SEASONAL ANALYSIS OF A SOUTHWESTERN NEW MEXICO RIPARIAN BIRD COMMUNITY WILLIAM H. BALTOSSER, Department of Biology, New Mexico State University, Las Cruces, New Mexico 88003 (present address: 5022 La Cienega NW, Albuquerque, New Mexico 87107) The lower Gila River Valley of southwesternNew Mexico hassome of the finestriparian habitat and the mostdiverse association of wildlifeto be found in the entire lower Colorado River drainage (Zimmerman 1970, 1975; Hubbard 1971; Johnson et al. 1974). Over two-thirds of New Mexico's total of 449 speciesof birdsare known from the valley (Zimmerman 1975, Hub- bard 1977). In addition, the New Mexico portion of the Gila River Valley containsthe •reatest diversity of raptorsin thelower Colorado River drainage and the largest number of endangered, threatened and peripheral bird species(johnson et al. 1974). The presentstudy was conductedto examine and quantifyseasonal changes in avian composition,density and diversity along the Gila River. Comparisonsof the presentstudy with similarstudies indicate numerous similarities,but the resultsreaffirm the richnessof the avi- fauna in the New Mexico portion of the Gila River Valley. METHODS I establishedtwenty contiguousstudy plots along the Gila River on the northeastside of the U.S. Highway 180 bridge (T15S, R17W, Sec. 33 & T16S, R17W, Sec. 4) in Grant County, New Mexico. Two types of habitat were recognizedand subdividedaccordingly: sandy riverbottom(5.30 ha) and adjacent woodlands (9.15 ha) composed of Fremont's Cottonwood (Populusfremontii), Box-elder (Acer negundo), and Goodding's Willow ($alix gooddingii)stands (see Figures1 & 2). Twenty-threeweekly surveys were made between4 Januaryand 30 June 1975. Surveyswere conducted accordingto methodsoutlined under the heading"Wi•ter Bird-Population Study" in AudubonField Notes (Anonymous1950). The spot-mapmethod (Williams1936, Kendeigh 1944) was used in conjunctionwith the former methodsto estimatebreeding bird populationsduring May and June. Den- sitiesand size classesof trees in the wooded standswere measuredduring winter months by direct counts and checked against low-altitudeaerial photographs.No effort was made to measure vegetationin the riverbottom becausethere was virtuallyno terrestrialand very little aquaticvegetation. Speciesdiversity was calculatedusing the Shannon-Weaver(1949) infor- mation function, $ H' = -•' p• In p• i=1 where $ is the numberof species,and p, is the proportionof the total number of individuals consistingof the ith species. This measure (H •) has two separatecomponents, species richness ($) and the equitabilityor evennessof speciesabundance (Lloyd and Ghelardi 1964, Tramer 1969). Species Westerri Birds 17:115-131. 1986 115 NEW MEXICO RIPARIAN COMMUNITY richnessis simplythe number of speciesin the sample. To measurethe even- ness of abundance, I used the index d' = l-l•/l-I • max in which/-/• max is In $. This index representsthe ratio of the observeddiversity to the maximum diversitypossible for the same numberof species.It has a maximumvalue of one when all speciesare equally abundant. AREA DESCRIPTION The studyarea is at an elevationof 1370 m and wasthe largestsingle stand of riparian woodland in the immediatevicinity. Surrounding areas included abandoned farm land to the west, farmed land to the east, and land similarto that of the study area to the north and south. Box-elder, cottonwoodand willowtrees within the studyarea produceda combinedcanopy cover of ap- proximately80%, excludingthe riverbottomand two smallopen areas. Box- eldersattained heights up to 15 m, averagedensities of 41 treesper ha (ex- cludingsaplings, which averaged102 per ha), and DBH (diameterat breast height)values from 26 to 64 cm (basedon pooledaverages from each of the 12 wooded studyplots). Fremont'sCottonwoods reached heights of 27 m, averagedensities of 14 treesper ha (virtuallyno saplings),and DBH values from 36 to 128 cm. Goodding'sWillow standsattained heights of 12 m, averagedensities of 11 treesper ha, and DBH valuesfrom 29 to 47 cm. Cot- tonwoods and willows were well dispersedthroughout the area, as were Box-elders, but the latter tended to be somewhat more concentrated in Figure 1. Panoramicview of a portion of the Gila River Valley study area looking south:area of study includesboth the open sandyriverbottom in the foregroundand denselywooded areas in the background. 116 NEW MEXICO RIPARIAN COMMUNITY southernportions of the studyarea. Small standsof Emory Baccharis(Bac- charisemory/) in southernportions of the studyarea were replacedin the more northern plots by Seepwillow (Baccharisglutinosa). The dense understorywas also composedof fallen limbs, snags and the following plants: Chuchupate (Ligusticumporteri), Buffaloweed (Ambrosiatrifida), Sweet Four-O'Clock(Mirabilis !ongifiora), Skunk-bush (Rhus trilobata), and Western Virginia-Creeper (Parthenocissusinsefta). RESULTS Species Composition and Seasonal Occurrence. Each of the 112 bird speciesobserved during the studywas groupedinto one of four categories based on its seasonaloccurrence and breedingstatus (Table 1). The four categoriesare: (1) winter birds (30 species),those present during January and Februarybut not remainingto breed; (2) migrants(29 species),those presentduring months other than January and Februarybut not remainingto breed; (3) summer residents(24 species),those not membersof the former categoriesbut nestingor usingthe area extensivelyduring May and June; and (4) permanentresidents (29 species),birds generally present throughout the 6-month period. Of the 30 winter species,only the Red-raftedHawk, Brown Creeper, Ruby-crowned Kinglet, Yellow-rumped Warbler, Rufous-sidedTowhee, Song Sparrow, Lincoln'sSparrow, White-crowned Sparrow and Dark-eyed Junco were consistentlypresent. Others such as Green-wingedTeal, Nor- thern Pintail, Northern Harrier, Golden Eagle, Prairie Falcon, Bushtit, Figure2. Typicalview of Gila RiverValley woodedareas showing Goodding's Willows in the foregroundwith Box-eldersand Fremont'sCottonwoods in the background. 117 NEW MEXICO RIPARIAN COMMUNITY WesternBluebird and LoggerheadShrike were recordedwithin the area only once, althoughmost were presentin nearbyareas throughout much of the period.The one Ring-billedGull wasmerely a vagrantto the studyarea; the speciesis irregularin the lowerGila RiverValley. The singleHermit Thrush observedon the firstof Februarymay have winteredin nearbyhabitats. Very few of the 29 migrantspecies remained for an extendedperiod. Thoselingering included Wilson's Warbler, House Wren and ChippingSpar- row. Most of the remainingspecies were simplypresent one week and were gone by the next. Some, such as Broad-tailedHummingbird, Painted Redstart, Western Tanager and Green-tailedTowhee, were common in other, generallymore elevated,areas of the lower Gila RiverValley during latespring and summer.Say's Phoebe and ChihuahuanRaven were relative- ly commonin open habitatsadjacent to the studyarea but rarelyventured in- to the area itself.The Gray Flycatcheris a regularmigrant in thisarea of New Mexico,but it is moretypical of evergreenwoodlands (John Hubbard in litt.). The Winter Wren and American Redstart are uncommon in the area but nonethelessmay occur somewhatregularly in very low numbers. Mostof the 24 summerspecies were continuouslypresent after their initial arrival and all but three nestedwithin the study area. Turkey Vulturesand Cooper's Hawks occasionallyentered the study area but I obtained no evidenceof nesting.The singleAmerican Crow observedin late May and early June was probably a vagrant, since the specieswas uncommon throughoutthe lower Gila River Valley prior to 1975. It has, however, become more common in the Valley during the past decade (Dale Zimmermanin litt.). The extendedpresence of Cliff Swallowsresulted from the establishmentof a nestingcolony under the U.S. Highway 180 bridge, which formed the southernboundary of the study. Only six of the 29 permanentresidents were observedon eachof the 23 weeklysurveys. All buttwo, however,were observed on 50% or moreof the surveys.Despite the fact that Brown Towheesand WesternMeadowlarks only occasionallyentered the study area, I consideredthem residents becausethey were consistentlyseen and/or heardthroughout the studyin adjacentareas. The Great Blue Heron, Mallard,Common Merganser,Spot- ted Sandpiper, Ladder-backed Woodpecker, Common Raven, Brown Towhee and Western Meadowlark did not nest within the confines of the study area, but all presumablynested in adjacentareas. Common Ravens are, however, known to have nestedwithin the area duringother years (Dale Zimmerman in litt.). The remaining21 permanent residentsare known to have nestedwithin the studyarea duringthe courseof my study(Table 1). Seasonal Variability. Variability in avian density, species richness, equitability,and birdspecies diversity in eachhabitat type are shownin Table 2. Avian densitywithin riverbottomareas showed considerablemonthly variation;there was lessvariation within the wooded areas.Seasonal density patternswithin both riverbottomand wooded habitatsshowed expected seasonaltrends; i.e., numberswere lowest during winter, highestduring springmigration, and of intermediatemagnitude during the breedingseason. The numberof speciespresent throughout the entire6-month period in river- bottom areas remainedrelatively constant, whereas that for wooded areas showed expectedseasonal fluctuations. 118 NEW MEXICO RIPARIAN COMMUNITY 119 NEW MEXICO RIPARIAN COMMUNITY O o0 u'• • O o0 o0 120 NEW MEXICO RIPARIAN COMMUNITY 121 NEW MEXICO RIPARIAN COMMUNITY