Finches  — Family Fringillidae

Purple Finch Carpodacus purpureus

The Purple Finch has long been known as a fairly common resident of the coniferous woodlands of San Diego County’s mountains.  It still plays that role, but atlas observers also discovered an unexpected new one, as an uncommon and newly established resident of oak and riparian woodland at low elevations in northwestern San Diego County.  Yet another role, this one traditional, is as an irregular winter visitor all over the county’s coastal slope to woodland and dense chaparral with fruiting shrubs like toyon and California coffeeberry.  In this role, however, the Purple Finch seems to be on the decrease.  How changes in habitats and climate may be contributing to these seemingly contradictory changes in status is unclear.

Breeding distribution: The Purple Finch occurs in all of the larger stands of coniferous forest in San Diego County’s mountains, being more common in denser stands of bigcone Douglas fir, incense cedar, and canyon live oak than in open groves of Jeffrey or Coulter pine.  Seldom does one find more than a dozen birds in a morning even in prime habitat, but counts occasionally run as high as 25 from 0.25 to 1.5 miles south of Burned Rancheria Campground, Laguna Mountains (P23), 23 June 2000 (E. C. Hall, J. O. Zimmer), 30 on Palomar Mountain (E15) 3 July 2000 (C. R. Mahrdt, E. C. Hall), and 40 on Middle Peak, Cuyamaca Mountains (M20) 11 June 2000 (R. E. Webster).  Between Palomar and Volcan mountains, Purple Finches inhabit a narrow band of oak woodland with few or no conifers around Mesa Grande and in the gorge of the San Luis Rey River below Lake Henshaw.  In this band the species is generally uncommon, but W. E. Haas noted 11, including a mist-netted female with a brood patch, near the San Luis Rey Day Use Area (G16) 3 July 1999.

            One of the biggest surprises generated by the field work for this atlas was the discovery that the Purple Finch has colonized woodland of sycamores and coast live oaks at low elevations in northwestern San Diego County.  We noted the species during the breeding season on 76 occasions in 19 atlas squares where no elevation covered exceeds 2000 feet.  The area colonized extends from the Santa Margarita Mountains east through Fallbrook to Valley Center and Palomar Mountain.  In this zone, the birds were most concentrated along De Luz Creek (B6/C6), with ten, all singing males, in B6 on 10 July 2000 and seven, including six singing males, in C6 on 26 June 1999 (K. L. Weaver).  The record nearest the coast was of one singing male near O’Neill Lake (E6), elevation barely over 100 feet, 18 May 1999 (P. A. Ginsburg).  Breeding confirmations at low elevations were observations of nest building and feeding young at 840 feet in Fallbrook (C8) May 1998 (L. Ale), an occupied nest (female apparently incubating) at 1900 feet between Magee Creek and Castro Canyon (C12) 30 April 2000 (J. Determan), and nest building at 1600 feet in Marion Canyon (D12) 18 June 2001 (K. L. Weaver).  Two singing males along Temecula Creek about 2600 feet elevation (C16) 10 April 1999 (K. L. Weaver) suggest spread north as well as west from Palomar Mountain.

Nesting: Before field work for this atlas began in 1997, no specific data on Purple Finch nesting in San Diego County had been published, and no eggs of the species had been collected.  Of the seven nests reported since 1997, one was in a Jeffrey pine, one in a Coulter pine, one in a sycamore, and four in coast live oaks.  Most were rather high in the trees, at heights of 28 to 45 feet, but one along the San Luis Rey River at Prisoner Creek (G16) 4 June 2001 (W. E. Haas) was barely over 3 feet above the ground in a coast live oak.

Our observations suggest the birds begin nesting in late April (occupied nest in Lost Valley, D20, 28 April 2000, W. E. Haas; adult feeding young along Nate Harrison Grade, E13, 18 May 1997, C. Sankpill).  Most nesting activity concludes in mid July, but M. Dougan noted nest building as late as 18 July 2000 west of Dyche Valley (F15) and V. S. Moran noted fledglings as late as 2 August 2000 on North Peak, Cuyamaca Mountains (L20).

Migration: The Purple Finch’s colonizing low elevations as a breeding species complicates tracking the species’ dispersal.  For example, four in Couser Canyon (E10) 4 April 1998 (K. Aldern, M. Bache) could well have been pioneers rather than late winter visitors.  During the atlas period the latest observation far from a likely breeding location was of two in Oak Hill Cemetery, Escondido (I12), 13 April 1999 (C. Rideout), and that was the only one after 4 April.  Extreme dates in previous years range from 5 November (1972, five at Point Loma, S7, G. McCaskie) to 18 April (1975, two at Live Oak Park, D8, J. L. Dunn).

Winter: The Purple Finch is well known as an irruptive species in winter, but this irregularity was not well featured during the five-year atlas period.  Numbers were low in 1997–98 and 2001–02 in comparison to the three intervening winters, but there was no broad-scale invasion as in 1974–75 or 1987–88.  Coinciding with an invasion of Cassin’s Finch, the highest winter counts of the Purple were in 2000–01: 52 in Palomar Mountain State Park (E14) 28 December 2000 (J. D. Barr), 79 around the Palomar Observatory (D15) the same day (K. L. Weaver), and 55 at Oak Grove (C16) 24 February 2001 (K. L. Weaver).  Few observations during the atlas period were far from the breeding range, though past winter records are from as far from it as Point Loma.  There was only one winter report from the Anza–Borrego Desert 1997–2002, of two at Agua Caliente Springs (M26) 26 February 1999 (J. L. Coatsworth).  There are only about four earlier records from the desert floor, somewhat more from desert-edge locations like Culp Valley (H23) and Chariot Canyon (K21) (Massey 1998, ABDSP database).

Conservation: The Purple Finch’s invasion of northwestern San Diego County just happened to coincide with the initiation of field work for this atlas.  In spite of 20 years of field experience in the Fallbrook area, in areas covered well for the atlas, Kenneth L. Weaver had not found Purple Finches summering there before 1998.  At low elevations in northwestern San Diego County, both the total number and the number reported per hour increased every breeding season of the five-year term.  The reason for this spread is still a mystery.  The avocado orchards widespread in this area give it a more forested aspect than it would have naturally, but the Purple Finches are using native woodland almost entirely.

            In contrast to this trend during the breeding season, the Purple Finch seems to be on the decline as a winter visitor.  It has not invaded San Diego County on any large scale since 1987–1988.  It was noted on all 11 of the San Diego Christmas bird counts from 1965 to 1975, but on the 11 counts from 1991 to 2002 it was noted on only two, in 1992 and 2002.  Trends on the Rancho Santa Fe, Oceanside, and Escondido counts are similar, if not quite so stark.  Could climatic warming be obviating the need for this facultative migrant to move south?

Taxonomy: Only the western subspecies of the Purple Finch, C. p. californicus Baird, 1858, is known from San Diego County, though there are a few sight records and one specimen of the more eastern C. p. purpureus (Gmelin, 1789) from elsewhere in southern California (Patten et al. 2003).

Cassin’s Finch Carpodacus cassinii

Cassin’s Finch is another of those high-mountain birds whose breeding range skips over San Diego County’s comparatively low mountains.  It breeds in the San Jacinto and Santa Rosa Mountains to the north and in the Sierra San Pedro Mártir to the south, but in San Diego County there is only a single summer record.  Some winters pass with no records of Cassin’s Finch either—the species is quite irregular.  On rare occasions, though, Cassin’s Finches invade, and flocks of dozens can be seen, mainly in the mountains.  One of these incursions took place in the winter of 2000–01, because of the atlas the best documented Cassin’s Finch invasion in San Diego County history.

Winter: In most winters, Cassin’s Finch is a rare visitor to San Diego County’s mountains, especially the Laguna Mountains, which offer more of the open dry pine forest that resembles the species’ breeding habitat.  In some years there are none, in some substantial incursions.  The atlas’ five-year term captured the full range of this variation.  The annual total was 0 in 1997–98, 15 in 1998–99, 0 in 1999–2000, 503 in 2000–01, and 4 in 2001–02.  The invasion of 2000–01 may have been the largest in San Diego County history, rivaled only by that of 1975–76, when B. Cord counted 103 in the Laguna Mountains 26 January.  In 2000–01, numbers in pine-dominated woodland ranged up to 45 around Crouch Valley (P22) 7 January (P. Unitt) and 38 in Lost Valley near Shingle Spring (D21) 23 December (L. J. Hargrove).  The birds spread into oak and riparian woodland as well, however, sometimes in numbers just as large: 48 in Thing Valley (Q24) 7 January (J. R. Barth) and 36 in the Manzanita Indian Reservation (R25) 10 February (K. J. Winter, A. Mauro).  Few birds strayed from the mountains; the only records below 2500 feet elevation were of one near Bonsall (F9) from December to 18 February (J. Evans) and eight at Fernbook (M14) 10 February (B. Hendricks).  Lowland records were rare during past incursions as well.

Migration: As expected in an irruptive species, Cassin’s Finch does not follow a regular migration schedule.  With one exception, dates for San Diego County extend from 3 October (1987, three at Point Loma, S7, J. L. Dunn, AB 42:139, 1988) to 17 May (1997, two at Pine Valley, P21, J. K. Wilson).  The winter preceding initiation of field work for this atlas, 1996–97, was also an invasion year for Cassin’s Finch, as for several other mountain birds, so spring 1997 generated several late records, most notably of one in the Clairemont area of San Diego (P8) 10 April 1997 (C. G. Edwards).  After the incursion of 2000–01 the latest sightings were from Panawatt Spring, Los Coyotes Indian Reservation, of nine on 5 May and one on 6 May (J. Hargrove).

The first summer record of Cassin’s Finch for San Diego County was of a single singing male on Birch Hill, Palomar Mountain (E15), 3 July 2000 (C. R. Mahrdt, E. C. Hall).  Curiously, this sighting followed a winter in which no Cassin’s Finches reached San Diego County.


Conservation: No strong trends in Cassin’s Finch numbers have been reported, but these numbers fluctuate at some places even in the species’ core range (Hahn 1996), making long-term changes difficult to detect.

House Finch Carpodacus mexicanus

The House Finch is the most abundant bird in San Diego County, and the field work for this atlas only supported this statement further.  A year-round resident, the House Finch occupies all terrestrial habitats, from the coastal strand to montane coniferous woodland and sparse desert scrub.  Most of man’s modifications of the environment favor the House Finch: buildings and bridges offer nest sites, and disturbed weedy areas offer foraging habitat.  The House Finch is a leading patron of bird feeders—of both seeds and sugar water.

Breeding distribution: The House Finch is San Diego County’s most widespread breeding bird.  It is most abundant in the coastal lowland (up to 335 in lower Los Peñasquitos Canyon, N8, 2 July 2000, D. R. Grine) and around oases and developed areas in the Anza–Borrego Desert (up to 300 around Carrizo Palms and Indian Hill, R28, 6 May 1998, J. O. Zimmer).  But it occurs throughout the county, being uncommon only in montane coniferous forest and in sparsely vegetated desert miles from water.  During the breeding season, the House Finch was missed from only one well-covered atlas square, D14 on Palomar Mountain.

Nesting: With close to 1500 records, the House Finch was the bird we confirmed nesting most frequently.  Many confirmations are of more than one nest; the House Finch sometimes nests colonially where good nest sites are scarce, as at desert oases with California fan palms.  The House Finch has been said to nest “anywhere” (Adams 1899), but scanning atlas participants’ descriptions of nest sites reveals quickly that the birds prefer placing the nest on as solid a surface as possible, accounting for the abundance of nests on buildings.  A covered situation is also desirable, leading the birds to nest in drain holes, under roof tiles, inside the hollow arms of power poles, behind slabs of buckled cottonwood bark, and in old nests of Black Phoebes, Cliff Swallows, and Hooded Orioles.  If the birds nest in trees, the trees are usually ones that offer dense screening foliage, especially cultivated trees like orange and Italian cypress.  In treeless chaparral, the inflorescences of Yucca whipplei appear to be the preferred nest site.  With 26 of 145 records, however, the most frequently described nest sites were cholla and prickly pear cacti, from the Anza–Borrego Desert to the cactus garden in Balboa Park.  House Finches nest colonially in stands of the teddy-bear or jumping cholla.  They evidently recognize the ability of cactus spines to deter predators.

            House Finches may begin laying in San Diego County as early as the end of February, as attested by observations of nest building as early as 19 February, occupied nests on 27 February, nests with nestlings on 15 March, and fledglings on 21 March.  They continue into July, with young still in the nest as late as 29 July.  These dates extend the season known from 33 egg sets collected in San Diego County, 30 March–29 June, but agree with the 28 February–7 August spread of egg dates from California as a whole (R. S. Woods in Austin 1968).

Migration: After breeding, House Finches gather into large flocks and move nomadically, searching for good foraging.  But they do not engage in regular migration, and their distribution’s uniformity masks seasonal shifts, if any.

Winter: The House Finch’s distribution in San Diego County in winter differs little from that in the breeding season, being concentrated at low elevations, sparser in the mountains and in regions of extensive unbroken chaparral.  Flocks may be larger in winter than in the breeding season (up to 690 in and near June Wash, M27, 10 January 1998, R. Thériault).  Numbers in the Anza–Borrego Desert varied positively with rainfall, increasing in 1998, decreasing thereafter.  But the magnitude of the fluctuation was less than with other seed-eating birds.  The total in 2001–02 was still 43% of that in 1998–99.

Conservation: The House Finch has been abundant in San Diego County since the earliest naturalists reported on the county’s birds.  There is no evidence for significant recent change in the species’ numbers.  Nevertheless, buildings, disturbed openings, irrigation, and bird feeders are all continuing changes to the environment that favor House Finches in man-modified habitats more than in natural ones.

Taxonomy: Carpodacus m. frontalis, widespread on the mainland of western North America, is the only subspecies of the House Finch in San Diego County.

Red Crossbill Loxia curvirostra

Many members of the finch family migrate irregularly, as demanded by the irregularity of their food supply.  The crossbills, with their unique specialization and dependence on conifer seeds, take this irregularity to an extreme.  With only small areas forested in conifers, San Diego County cannot support a population of the Red Crossbill, so the species is a sporadic visitor here, though it has attempted nesting in the county at least once.  Studies of morphology, genetics, and voice suggest the Red Crossbill, with its great variation across North America in size and bill shape, may consist of multiple cryptic species.

Winter: The winters of 1966–67, 1984–85, and 1996–97 saw the biggest incursions of the Red Crossbill known in the history of San Diego County.  Thus the 5-year atlas period 1997–2002 began with the winding down of an incursion in which small flocks were seen through much of the county such as 10 at the Vineyard Golf Course, Escondido (K11), 17 February 1997 (E. C. Hall).  The Oceanside Christmas bird count yielded the maximum of 24 on 29 December 1996, and up to 10 occurred even at Borrego Springs that winter (M. L. Gabel, NASFN 51:119, 1997).  The irruption of 1966–67 yielded up to 150 at Point Loma in November 1966 (AB 21:80, 1967), that of 1984–85, up to 25 there in March 1985 (AB 39:351, 1985).

            The remainder of the atlas period was more typical.  In the five winters following 1996–97 no crossbills were noted in San Diego County at all, except in 2000–01, which yielded four occurrences, three of them in conifer-wooded mountains, with a maximum of eight birds in Lower Doane Valley, Palomar Mountain State Park (D14), 22 December 2000 (P. D. Jorgensen).

Migration: The movements of the Red Crossbill are famously unpredictable, with little correspondence to the calendar.  Nevertheless, in San Diego County, the birds seldom if ever arrive before late October.  After larger invasions they may remain quite late in the spring, being recorded at Point Loma as late as 4 June in 1967 (AFN 21:542, 1967) and 3 June in 1985 (AB 39:351, 1985).  In 1997 the latest report was of two near Descanso (O19) 13 May 1997 (R. A. Hamilton).

Breeding distribution: The crossbill’s only breeding activity noted in San Diego County was at Point Loma in late March 1967, when some birds were paired, carrying nest material, and engaging in apparent courtship feeding.  No fledglings were seen subsequently.  The species occurs rarely in San Diego County’s mountain forests in summer, and not only in summers following irruptions.  The maximum number in such a role was up to 15 in the Laguna Mountains in late July 1993 (G. L. Rogers, P. A. Ginsburg, AB 47:1152, 1993).  From 1997 to 2001 the only such reports were from Middle and Cuyamaca peaks (M20), with one on 19 May 1998 and one or two 23–24 June 2001 (S. Peterson, D. Holway).


Conservation: The crossbills breeding in the southwestern quadrant of the contiguous United States have bills adapted to feed on the seeds of pines.  Prolonged drought, as seen at the beginning of the 21st century, brings death to large numbers of pines by way of bark-beetle attack or forest fire.  The Red Crossbill, being so specialized, is thus one of the most likely of the coniferous forest birds to suffer the effect of a drying climate.


Taxonomy: The Red Crossbill presents a taxonomic conundrum unique in North America.  Various populations differ to a greater or lesser degree in size, bill shape, learned calls, genetic makeup, and specialization for certain conifers.  The breeding ranges of these populations overlap to varying and still uncertain degrees.  In spite of considerable study (Groth 1993, Benkman 1993), the proper interpretation and categorization of this variation is far from clear (DeBenedictis 1995).  All but one of the eight specimens from San Diego County in the San Diego Natural History Museum, including three from the irruption of 1996–97, are of the medium-large size prevalent in the pine forests of western North America.  That is, they are of size class III of Phillips (in Monson and Phillips 1981) or call types 2, 5, or 7 of Groth (1993), for which the oldest name is L. pusilla Gloger, 1834, or L. c. bendirei Ridgway, 1884.  The exception is SDNHM 873, collected at Campo (U23) 6 March 1877.  It has the very large bill identifying it as L. c. stricklandi Ridgway, 1885, which breeds in the mountains of mainland Mexico, north to southeastern Arizona, and in the mountains of northern Baja California.

Pine Siskin Carduelis pinus

In its breeding range, the Pine Siskin is indeed a bird of pines, as well as other conifers.  But in San Diego County, where it occurs primarily as a winter visitor, the siskin is just as likely to be seen in riparian woodland, where it feeds on the catkins of willows and alders.  Its abundance varies enormously from year to year.  In some winters flocks are common; in other years there are hardly any.  Though the Pine Siskin breeds regularly south to the San Jacinto Mountains, Riverside County, in San Diego County’s mountains there are only five summer records, one of a juvenile with a flock of adults.

Winter: In general, the Pine Siskin is most numerous in the mountains, becoming less so at lower elevations toward the coast.  Nevertheless, variation from year to year is much greater than that from site to site.  There was a substantial irruption in 1996–97, so that the first spring of the five-year atlas period yielded fair numbers of siskins, but in the next five winters the numbers were average to low.  In the winter of 2001–02 the species was almost absent, with only three individuals reported.  By far the largest count made during the atlas period was of 235 around North Peak, Cuyamaca Mountains (L20) 6 February 1999 (R. Breisch), but this could have been equaled easily in years of big invasions, such as 1975–76, 1981–82, 1984–85, 1987–88, and 1992–93.  The highest number on any of the county’s Christmas bird counts was 553 on the Lake Henshaw count 29 December 1981.

            At lower elevations, during the atlas period, no Christmas bird count yielded more than 29, and high numbers in one atlas square were of 49 at Oak Hill Cemetery (I12) 16 March 2001 (C. Rideout) and 30 at Guajome Lake (G7) 8 December 2000 (G. C. Hazard).  But totals of lowland Christmas bird counts during irruptions are as high as 182 in Oceanside 27 December 1981, and G. McCaskie noted 800 in Presidio Park (R8) 17 November 1963.

            No siskins were found in the Anza–Borrego Desert during the atlas period, but small numbers reach there during major irruptions, with up to 20 at Lower Willows (D23) 20 December 1987 (A. G. Morley) and 29 in the north Borrego Valley (F24) 2 January 1993 (Christmas bird count).

Migration: Like other erratic winter visitors, the Pine Siskin has no regular migration schedule.  Early November to mid April is its principal season, but it has been noted in fall as early as 9 September (1963, Tijuana River valley, G. McCaskie; 1992, Borrego Palm Canyon, F23, R. Thériault).  During the atlas period, our latest spring date was 10 May (1999, 10 in Lost Valley, D20, J. M. and B. Hargrove), except for a remarkably late straggler, the latest ever in southern California’s lowlands, at the Dairy Mart pond, Tijuana River valley (V11), 6 June 1999 (G. McCaskie).

Breeding distribution: No nests of the Pine Siskin have ever been found in San Diego County, and there are only five records in the county’s mountains from late spring and summer.  One of these, however, was of a juvenile, probably raised locally, with four adults on Middle Peak, Cuyamaca Mountains (M20) 19 July 1987 (R. E. Webster, AB 41:1489, 1987).  Of the other records, two are from Palomar Mountain (one on 19 July 1966, AFN 20:600, 1966; one at Bailey’s Meadow (E14) 28 May 1999 (C. R. Mahrdt, E. C. Hall, J. O. Zimmer), one from the Cuyamaca Mountains (one on 9 July 1967, G. McCaskie), and one from the Laguna Mountains (one near Laguna Campground, O23, 13 June 1998, C. G. Edwards).


Conservation: Pine Siskins feed on the seeds of some common ornamental trees and so take advantage of areas that once had little to offer them.  The white alder, a native riparian tree used frequently in landscaping, is one of the siskins’ favorite seed sources.  Gander (1929) commented on their feeding on eucalyptus seeds in Balboa Park.

Taxonomy: Pine Siskins in San Diego County, like those throughout the United States, are of the small subspecies C. p. pinus (Wilson, 1810).

Lesser Goldfinch Carduelis psaltria

The Lesser Goldfinch is one of San Diego County’s most widespread birds.  It is a year-round resident and a habitat generalist, taking advantage of any weedy area for foraging.  Nesting birds need shrubs or trees for nest sites and water for drinking within an easy commute of the foraging habitat. Our most interesting discovery about the Lesser Goldfinch was its response to El Niño: following the wet winter of 1997–98, the birds spread over the Anza–Borrego Desert, some nesting far from the oases to which they are usually restricted there.  When drought returned, the spread proved as ephemeral as the bloom of desert wildflowers.

Breeding distribution: The Lesser Goldfinch is almost ubiquitous in San Diego County.  It is especially abundant in the inland valleys, where as many as 150 have been counted during a day in the breeding season, as near Monserate Mountain (D9) 19 May 1999 (E. C. Hall), in lower Boden Canyon (J14) 1 June 1999 (C. R. Mahrdt), and near Ramona (K15) 18 June 1999 (M. and B. McIntosh).  Because male Lesser Goldfinches defend only a small area around the nest, the species may breed semicolonially; Weaver (1992) recorded 11 territories in one 11.7-acre study plot along the Santa Margarita River north of Fallbrook (C8).  Large numbers occur also in the mountains (up to 125 in Matagual Valley, H19, 18 June 2000, S. E. Smith) and in the Borrego Valley (up to 65 in the north end of the valley, E24, 12 March 2000, P. D. Ache).  The Lesser Goldfinch occurs in heavily urbanized areas though less commonly than where development is sparse or none; we did not find it in the breeding season in the two most completely developed atlas squares, R7 (Ocean Beach) and S8 (North Island).

Only in the remoter waterless reaches of the Anza–Borrego Desert do we see large holes in the distribution of this species that must drink regularly.  Even there the Lesser Goldfinch is surprisingly widespread.  The lack of nesting confirmations from many desert areas, however, suggests that many of the goldfinches there during the breeding season are not nesting.  It was primarily in the wet year 1998 that Lesser Goldfinches spread over the Anza–Borrego Desert and were confirmed nesting in washes far from oases.  Only two nestings in such dry habitats were noted in the other four years of the study (nest building on the northeast slope of the Santa Rosa Mountains, C28, 2 May 2000, R. Thériault; occupied nest along Fish Creek Wash at Split Mountain, L29, 11 April 2000, J. R. Barth).

Nesting: The Lesser Goldfinch usually builds its nest in a dense-foliaged shrub or tree, generally placing the nest toward the tip of a limb in a situation where it will be shaded at least part of the day (Linsdale 1957, Dawson 1923).  Of 29 nests atlas observers described, nine were in coast live oaks; others were in Engelmann oak, sycamore, willow, cottonwood, pine, ash, eucalyptus, avocado, mulefat, laurel sumac, coyote brush, and Chinese weeping elm.  Clearly, the Lesser Goldfinch is a generalist when it comes to nesting.

            We observed quite a bit of nesting activity beginning in mid March, especially in the wet year 1998.  An occupied nest as early as 13 March 1998 (near Lake Murray, Q11, N. Osborn), an adult feeding a nestling as early as 21 March 1998 (Lower Willows, D23, B. Peterson), and a fledging as early as 7 April 1998 (west end of Batiquitos Lagoon, J6, M. Baumgartel) all imply egg laying in the second week of March, and therefore earlier than the 6 April attested by collected egg sets from San Diego County or 22 March from all of California.  The season winds down in late July and early August (nest with nestlings in Poway, L12, as late as 16 August 1999, K. J. Winter).  But the Lesser Goldfinch is also known to nest in the fall, in San Diego County (Sharp 1908, Carpenter 1919) as elsewhere.  During the atlas period we noted fall nesting twice, with an adult male feeding a fledgling in the Rolando neighborhood of San Diego (R11) 3 December 1998 (F. Shaw) and a nestling found fallen out of a nest on the campus of San Diego State University (Q11) 20 October 2000 (SDNHM 50489).

Migration: The Lesser Goldfinch does not undertake any regular migration in San Diego County.  Nonbreeding birds, however, flock and wander.  In open desert scrub, where few if any Lesser Goldfinches nest, small flocks have been seen as late as 7 April (1998, eight at Yaqui Meadows, H24, P. K. Nelson).

Winter: The distribution of the Lesser Goldfinch in San Diego County in winter differs little from that during the breeding season.  The species’ flocking in winter leads occasionally to counts as high as 293 east of Chula Vista (U12) 19 December 1998 (W. E. Haas) and 207 at Sentenac Ciénaga (J23) 16 February 1998 (R. Thériault).  There is some shifting downslope from the highest elevations; we did not find the species in winter near the summits of Hot Springs Mountain (E20) and the Laguna Mountains (O23), where it occurs in summer.  The Lesser Goldfinch is scattered over the Anza–Borrego Desert in winter but more sparsely than in spring.  Probably the difference is due to the birds’ being able to take greater advantage of the desert when spring growth follows the rains.

Conservation: The Lesser Goldfinch benefits from many of man’s alterations of the southern California environment.  It feeds heavily on introduced weeds like the common sow-thistle and yellow star-thistle (Beal 1910, Linsdale 1957).  Rural ranches offer disturbed open areas for foraging, shade trees for nesting, and water sources for drinking lacking in undeveloped native chaparral.  The widespread use of the native white alder in landscaping puts another of the goldfinch’s favored seed sources in many places where it does not grow naturally.  These factors likely outweigh negative ones like overgrazing, overpumping of groundwater, and landscaping so intensive it eliminates weedy edges.

Taxonomy: The subspecies of the Lesser Goldfinch resident in San Diego County is the Green-backed Goldfinch, usually called C. p. hesperophilus (Oberholser, 1903).  There are at least two sight records of black-backed males elsewhere in southern California (Patten et al. 2003), apparently nominate C. p. psaltria (Say, 1823).  Such birds could be escapees from captivity, however, rather than vagrants or variants.  The black-backed subspecies is widespread in mainland Mexico, the source of many escaped cage birds.

Lawrence’s Goldfinch Carduelis lawrencei

Though found in San Diego County year round, Lawrence's Goldfinch is notoriously nomadic, exploiting food sources that are abundant but ephemeral.  In summer, these are seeds of wildflowers of the family Boraginaceae, especially the fiddleneck, which grows in scattered meadows, often in disturbed soil.  In winter, the goldfinches shift to gleaning seeds from the dried flowers of chamise, the shrub dominating San Diego County's chaparral.  Nevertheless, the bird’s irregularity in winter is even greater than in summer.  Lawrence's Goldfinch responded to El Niño rains of 1997–1998, capitalizing on the bloom of desert wildflowers.  Lawrence's, like the other goldfinches, must drink regularly, making long flights to water sources scattered through its arid habitat.

Breeding distribution: In San Diego County, Lawrence’s Goldfinches are concentrated in the mountains.  For suitable habitat, a meadow, a creek, and a grove of oaks is the ideal combination.  In such areas, one may occasionally encounter concentrations as high as 50 in Chariot Canyon (L21) 16 June 1999 (J. K. Wilson), 56 on East Mesa, Cuyamaca Mountains (N21), 8 July 1999 (D. C. Seals), and 60 along upper Pine Valley Creek (O21) 4 July 1997 (P. Unitt).  Even in this core range, however, the species is patchy and irregular, often lacking.  At lower elevations, toward the coast, it becomes ever more so, though numbers as high as 40 in Pauma Valley (E12) 13 May 2000 (P. Unitt) and 20 along the San Diego River in Santee (P13) 24 June 1997 (D. C. Seals) may still occur in the middle of the breeding season in the inland valleys.  Along the coast, Lawrence’s Goldfinch is rare, especially as a breeding bird.  Susan E. Smith noted two coastal nestings, one barely more than 1 mile from the beach near the intersection of Interstate 5 and Del Mar Heights Road (N7), where apparently three pairs nested in Canary Island pines planted around an apartment complex; nestlings were visible in one nest 8 May 1999.  The other was 0.3 mile from the beach near the University of California (O7), where a male fed a female on a nest in a eucalyptus 9 May 2000.

            In the Anza–Borrego Desert Lawrence’s Goldfinch’s irregularity is accentuated further.  Of the atlas period’s five years, the wet 1998 saw by far the greatest numbers; in other years there were few to none.  Butterfield Ranch in Mason Valley (M23) is the only desert location where Lawrence’s Goldfinch is even moderately regular.  All our desert confirmations of Lawrence’s Goldfinch nesting were in 1998, except at the desert-edge locations of Earthquake Valley (K23), Mason Valley (M23), and In-Ko-Pah (T29), where the birds nested in 2001 as well.  Lawrence’s Goldfinch nesting in the Anza–Borrego Desert was unknown before 1998 (Massey 1998).

Nesting: Lawrence’s Goldfinch’s strategy of opportunism results in its being the most colonial of San Diego County’s finches.  Several pairs may nest simultaneously in a small grove of trees.  In addition to coast live and Engelmann oaks, sycamores, and pines, reported nest sites included deodar cedar and Italian cypress.  Evidently the dense screening foliage of these exotic trees makes them especially attractive.  An atypical apparent nest site was in a rotted out cavity in a horizontal willow branch, from which I flushed a pair near Moretti’s Junction (H18) 12 May 2001.

            The timing of Lawrence’s Goldfinch nesting varies with the rains and food supply as well.  Most nesting takes place from April to July, as shown by the 7 April–9 July spread of 12 collected egg sets.  After the wet winter of 1997–1998, however, Lawrence’s Goldfinches began nesting exceptionally early, as attested by an occupied nest in Culp Valley (H23) 23 March 1998 (M. L. Gabel), a nest with nestlings at Stelzer County Park (O14) 4 April 1998 (M. B. Mulrooney), and a fledgling at the Roadrunner Club, Borrego Springs (F24), 25 March 1998 (M. L. Gabel).  The last implies eggs laid at the end of February, apparently the earliest the species’ nesting has ever been reported.  On the late side, Lawrence’s Goldfinches were still building nests in Pine Valley (P21) 5 July 1997 (J. K. Wilson) and at Lake Morena (T21) the same day (R. and S. L. Breisch).

Migration: Lawrence’s Goldfinch is an irregular, partial migrant, a variable fraction of the population moving east of the Colorado River for the winter.  In San Diego County, migrants are most notable in March and April, when flocks are seen occasionally and the birds show up more frequently in the Anza–Borrego Desert and along the coast.  Such flocks may be as large as 60 at Lower Otay Lake (U14) 2 April 2000 (S. Buchanan), 350 in the Borrego Valley’s mesquite bosque (G25) 10 April 1995 (Massey 1998), and “hundreds” in Santee (P12) 24 March 2001 (W. McCausland).  In a nomadic, opportunistically nesting species like Lawrence’s Goldfinch the notion of a migration schedule is fuzzy.  Especially in 1998 we noted occasional pairs or flocks in habitat atypical for breeding in late spring and summer: pair near Sunset Mountain (J26) 19 April 1998 (F. L. Unmack, R. Orr); pair at Azalea Park, East San Diego (R10) 28 May 1998 (J. A. Dietrick); 10 on the east side of Chula Vista (U12) 24 June 1998 (T. W. Dorman).

Winter: In winter opportunism rules Lawrence’s Goldfinch’s behavior and distribution even more strongly.  The birds descend on a source of seeds, exhaust it, and move on.  They are concentrated at this season between 1500 and 4500 feet elevation and most widespread in south-central San Diego County, where the chamise on which they feed covers vast areas.  During the atlas period the largest winter flocks noted were 236 east of Corte Madera Valley (R21) 20 February 1999 (W. E. Haas), 200 near Corral Canyon (S20) 30 January 1999 (D. C. Seals), and 150 near Swan Lane (F18) 29 December 1997 (G. L. Rogers).  We found the species rare along the coast in winter, too, with a maximum of six in Torrey Pines State Reserve 23 December 2001 (S. Walens).  The irregularity of Lawrence’s Goldfinch is exemplified by the results of the Escondido Christmas bird count: species recorded on 10 of 17 counts, mean 22, standard deviation 71, maximum 224.

Conservation: Lawrence’s Goldfinch’s irregularity means that much more data are needed for trends to be identified in this species than in most others.  None are evident yet, though the flocks of hundreds noted occasionally in the Tijuana River valley in the 1960s did not recur during the atlas period. The planting of exotic conifers and the irrigation of rural ranches enhance Lawrence’s Goldfinch habitat, but intensive development, obliterating meadows and weedy fields, eliminates it.  Another likely negative factor is the proliferation of foreign weeds that have so largely displaced the native wildflowers that provide the staples of the goldfinch’s summer diet.

American Goldfinch Carduelis tristis

The American is the most widespread of the three goldfinches in North America but the most restricted in San Diego County, which is virtually at the southern tip of the bird’s range. As implied by the names of the California subspecies, the Willow Goldfinch, C. t. salicamans, the American is a riparian species in this region. It is a common resident in northwestern San Diego County but common in only a few places elsewhere.  During the nonbreeding season flocks readily depart riparian woodland to forage on seeds of plants like sunflowers and the great marsh evening primrose.

Breeding distribution: As a breeding bird, the American Goldfinch is closely tied in San Diego County to riparian woodland.  By far the largest numbers are found along the Santa Margarita River in Camp Pendleton, where daily counts in a single atlas square ran as high as 120 near the confluence of De Luz Creek (D6) 20 July 1999 (D. C. Seals) and 100 between Rifle Range Road and Ysidora Basin (F5) 16 May, 13 June, and 11 July 1998 (R. E. Fischer).  The species is common elsewhere in the lowlands of northwestern San Diego County (24 along the Santa Margarita River north of Fallbrook, C8, 16 July 1999, W. Pray; 50 along the San Luis Rey River near Gird Road, E8, 26 April 1999, P. A. Ginsburg) and found along small creeks as well as the Santa Margarita and San Luis Rey rivers.  Farther south the American Goldfinch becomes more and more localized to the major riparian corridors but is still common in a few places, especially along the Sweetwater and Tijuana rivers (up to 50 along the Tijuana River east of Hollister Street, W11, 27 June 1998, P. Unitt).  Above 1500 feet elevation the American Goldfinch is sparse and possibly irregular at any particular site.  During the atlas period, the only place in this zone where the species was found repeatedly was along Buena Vista Creek in Warner Valley (G18/G19; up to 20 near Warner’s Ranch, G19, 17 June 2000, J. D. Barr).  The species is absent from southern San Diego County east of Descanso and Potrero.  On the desert slope the American Goldfinch occurs in San Felipe Valley but not consistently.  Near Paroli Spring (I21) there were 10, including multiple singing males and pairs, 26 April 1999, but on 13 June 1999 there was only a single bird (J. O. Zimmer).  Farther downstream, near Scissors Crossing (J22), there were two on 14 May 1998 (E. C. Hall), but intensive coverage of this area May–July 2002 did not reveal any.

Nesting: Unsurprisingly for a species so closely linked to riparian woodland, the American Goldfinch nests primarily in willow trees.  Of 18 collected egg sets whose site was described, 15 were in willows.  Other known sites in San Diego County are cottonwood, cypress, orange trees, and goldenrod.  Thus the American Goldfinch may nest high or low.  In the eastern United States, the American Goldfinch is famed for nesting late in the summer, but in southern California its nesting season is typical of that of other riparian woodland birds.  Dates of 33 egg sets collected 1889–1940 range from 21 April to 6 July, and the nesting activity we observed during the atlas period was largely consistent with this.  The exception was at the upper end of Sweetwater Reservoir (S13) in 1997, when P. Famolaro noted nest building as early as 26 March, an occupied nest as early as 10 April, and fledglings as early as 30 April.

Migration: The American Goldfinch is not as nomadic as some members of its family, but occasional birds show up in nonbreeding habitat even in the middle of the breeding season (e.g., one at the Chula Vista Nature Center, U11, 30 May 1999, B. C. Moore).  The species is a very rare wanderer to the Anza–Borrego Desert, recorded from 19 September (1991, three in Culp Valley, H22, M. L. Gabel) to 22 April (1957, two at Yaqui Well, ABDSP database).  The only spring migrants reported from the desert 1997–2001 were three at Mountain Palm Springs (O27) 22 March 1998 (S. V. Fukuman) and one in Blair Valley (L24) 13 April 1998 (G. P. Sanders).  Note that these records were in the wet spring of 1998 that induced the Lesser and Lawrence’s Goldfinches to spread into the desert.

Winter: At this season the American Goldfinch loosens its attachment to extensive riparian woodland and takes advantage of weedy areas along minor creeks and even alder and Liquidambar trees planted in parks and residential areas.  The pattern of winter records, though, suggests that few birds move more than a few miles.  High counts in winter were of 60 at the east end of Lake Hodges (K11) 31 December 1997 (E. C. Hall), 50 in Reidy Canyon (H10) 12 December 1999 (D. and D. Bylin), and 50 near the upper end of Sweetwater Reservoir (S13) 31 January 2002 (T. W. Dorman).  A few American Goldfinches were reported in southeastern San Diego County, east to Live Oak Springs (S25; three on 21 January 2001, W. Dallas) and as high as 5200 feet elevation along La Posta Creek (P24; two on 19 December 2001, E. C. Hall, J. O. Zimmer).  The latter is the only winter record above 4000 feet elevation.  During the atlas period we noted the species twice in the Anza–Borrego Desert, one in the northern Borrego Valley (E24) 17 February 1999 (J. E. Fitch), eight at Agua Caliente Springs (M26) 1 December 1997 (E. C. Hall).  The American Goldfinch has been recorded on only three of 18 Anza–Borrego Christmas bird counts 1984–2001, maximum eight on 28 December 1986.  The winter irregularity of so many cardueline finches is not typical of the American Goldfinch in southern California, but the 262 reported on the Lake Henshaw Christmas bird count 3 January 1987 far exceeded the count’s second highest total of 18.

Conservation: The range of the American Goldfinch may have retracted over the 20th century in Baja California.  Currently, it is only a rare visitor south of the Tijuana River, despite its abundance along this river on the U.S. side of the border (Erickson et al. 2001).  Within San Diego County, although much riparian woodland has been lost, the American Goldfinch has sustained its numbers in the habitat that remains.  Cowbird trapping may benefit the American Goldfinch, as the goldfinch is a frequent victim of the cowbird in California (Friedmann 1963).  With its vegetarian diet, however, the goldfinch is not a good host for the cowbird.  Cowbird parasitism may reduce the goldfinch’s nest success significantly even if it fails to increase the cowbird’s.

Taxonomy: The only subspecies of the American Goldfinch known from southern California is the one breeding locally, the Willow Goldfinch, S. t. salicamans Grinnell, 1897.

Evening Grosbeak Coccothraustes vespertinus

A denizen of boreal forests, the Evening Grosbeak breeds in the Sierra Nevada but not in southern California.  In San Diego County it is a rare and sporadic visitor, mainly to the mountains in winter.

Winter: Most reports of the Evening Grosbeak in San Diego County are from Palomar Mountain (up to 40 on 9 February 1987, K. L. Weaver, AB 41:332, 1987), the Cuyamaca Mountains (up to eight on 11 November 1997, P. A. Ginsburg, NASFN 51:119, 1997), and the Laguna Mountains (up to 20 from 2 to 18 January 1987, B. McCausland, AB 41:332, 1987).  There are only 12 records below 1500 feet elevation, one as far southwest as the Tijuana River valley (20–22 October 1966, AFN 21:80, 1967).

            Like many members of the finch family, the Evening Grosbeak is irregular in its movements.  Years may pass with no reports for San Diego County, then, as in 1955–56, 1972–73, 1984–85, or 1986–87, there is a small irruption, often coinciding with irruptions of other finches.  During the atlas period, there were no records of the grosbeak between 1997, when an irruption the preceding winter yielded a spring straggler in Noble Canyon, Laguna Mountains (O22), 15 May (R. A. Hamilton, FN 51:927, 1997), and 2001, when at least five were at Paso Picacho Campground, Cuyamaca Rancho State Park (M20), 6 January–4 February (G. C. Hazard, M. B. Mulrooney).


Migration: San Diego County records of the Evening Grosbeak extend from 9 October (1972, one at Palomar Mountain, AB 27:125, 1973) to 13 February (1985, two at San Diego, R. E. Webster, AB 39:211, 1985) and from 26 April (2001, one at Paso Picacho Campground, R. Thériault) to 25 May (1971, three at Palomar Mountain, AB 25:804, 1971).  The eight spring records include the only desert or desert-edge records, from Yaqui Well (I24) 16 May 1964 (AFN 18:488, 1964) and Jacumba (U28) 30 April 1978 (AB 32:1057, 1978).

Taxonomy: In lack of a specimen, the subspecies of the Evening Grosbeak reaching San Diego County is not definitely known.  But all specimens from elsewhere in California, migrants as well as breeding birds, have been ascribed to C. v. brooksi (Grinnell, 1917), which breeds from British Columbia south through the western United States to central Arizona.  Shorter-billed nominate C. v. vespertinus (Cooper, 1825), breeding in Canada, has reached as far as Arizona, however, as attested by two specimens collected during the invasion of 1955–56 (Phillips et al. 1964).

Geography 583
San Diego State University