity at the time of death. Because of eventual decay of cavity boots, inclusion of dead cacti in extreme states of decay could result in an underestimate ofthe original frequency ofnest cavities. For each dead sahuaro, evidence of possible causes of mortality, including lightning strikes and internal tunneling by packrats (Neot- oma albigula) was recorded. Sahuaros in which branches or portions of the stem were haphaz- ardly scattered in several different compass di- FIGURE 2. Height distributions of nest cavities rections were classedas possible lightning strike within the main stems of sahuaros.The boundary be- victims. The original heights of fallen, dead sa- tween the shaded and unshaded regions indicates the huaros were measured directly when the entire maximum height of the stem, the stem apex. The bro- stem was intact. For dead sahuaroswhose heights ken lines indicate the upper 3 m of stem apexes for could not be measured directly, the diameter of cacti ofdifferent sizes.Open trianglesindicate locations of individual nestsof respectivespecies when the other the woody skeleton was measured 1 m above the speciesis absent from the same sahuaro.Solid circles base. This skeletal measurement was used to es- indicate nest heights when nest cavities of the other timate original height since there is a tight linear speciesare located in the same sahuaro. relationship between these two variables (Y = -62.3 + 41.04X; r2 = 0.90; X and Y are the located-almost exclusively within the upper 3 skeleton diameter and sahuaro height, respec- m of the stem, regardlessof cactusheight. There tively, in centimeters). This relationship was de- is no correlation between the absolute values of rived from data collectedbetween 1984 and 1987 residuals about a regressionfit through the data on 35 fallen sahuarosin the vicinity of Tucson, shown in Figure 2A and sahuaro height (r, = Arizona. 0.096, P > 0.45).

RESULTS COMPETITIVE RESTRICTION OF DISTRIBUTION OF NEST-CAVITY NEST-CAVITY LOCATIONS? HEIGHTS We initially hypothesized in 1983 that interspe- Cavities excavated in the main stem by Gilded cific competition for nest-site locations along the Flickers are almost always located within 3 m of vertical stems of sahuaroswas an important de- the stem apex, regardlessof sahuaro height (Fig. terminant of nest-cavity distributions. The pres- 2A). Throughout nearly the entire range of sa- ence of nest cavities excavated by Gila Wood- huaro sizes, the average height of Gila Wood- peckersin lower sectionsof sahuaro stems could peckernest cavities is lessthan the averageheight potentially preempt the excavation of cavities by of Gilded Flicker nest cavities (Fig. 2B). In ad- Gilded Flickers and restrict Gilded Flickers to dition, the vertical range in the location of Gila upper sections of the stem. Conversely, the oc- Woodpecker nest cavities increaseswith increas- cupation of the apical region by cavities exca- ing cactusheight. In sahuaros6 to 8 m tall, Gila vated by Gilded Flickers might displace Gila Woodpecker nest cavities are found only within Woodpecker excavation activities to lower parts the top 3 m of the main stem. However, in 12- of the stem. Such interactions could also poten- to 15-m tall sahuaros,the vertical range of nest tially restrict the vertical range over which cav- locations extends from within 2 m of the cactus ities of both speciesare found. apex to as much as 8 m below the apex (Fig. 2B). To test these hypotheses, we subdivided the This pattern is statistically significant: there is a data collectedat ORPI in 1983 into three groups: positive correlation between the absolute values (1) sahuaroscontaining only the nest cavities of of residuals about a regression fit through the Gilded Flickers, (2) sahuaroscontaining only the data in Figure 2B and sahuaro height (Spear- cavities excavated by Gila Woodpeckers, and (3) man’s rank correlation, r, = 0.353, P < 0.0001). sahuaros in which cavities excavated by both This contrastssharply with the invariant vertical species were present. For each woodpecker range over which Gilded Flicker nest cavities are species,we compared heights of nest cavities in

WOODPECKER NEST-CAVITY HEIGHTS IN SAHUAROS 795

Thickness of Outer Cortex (cm)

0 10 20 0 I

100

200

FIGURE 3. A. Cross section of a sahuaro approximately 50 cm from the stem apex. Note extremely thin, indistinct woody skeleton. B. Cross section taken approximately 2 m below the stem apex. C. Standing skeleton of a sahuaro. Small horizontal processes on upper left sideare skeletalremains of small branches. 300 an outer cortex and a central pith. The pith is Ix Gila Woodpecker Cavities surrounded by a skeleton consisting of a ring of parallel woody rods (Fig. 3). Near the base of the cactus, the skeletal rods can be several centi- I d metersthick. Moving upward from the base,these x Horizontal Depth: Gilded Flicker Cavities woody rods become thinner and weaker. At 0.5 m below the stem apex, individual skeletal rods FIGURE 4. Thicknessof outer cortex of sahuarosas a function of distance below the stem apex. The left ranged between 5 and 6 mm in diameter. In the margin indicatesthe boundary of the outer cortex and same cacti, the skeletal rods were 9 to I4 mm in the woody skeleton.The dark shadedregion showsthe diameter 2 m below the stem apex. The succulent approximate minimum thicknessesmeasured for the layer of the outer cortex is also thinnest near the outer cortex; the right-hand margin of the light shaded stem apex and attains a maximum thickness of region denotes maximum cortical thicknesses measured from the same sections.Open circles are actual approximately 16 cm with increasing distance data on minimal cortex thickness;solid circlesare mea- below the apex (Fig. 4). surementson maximal thickness. Nest cavities excavated by the two woodpecker speciesdiffer in size and shape. These differ- encesapparently affect the ability of each species contact with the woody skeleton (Fig. 5D). How- to excavate cavities in various parts of the stem. ever, horizontal depth of cavity excavation by Nest cavities excavatedby Gila Woodpeckersare Gila Woodpeckersusually halts when the woody considerably smaller than Gilded Flicker nest skeleton is encountered and typically none of the cavities (Table 1, Figs. 5A, B). The average hor- woody rods are severed (Table 2). Occasionally izontal depth of Gila Woodpecker cavities is ap- (two of 14 nest-cavity boots examined), Gila proximately 12 cm; average horizontal depth of Woodpeckerswill chisela 5- to 6-cm hole through Gilded Flicker cavities is approximately 19 cm. the cortex and two or three skeletal rods, and The smaller nest cavities of Gila Woodpeckers then excavate the larger chamber entirely within are somewhat flattened; the inner chamber has the confines of the woody skeleton in the soft greater width than depth (Table 1, Figs. 5A-C). pith. However, by excavating nest cavities in Gila Woodpecker cavity boots usually bear ver- lower portions of the stem, especially more than tical, linear scarson the back side that indicate 2 m from the stem apex, Gila Woodpeckers can 796 JOSEPH R. McAULIFFE AND PAUL HENDRICKS

FIGURE 5. A, B. Front and side views, respectively,of Gilded Flicker (left) and Gila Woodpecker (right) nest- cavity boots. See Table 1 for averagedimensions. C. Top view of a Gila Woodpecker nest-cavity boot. Notice the flattened back of the boot (lower side of photo) and the compressedcross-sectional shape of the nesting chamber. D. Rear view of the same nest-cavity boot. Parallel lines indicate the former contact with, but not penetration of, the sahuaroskeleton. Scale bar = 10 cm. E. Top view of three Gilded Flicker nest-cavity boots. Circular patterns on tops of the boots are the scarsdue to complete severing of the woody skeleton. F. Cross WOODPECKER NEST-CAVITY HEIGHTS IN SAHUAROS 797

TABLE 2. Extent of severingof skeletalrods by Gila Woodpeckersand Gilded Flickers during nest excavation. Data taken from nest-cavity boots associatedwith remains of dead sahuaros.

Percentage of skeletal rods severed by nest-cavity excavation 0 l-25 26-50 51-75 76-100

Gila Woodpecker nest cavities 9 4 0 0 1 Gilded Flicker nest cavities 2 2 5 5 27

apparently avoid the necessity of chiseling apex. The closer a Gilded Flicker excavates its through any part of the woody skeleton (Fig. 4). cavity near the apex (as indicated by the diameter The considerably greater horizontal depth of of the skeletal scar on top of a cavity boot), the Gilded Flicker nest cavities precludes their ex- greater is the likelihood that all of the skeletal cavation entirely within the outer cortex. No- rods will be cut. The estimated percentage of where along the stem is the outer cortex thick woody rods severedis negatively correlated with enough to entirely accommodate a nest cavity the diameter of the skeletal scar measured on the much over 15 cm in horizontal depth (Fig. 4). top of cavity boots (r, = -0.53, n = 35, P = Gilded Flickers must therefore chisel through the 0.001). skeleton. By excavating nest cavities near the Only rarely do Gilded Flickers exhibit other cactusapex, only relatively thin, soft woody rods ways of dealing with the barrier imposed by the would require removal. Flickers excavating nest woody skeleton. Out of 35 nest-cavity boots col- cavities at increasing distances from the stem lected, only one was positioned entirely within apex would encounter increasingly larger and im- the outer cortex. The nest chamber was consid- penetrable skeletal rods. erably flattened. From the same sample of nest Flicker nest-cavity bootsbear distinct scarsthat cavities, one was constructedby chiselinga small, reveal their original proximity to the stem apex 7- to 8-cm diameter hole through the cortex and as well as the number of woody rods that were three woody rods. This long throat extended hor- completely severed during excavation. These izontally to the pith. The entire chamber was scarscan be clearly seen on the upper side of a excavated within the soft pith on the interior of nest-cavity boot (Fig. 5E). If the entire woody the woody skeleton. The nest cavity was consid- skeleton was severed, the scar will have a ring- erably smaller in diameter than others,indicating like pattern on the top of the nest chamber. If the probable restriction on internal diameter im- only a portion of the woody skeleton was sev- posed by such positioning. ered, the scarwill have a semicircular or crescent pattern when viewed from above. Out of the 35 IMPACTS OF NEST-CAVITY EXCAVATION nest cavities examined, 66% possessedscars re- ON SAHUAROS vealing the entire skeleton was severed during The two woodpecker species differ considerably excavation (Table 2). In live sahuaros, the di- in the damage they inflict on sahuaros during ameter of the woody skeletal ring increaseslin- nest excavation. The typical Gila Woodpecker early with distance from the stem apex (Fig. 6B). nest cavity located entirely within the soft outer Therefore, the approximate distance between the cortex apparently causes little harm. The exca- site of original cavity excavation and the stem vation heals over with the same response as to apex can be estimated from the diameter of the any wound. The main damage inflicted is the scar present on the top of the nest-cavity boot. loss of a relatively small amount of water-storing In most boots, the diameters of these scarsis less cortex along with a small amount of photosyn- than 12 to 14 cm (Fig. 6A), indicating that these thetic surface at the site of the nest-cavity open- nestswere originally excavatedvery near the stem ing (Fig. 5F). c section through a relatively small nest cavity excavated by a Gila Woodpecker. None of the skeletal rods are severed.This section was taken approximately 200 cm below the stem apex. G, H. Cross sectionsthrough nest cavities excavated by Gilded Flickers. In each, all of the woody rods were severed. I. Nearly complete healing of woody tissue around a Gilded Flicker nest cavity that originally severed all of the skeletal rods. 198 JOSEPHR. McAULIFFE AND PAUL HENDRICKS

amined, only one showeda nearly complete heal- A ing and regrowth of woody tissue around the nest cavity (Fig. 51). Sahuaros weakened by Gilded Flicker nest cavities may be especially subject to decapitation during violent storms. We have commonly observed decapitated sahuaros with the decaying remains of the apex near the base.

I It is far more rare to encounter living, decapi- 10 20 tated sahuarosin which the fallen apex has com- Diameter of Skeletal Scar pletely decayed (a process that probably takes on Cavity Boot (cm) about 10 to 20 years at most in the Sonoran Desert near Tucson (Ebert and Zedler 1984). This suggeststhat once decapitation occurs, death of the cactus is common within the period of time required for complete decay of the fallen apical section. In addition to providing skeletal support, the woody rods comprise the principal vascular tissue. The apical portions of cacti located above Gilded Flicker nest cavities are commonly shrunken with considerably smaller cross-sec- 200 tional diameters than normal. This indicates the diminished transport of water to the apex due to 1 the severing of the vascular tissue. The exterior of the cactusis coveredwith accordion-like pleats that allow for expansion or contraction of internal volume without changing external surface 300 - area. The loss of internal volume greatly in- creasesthe surface area to volume ratio which could acceleratethe rate of unavoidable evapo- 0 10 h0 rative water loss on a per unit volume basis. It is possible that extreme water deficits could di- Diameter of Woody Skeleton (cm) rectly cause death of the apex. FIGURE 6. A. Diametersof the skeletalscars found Increased surface area/volume ratios of the on topsof Gilded Flicker nest-cavityboots. B. Data apexcan have another seriousresult. Tissue death taken from crosssections of fallen sahuarosshowing due to freezing is a principal source of mortality the diameterof the ring of woodyskeletal rods at different&stances below the stemapex. Individual solid in northern and eastern portions of the range of symbolsindicate actual measurements; the shadedre- the sahuaro (Steenberghand Lowe 1977, 1983). gion indicatesthe approximaterange of values(fitted An increase in the surface/volume ratio would by eye). increasethe rate ofheat lossfrom internal tissues, thereby increasing the susceptibility to freezing. While a nest cavity may not be the direct cause The complete or even partial severing of the of mortality in all cases, the interaction of the woody skeleton by Gilded Flickers has more det- effectsof the excavation with the effectsof freez- rimental impacts. One consequenceis the loss of ing may bring about increased mortality (see skeletal support. Cross sections made at the lo- SteenberghandLowe 1976, fig. 23; 1983, p. 128). cation of Gilded Flicker cavities in fresh, wind- Data collected from dead sahuarosat the Tor- thrown sahuarosshow that only a relatively thin tolita Bajada site provided information on the layer of soft outer cortex and epidermis may re- extent of mortality that may be attributed to nest- main to support the stem apex if all of the woody cavity excavation by Gilded Flickers. Of the 55 skeletal rods are cut (Figs. SG, H). Complete dead sahuarosexamined, 2 1.8% contained Gild- regeneration of severed skeletal rods rarely oc- ed Flicker nest cavities. No Gila Woodpecker curs. Of all the 35 flicker nest-cavity boots ex- nest-cavity boots were found associatedwith any WOODPECKER NEST-CAVITY HEIGHTS IN SAHUAROS 799

of the dead sahuaros. Destruction by lightning TABLE 3. Distribution of Gilded Flicker nest-cavity was an apparent sourceof mortality in 30.9% of boots among live and dead sahuaros. the dead sahuarosexamined. Extensive internal Sahuaros taller tunneling by the woodrat, Neotoma albigula, may Sahuaros 5 to 8 m tall than 8 m have been a principal causeof death in only one Nest cavities Nest cavities Sahuaro _ of the 55 cacti examined. A possible cause of conditmn PFZS~llt Absent Present Absent death could not be determined in 45.5% of the Live 0 91 6 15 dead sahuaros. Dead 7 28 5 15 When live and dead sahuaroswere compared, x2 = 18.96, x* = 0.07, Gilded Flicker nest cavities were more com- P < 0.001 P > 0.10 monly found in dead cacti than in live cacti. Flicker nest-cavity boots were found in the remains of 2 1.8% of dead sahuaroswhose original cottonwoods and willows (Gilman 1915, Bent height was estimated at more than 5 m. Flicker 1939). Yet, nest cavities excavated by Gila nest cavities were found in only 5.3% of living Woodpeckers are small enough that they can be sahuarostaller than 5 m (x2 = 10.4, P < 0.01). positioned entirely within soft tissuesof sahuaros The greater incidence of flicker nest cavities in without ever penetratingthe woody skeleton.The dead sahuarosindicates that once Gilded Flick- typical flattening of Gila Woodpecker nest cav- ers excavate cavities in sahuaros,these damaged ities against the inner skeletal rods (Figs. 5C, D) cacti are recruited into the population of dead suggestsan avoidance of chiseling on the skele- sahuaros at a faster rate than are cacti undam- ton. aged by flickers. Small sahuaros are apparently The larger nest cavities of Gilded Flickers can- more prone to mortality as a result of flicker nest- not fit entirely within the soft outer cortex of cavity excavation than are older sahuaros. For sahuaros. Hence the woody skeleton must be sahuarosbetween 5 and 8 m tall, 20% of the dead penetrated. This presentsa dilemma for Gilded individuals contained nest cavities, whereasnest Flickers because their wood-chiseling capa- cavities were never found in living sahuaros.In bilities are considerably less than those of Gila the larger size classesof sahuaros, the presence Woodpeckers.The Gilded Flicker foragesalmost or absence of Gilded Flicker nest cavities was exclusively on the ground; much of its diet con- evenly distributed among live and dead sahuaros sistsof ants (Bent 1939, Short 1982, Vander Wall (Table 3). and MacMahon 1984). Accompanying the evo- lutionary specialization of ground feeding is a DISCUSSION commensurate loss of wood-chiseling ability Two processesnot involving competition for nest (Short 1982). Flickers possessnarrow, pointed sites can account for much of the difference in bills as opposedto the stouter, chisel-tipped bills the vertical distributions of Gila Woodpeckerand of Gila Woodpeckers and other woodpecker Gilded Flicker nest cavities. These are (1) con- speciesmore capable ofwood excavation. Gilded straints imposed by sahuaro anatomy on initial Flickers apparently minimize the difficulty of nest-cavity excavation and (2) differencesin sur- chiseling through the woody skeletonsof sahua- vival of sahuaros following excavation of nest ros by excavatingnest cavities near the stem apex. cavities by the two species. In this location, the skeletal rods are extremely The two woodpecker species experience dif- thin and present the least resistanceto removal. ferent constraints imposed by sahuaro anatomy But by excavating near the cactus apex, usually becauseof their different body sizes (and hence all of the thin skeletal rods are severed, resulting requirements for different sizes of nest cavities) in much more severe damage. This damage ap- and wood-chiseling abilities. The woody skele- parently is often directly or indirectly responsible ton of a sahuaro is undoubtedly the structure for death of sahuaros. most resistant to removal by the chiseling efforts Even if Gilded Flickers and Gila Woodpeckers of either species.Gila Woodpeckers are capable constructednew nest cavities in the same relative of chiseling through wood considerably tougher location near the stem apex, differential mortal- than the skeletal rods of sahuaros. Along river ity of cacti containing nest cavities of different floodplains in Arizona, Gila Woodpeckers com- specieswould eventually produce dissimilar dis- monly excavate nest cavities within trunks of tributions of nest-cavity heights. A newly exca- 800 JOSEPH R. McAULIFFE AND PAUL HENDRICKS vated nest cavity leaves an external scar on the cactus whose absolute height remains the same INFLUENCES OF COMPETITION over time-just like initials carved in the bark Our initial interpretation of the different vertical of a tree. All vertical growth occurs at the mer- distributions of nest cavities excavated by Gila istem tissue at the apex of the main trunk or Woodpeckersand Gilded Flickers was that com- arms. If nest cavities do not lead to death, a petition was a major factor determining this pat- record of original heights of older excavations tern. Competition may actually have very little, will be preserved as a cohort of sahuarosgrows if anything, to do with it. Although a slight shift in height. Simultaneously, growth ofthe cacti will in Gila Woodpecker nest-cavity heights was as- increase the vertical range available for the ex- sociatedwith the presenceof Gilded Flicker nest cavation of new nest cavities. The vertical dis- cavities, it is impossible to clearly separatecause tribution of Gila Woodpecker nest cavities there- and effect with these data. This is becauseinfor- fore representsa long recorded history of recent mation on the sequence of nest-cavity excava- and old nest-cavity excavations. This accumu- tions is unknown. Rather than Gila Woodpeck- lation of nest cavities over time is probably large- ers shifting their excavation activities in response ly responsible for the increasing range of Gila to the original presenceof a Gilded Flicker cav- Woodpecker nest-cavity heights with increasing ity, Gilded Flickers may actually choose to ex- sahuaro height (Fig. 2B). For Gilded Flickers, cavate cavities in sahuaros where the average however, excavation of new nest cavities occurs height of Gila Woodpecker nest cavities is lower. near the stem apex and these excavations com- Either scenario could produce the same pattern. monly lead to death of the sahuaro. In this case, Observations of agonistic behavior of Gila old nest cavities would rarely, if ever, be found Woodpeckers toward Gilded Flickers and the in lower sectionsof the stem, regardlessof cactus eviction of a pair of Gila Woodpeckersby a Gild- height, because only a relatively brief span of ed Flicker (Martindale 1982) indicate that these time elapses between the time of cavity exca- two woodpecker speciesdo compete. However, vation by a Gilded Flicker and death of the sa- these types of interactions may not have much huaro. This mortality of sahuaros (and of the effect on the overall vertical distributions of nest Gilded Flicker nest cavities they contain) ex- cavities excavated by the two species.One com- plains the triangular zone in the lower right sec- petitive effect that may conceivably occur is a tion of Figure 2A in which Gilded Flicker nest negative association of nest cavities of the two cavities are completely absent. speciesin the same sahuaro. Evidence for this Patterns of sahuaro use by Gila Woodpeckers kind of competitive displacement could be ob- observed by other authors can be explained by tained by collecting data on the numbers of sa- the accumulation of nest cavities over time. Kor- huaros in a population that fall within four dis- 01 and Hutto (1984) reported that taller sahuaros crete categories: (1) those with only Gila more frequently contained Gila Woodpecker nest Woodpecker cavities, (2) those containing only cavities than did smaller sahuaros. In addition, Gilded Flicker cavities, (3) those containing cav- among sahuaros containing at least one Gila ities excavatedby both species,and (4) cacti lack- Woodpecker nest cavity, the number of nest cav- ing nest cavities altogether. The observed num- ities per sahuaro was positively correlated with bers of sahuaroscontained in each categorycould sahuaro height. Korol and Hutto suggestedthat be compared with the null expectation, given a thesepatterns may be the result of problems such random distribution of the observed number of as excavation difficulties, increased predation nest cavities of each speciesamong sahuaros. rates, or thermal stressassociated with nesting Although our initial hypothesisbased on com- in shorter sahuaros.An alternative explanation petition-induced resource partitioning was ap- is simply that younger sahuaroshave been avail- pealing to us, it also temporarily shackledus with able for nest-cavity excavation for only a rela- observational “blinders” that may have actually tively short period of time, whereas large, older delayed the observation and consideration of sahuarospossess an accumulated record of past constraints imposed by sahuaro skeletal anato- nest excavations that may be as much as a cen- my and the implications of flicker-induced sa- tury older (see data on age-height relationships huaro mortality. Understanding the woodpecker for sahuaros in Steenbergh and Lowe [1977, nest-height distributions, and possibly other pat- 19831). terns of differential resource utilization requires WOODPECKER NEST-CAVITY HEIGHTS IN SAHUAROS 801 detailed knowledge of the natural histories of the KLEINBAUM,D. G., ANDL. L. KUPPER. 1978. Applied component species.Without such knowledge,the regressionanalysis and other multivariate methods. Duxburv Press. North Scituate. MA. uncritical acceptanceof a given pattern as prima KOROL,J., AND k. L. H&TO. 1984. Factors affecting facie evidence for a theoretically expected set of nest site location in Gila Woodpeckers. Condor ecological interactions is far more likely, simply 86:73-78. becauseof the lack of alternative explanations. MARTINDALE,S. 1982. Nest defenseand central place foraging: a model and experiment. Behav. Ecol. ACKNOWLEDGMENTS Sociobiol. 10:85-89. SHORT,L. L. 1982. Woodpeckersof the world. Del- We thank the United StatesNational Park Serviceper- aware Mus. Nat. Hist. Monogr. Ser. No. 4. Green- sonnel at Organ Pipe Cactus National Monument: es- ville, DE. peciallv Harold Smith and Bill Mikus. for orovidina STEENBERGH,W. F., ANDC. H. LOWE. 1976. Ecology logistical support for part of this study in I983 and of the saguaro.I. The role of freezing weather on Carla Mae McAuliffe for her expert editing. a warm-desert plant population, p. 49-92. In Re- searchin the parks.Natl. Park Servicesymposium LITERATURE CITED series No. 1. U.S. Government Printing Office, BENT,A. C. 1939. Life histories of North American Washington, DC. woodpeckers.U.S. Natl. Mus. Bull. 174. STEENBERGH,W. F., ANDC. H. LOWE. 1977. Ecology BRANDT,H. 1951. Arizona and its bird life. The Bird of the saguaro.II. Reproduction, germination, es- tablishment, and survival of the youngplant. Natl. ResearchFoundation. Cleveland. OH. Park Service Sci. Monogr. Ser. No. 8. U.S. Gov- EBERT,T. A., AND P. H. ZEDLER. 1984. Decompo- ernment Printing Office, Washington, DC. sition of ocotillo (Fouquieria splendens)wood in STEENBERGH,W. F., ANDC. H. LOWE. 1983. Ecology the Colorado Desert of California. Am. Midl. Nat. 111:143-147. ofthe saguaro.III. Growth and demography.Natl. Park Service Sci. Monogr. Ser. No. 17. U.S. Gov- GILMAN, M. F. 1915. Woodpeckers of the Arizona ernment Printing Office, Washington, DC. lowlands. Condor 17:15 l-l 63. INOUYE,R. S., N. J. HLJN~LEY,AND D. W. INOUYE. VANDERWALL, S. B., AND J. A. MACMAHON. 1984. 1981. Nonrandom orientation of Gila Wood- Avian distribution patterns along a Sonoran Des- pecker nest entrancesin saguarocacti. Condor 83: ert bajada. J. Arid Environ. 7~59-74. 88-89.