The sapsuckers showed seasonal preferences among the species they were seen feeding on (table 1). In early spring they fed on hemlock, spruce, and aspen for several weeks. Then they used red maple heavily in May and at least until mid-June. By late May the three birch species had assumed increasing importance, which continued throughout the summer. In dry years, apparent moisture stress in birch and other hardwoods drove the sapsuckers back to the hemlocks. Some other tree species provided important food in the fall.

The sapsuckers had feeding habits that made repeated observations easy. The same birds made frequent trips to the same tree daily. These visits continued for periods lasting several days to several months. Sapsuckers drilled only on live trees to get sap and bark. It was easier to identify sap feeding than bark feeding.

• A bird's visit to a number of existing fresh holes, and sap-collection activities not accompanied by drilling.
• Definite rapid siphoning of sap from holes – where sap had filled the holes and its removal could be clearly seen.
• Wet sap on the tip of a bill.
• Insertion of the bill tip at the bottom of sap holes and moving it from one side to the other, to wipe up, or to siphon, small quantities of sap.

When sap occurred in very small quantities, I could not always tell definitely whether the bird was feeding on sap, phloem, or other tissues.

My observations gave substantial evidence of sap feeding on 14 of 19 species observed in 1964. And by 1968 sap feeding had been observed on 23 of the 28 species listed in table 1: the exceptions were white spruce, tamarack, beech, hophornbeam, and elder. Sap feeding probably occurred on all species except elder, but there were too few observations to establish this definitely.

It has been reported that sapsuckers get intoxicated by feeding on fermented sap or fruit. I did not see any evidence of intoxication. Sapsuckers normally avoided fermented sap and fed only on fresh sap. Among the 532 birds I saw feeding on sap, only on two occasions did I see birds take fermented sap. One juvenile bird tasted it and went immediately to fresh sap. Another time, a pair of adults fed their fledglings on solidified fermented paper birch sap. The evidence suggests that this probably was necessary because red squirrels (Tamiasciurus hudsonicus loquax Bangs), feeding only on fresh sap, denied the birds access to their own sap holes.

Woody tissues. – Feeding on live bark tissues was common. I often saw small bits of it being eaten. I assumed that it was phloem, but specimens were not taken from the birds for identification. I do not know of any report that cites precise identification of the bark tissues sapsuckers eat.

A different kind of feeding occurred near the cambium, on what I assumed to be new callus tissues. It occurs around the perimeter of the sap pits. Specimens were not taken for examination. This kind of feeding is common and distinctive. The very rapid taps with the bill are noticeably faster than those used for other bark feeding or drilling.

None of the holes I saw extended noticeably into the xylem. A few of the soft outer cells might be eaten.

Drying of the xylem and bark at the holes is a serious matter. Once dead, the xylem does not join again on that plane, even if new wood eventually covers it. Some effects of that have been described (3, 4, 9).

Insects. – Insects were a major food of sapsuckers. My 1,301 recorded observations of insect feeding do not represent the total observed. Sapsuckers were experts at catching insects in mid-air. But they accepted other opportunities in virtually any location. Insects made up the bulk of the nestlings' food. Adult birds generally added sap to a bill full of insects just before flying to the nest. To a lesser extent, insects were fed without the sap.

Many of the birds' favorite hardwood trees probably will be killed by repeated drilling within future years – assuming that the sapsucker population remains fairly constant. Here is the tree mortality that sapsuckers had already produced among all damaged trees found within ten nesting territories in 1964:

Intensive feeding can kill a small tree in a single season, but generally two or more years of drilling are needed to kill trees 8 inches d.b.h. or larger.

Sapsucker damage was found on 51 species. Feeding was actually observed on 28 species (table 1). The other species, on which-only damage was observed, were:

The conifers and the less-favored hardwoods may continue to receive some damage annually, but most of them probably will live. This is based upon the assumption that drastic changes will not occur in either the stand compostion or the sapsucker population.

The quality of the wood in these living trees will depend upon the frequency and severity of the damage. Ring shake and other quality-reducing internal defects in trees remaining alive after being damaged by sapsuckers have been described else-where (3, 4, 8, 9). Bird peck defect caused by sapsuckers has been discussed in numerous publications. Many of our living trees will contain the same kinds of defects.

other damaged trees may heal and escape further damage for an unpredictable number of years – provided they are not used by subsequent sapsucker populations. Damaged wood fibers will remain within the trees and reduce the wood quality. Some trees may escape further damage for many years, and then may be revisited by sapsuckers and may be killed in one summer by intensive feeding.

Hemlock. – Eastern hemlock (Tsuga canadensis (L.) Carr) was the conifer that sapsuckers fed on most in this area (table 1). Repeated annual sapsucker attacks on hemlocks are common in this region. Both local and migrant birds are responsible.

This feeding on hernlocks was seasonal: it was not continuous throughout the growing season. Most feeding on hemlock occurred upon arrival of the birds about mid-April and continued into May. This period overlaps with early concurrent feeding on red spruce and later heavy feeding on paper birch and red maple in May.

Droughts apparently affected the feeding pattern. Sapsuckers fed heavily on hemlocks in spring 1964, switched to hardwoods, and then resumed feeding on hemlocks to a lesser extent later in the summer – after the sap in hardwood species diminished. Essentially the same sequence occurred in 1968, when severe moisture stress apparently occurred in paper birch and red maple about the last week of August, and the birds had to feed on hemlocks. Within 2 weeks, and after rains, sapsuckers had abandoned the hemlocks and returned to the paper birches.

When droughts did not occur, sapsuckers seldom fed on hemlocks in summer. With adequate rainfall in July 1966, and an abnormal abundance in the summer of 1967, sapsuckers used hemlocks very little in those summers. juvenile birds did a little casual drilling late in those summers, but the adults retained possession of their favorite paper birches.

Red spruce. – Sapsucker work on red spruce (Picea rubens Sarg.) occurred mostly during late April and early May, about the same time as on hemlock (table 1). The only drilling observed on red spruce after 11 May, 1964 was done by two juvenile birds, one on 16 July and one on 14 August. A similar pattern occurred in the other years.

Other conifers. – The sapsuckers showed little interest in black spruce (Picea mariana (Mill.) B.S.P), balsam fir (Abies balsamea (L.) Mill.), northern white-cedar (Thuja occidentalis L.), and eastern white pine (Pinus strobus L.) within the stands under observation (table 1). Most of the summer drilling on these conifers seemed to be casual exploration by juvenile birds.

Birches. – The birches were the favorite trees of sapsuckers. Paper birches (Betula papyrifera Marsh.), most abundant in the stands, were used heavily. Few yellow birches (B. alleghanien Britton) were available, but most of those found had been killed or damaged by sapsuckers. Gray birch (B. populifolia Marsh.) was used more than the data indicate. Damage found on gray birches was not always useable because birds were absent. In several areas gray birches seemed to be preferred over nearby paper birches.

Data were inadequate for comparing relative preferences for the three birches.

Paper birch. – Among all the trees available, paper birch appeared to be the favorite species. it normally was fed upon from April to October (table 1). Paper birches that earlier in summer had served as major sources of food for adults and nestlings were closely guarded later by adult birds, from late July until the migration began in late September or early October.

Paper birch remained the favorite species of sapsuckers even in August and September 1964 and 1968 when drought conditions appeared to diminish sap flow drastically, compared with earlier in the summer. Moisture stress in the birches as appeared to force sapsuckers to work on hemlocks in both years. They returned to birch later. Adult birds continued to guard their favorite band of holes and appeared to feed more heavily upon the birch bark. Bark was removed between vertical rows of holes, thus enlarging them into abnormally wide and long strips (0.5 inch to about 1.25 inch wide, and up to several inches long).

Although most holes were drilled on birch trunks, holes also were made on branches. Drilling on branches happened mostly in spring. The birds also commonly drilled substantial numbers of holes on birch branches in late summer, when their preferred holes on the trunk no longer produced sap. Improved sap production after heavy rains brought birds back to the trunk, except on trees rapidly dying from sapsucker damage.

Red maple. – Red maples (Acer rubrum var. rubrum L.) were used heavily by sapsuckers in spring of all 5 years (table 1). But in summer use varied widely and appeared to depend upon the amount of rain. After 25 June, 1964, use had noticeably lessened, about half the time consisting of casual drilling around old holes by juvenile birds. During the exceptionally dry summers of 1964 and 1968, red maples were used for casual drilling by both juveniles and adults while they awaited a turn at the sap holes on a favorite birch. But such feeding was of short duration and few fresh holes appeared on red maples during July and later. The last sap feeding on red maple in 1964 was observed on 8 June. It appeared that feeding on bark tissues, rather than sap, occurred from July on.

The abnormally wet summer of 1967 produced a marked difference in sapsucker use of red maples, as compared with the drought summer of 1964. In 1967 some red maples produced sap throughout the summer and were used into October. Even when that happened paper birches were more favored (table 1).

In relatively dry summers some red maples on the edge of marshy ground were used more and longer than nearby red maples on higher, drier ground. However, in general, these sapsuckers began leaving hemlocks and spruces and turning gradually to intensive feeding on red maples in early May, and this continued until about mid-June. By then the birds had become more interested in the three birch species. There was no sharp demarcation in use of different tree species; it was a gradual process, with considerable overlapping.

Sugar maple. – Only a few sugar maples (Acer saccharum Marsh.) occurred in the territories studied. On 18 June, 1964 strong flows of sap occurred at sapsucker holes on a 7-inch d.b.h. sugar maple and on a nearby 11-inch paper birch on our Maine forest. These trees served as major sources of sap for a pair of sapsuckers feeding nestlings. The same sapsucker activities occurred on 30 June, 1966 on a 12-inch sugar maple in northwestern Elk County, Pennsylvania.

Aspens. – Neither of the aspen species appeared to be favorite trees for sapsucker feeding except in late April, when sapsuckers commonly fed on sap from the branches of quaking aspen (Populus tremuloides Michx.) and bigtooth aspen (P. grandidentata Michx.) (table 1). The holes were made within 3 to 5 feet of the branch tips.

Aspens here normally are not used as much by sapsuckers as birch or maple. Sapsuckers fed lightly on the boles of the largest aspens. I found small quaking aspens damaged on clearcut strips. The smallest quaking aspen I have found with sapsucker holes was 1-3/4 inches d.b.h. it appeared that aspen here was used more when other preferred species were not available.

However, aspens are favorite sapsucker nesting trees in this area. Of the 29 nests found in 1964, 25 were in aspens.

Red oak. – Sapsuckers fed on northern red oak (Quercus rubra L.) more than is indicated by table 1. Other evidence of heavy feeding was found when the birds were not present. Few oaks occur in the stands where these observations were made.

When moisture stress increased in paper birch, resident sapsuckers frequently fed on red oaks available within their territory. This occurred with increasing frequency in September. It was continued by migrants into October.

Sapsuckers killed one or two northern red oak limbs about 5 inches diameter on several trees on the Penobscot Experimental Forest. But that is not common here because the birds prefer paper birch, and the supply is ample. Local birds and migrants do not feed on many of the oaks here long enough to form the large bands of holes necessary to kill the stems.

I do not know for sure that oaks become favorite trees as the birds proceed south. But the limited evidence available suggests that northern red oak and other oaks may be major species for sapsucker feeding during their fall migration. Other kinds of oaks observed in Massachusetts and Connecticut appear to have been damaged by fall migrants.

In any region where oaks are a favorite species, and sapsuckers stay long enough to make many holes in individual trees, greater damage probably would result than occurs here. Oaks are damaged on the migratory and wintering ranges south of here (4).

Ash. – Neither white ash (Fraxinus americana L.) nor black ash (F. nigra Marsh.) were fed on very much by sapsuckers in this area (table 1). Only juvenile birds were seen drilling white ash in 1964. An adult that I disturbed near its nest drilled one hole in a black ash; this appeared to be a diversionary action, rather than feeding. Intensive feeding does occur on a few ash here in early fall, but it is rare – compared with feeding on most other hardwoods.

Another common practice is light tapping and wiping that appears more to involve bill cleaning than feeding. Loafing and preening occur at the same places. I have seen this repeatedly on individual ash and northern white-cedar. It also occurred on the edges of dead bark on other species. This should not be misinterpreted as feeding.

Other hardwoods. – Serviceberry (Amelanchier laevis Wieg.) was used mostly in the fall (table 1). The birds paid it little attention until birch sap became scarce.

In the fall of 1967 and 1968, the same 8-inch d.b.h. serviceberry was used heavily by local birds and by migrants as late as 13 October, 1967. In early October 1967 profuse quantities of sap flowed, diminishing rapidly as the foliage colored. Sap flow in 1968 was noticeably less, perhaps because of the summer drought.

Is serviceberry a favorite species of fall migrants? It may be if it consistently produces sap later than other hardwoods along the migration route, as it has here. The literature does not provide an answer.

Striped maple (Acer pensylvanicum L.) generally is a small tree in this region. I observed feeding on it only once, in April. A few other damaged trees were found when birds were not present. It appeared to be low on the preference list when other and larger hardwoods were available.

Other hardwood species sapsuckers fed upon only occasionally, and generally lightly, were American basswood (Tilia americana L.), balsam poplar (Populus balsamifera var. balsamifera), American beech (Fagus grandifolia Ehrh.), and eastern hophornbearn (Ostrya virginiana (Mill.) K. Koch). Most of that feeding occurred here in late September and early October (table 1). Local and migrating birds were involved.

American elm (Ulmus americana L.) was fed upon lightly in Maine. It does not appear to be a preferred species if birch and maple are available and producing sap. If not, heavy feeding may occur on elm for short periods, especially late in summer. It also has served as a convenience species near some nests in May and June, generally when red maples were not available.

Elm (Ulmus spp.) and silver maple (Acer saccharinum L.) were damaged heavily by sapsuckers 18 August, 1965 and earlier, near Kentville, Nova Scotia. Severe damage had occurred on 10 elms and 6 maples, ranging from 3 inches to 14 inches d.b.h. All were shade trees planted around a large motel. The manager's deep concern about the damage was well justified.

Many berries of red-berried elder (Sambucus pubens Michx.) were fed to young sapsuckers by their mother on 6 July, 1964. No evidence of sap or bark feeding on this shrub species was seen.

Speckled alders (Alnus rugosa (DuRoi) Spreng.) with heavy sap flow held the interest of an entire family of sapsuckers on 13 July, 1964. Other stems of this species were found earlier and later where sapsuckers had fed heavily, but the birds were not seen feeding. Some alders were killed. Most of the feeding appeared to occur about mid-summer. Alder has been an important food species at some locations in Maine where trees higher on the preference list were not available.

In Maine, most of the feeding on apple (Malus spp.) trees apparently occurred after late summer. Local birds used it when birch sap was not available in sufficient quantities. The limited evidence I have for Maine indicated very heavy feeding on apple trees by fall migrants, and limited use by local birds at other seasons.

The damage would be much more severe on apple trees if the sap holes were as large as those made during summer feeding on other hardwoods. Small gimlet-type holes normally are found here. Some larger sap holes were made by local juvenile birds in August 1968.

If sapsuckers visit an area, one of the easiest places to see the damage is in an old apple orchard, or near wooded margins in producing orchards.

Most of the trees damaged had an average d.b.h. of 4 to 8 inches. However, the range of diameters observed varied from a minimum of 2 inches to a maximum of 16 inches.