Foraging behavior of the Eurasian Three-toed Woodpecker subspecies Picoides tridactylus funebris in southern Gansu, China

Introduction

The Three-Toed Woodpecker is circumpolal distributed across the Northern Hemisphere. This vast distribution area results in eight subspecies, which are currently divided into two separate species based on mitochondrial DNA: Picoides tridactylus in Eurasia and P. dorsalis in North America (Zink et al., 1995, 2002). The subspecies P. t. funebris (Fig. 1) is endemic to the Qinghai-Tibet plateau in China. No research is being ever conducted since its discovery in 1870 (Verreaux, 1870). Thus, P. t. funebris is a subspecies for which we are not aware of any reports on life history details including foraging behavior.

Figure  1.  The Three-toed Woodpecker subspecies Picoides tridactylus funebris

DownLoad: Full-Size Img PowerPoint

Foraging behavior in Three-toed Woodpeckers is strongly related to the availability of food which was shown to be the driving factor for selection of foraging sites (Hogstad, 1976, 1977, 1991; Pechacek, 2006). Bark beetles and wood boring beetle larvae that are both confined to the subsurface of dead and dying trees are reported to be the most important prey of Three-toed Woodpeckers (Pechacek and Kristin 2004). Morphological adaptations especially with regard to the shape of the chisel-like bill are the basic precondition to excavate these prey from foraging substrates.

Measurements of 12 museum specimens of P. t. funebris (6 males and 6 females, Y.Z. Zhu unpublished data) suggested that the differences in bill length between the sexes are less developed than in P. t. alpinus (< 4% vs > 8%) (Pechacek, 2006). This may have resulted in evolution of more serious competition between males and females because bill length differences are related to the pronounced foraging niche partitioning. Niche partitioning helps to prevent conflicts in cases when both sexes are foraging close to each other and was well described in P. t. tridactylus (Hogstad, 1976, 1977, 1991) and P. t. alpinus (Pechacek, 2006).

We therefore conducted field observations on P. t. funebris in Gansu Province, China to compare its foraging behavior with that of other subspecies we particularly aimed for 1) exploring foraging sites and foraging techniques, and 2) assessing the niche partitioning between the sexes.

Methods

The study was conducted at the Kache Forest Farm, Zhuoni County, Gansu Province, China (34°10′7″–34°38′20″N, 103°12′49″–103°48′50″E), with altitude ranging from 2500 to 4500 m above sea level. The Kache Forest Farm is composed of coniferous forests that are dominated by spruces (Picea asperata, P. wilsonii) and firs (Abies fargesii, A. faxoniana). Two large-scale logging operations were conducted in the area between 1949 and 1998, and afforestation took place in recent years.

A female P. t. funebris was caught in a mist net on 21 March 2007. It was banded and fitted with a 2 g radio transmitter (type BD 2G, Holohil, Canada). We followed the female by radio-tracking and collected foraging observation. The mated male individual was observed travelling together with the female frequently during the breeding season, and we therefore conducted observations on the foraging male as well.

Following field protocols by Pechacek (2006), foraging data was recorded every 15 s during an observed foraging bout. We recorded foraging behavior and associated parameters of foraging substrates. Foraging bout was considered as effective only if the woodpecker was observed foraging on a tree for more than 45 s and we could therefore record observations more than three times. Collected data reflected behavior or substrate use as a percentage of each foraging bout (i.e. percentage of time spent by displaying a particular foraging behavior or using a particular foraging substrate). To ensure the independency of foraging bouts, the next foraging bout was recorded after minimally one hour (Swihart and Slade, 1985), or alternatively, if the woodpecker moved at least 100 m from the previous location (Pechacek, 2006).

According to the sampling classification developed by Remsen and Robinson (1990) and Pechacek (2006), we distinguished nine types of behavior associated with foraging: pecking, peeling, sap-sucking, preening, territory defence, climbing, freezing, head swinging, and other. We summarized pecking, tapping and probing from the original classification (Remsen and Robinson, 1990; Pechacek, 2006) as pecking which represented foraging for prey hidden relatively deep in the foraging substrate. Scaling and gleaning from the original classification were summarized as peeling which represented foraging for prey available close to the substrate surface.

Parameters of foraging substrates were categorized and measured according to Pechacek (2006). We considered parameters that included foraging zone (i.e. base of the trunk, trunk, branches), substrate thickness, foraging height (i.e. lower, central and middle third of the foraging tree) and condition of the foraging substrate (i.e. alive, dead with and without bark). The thickness of foraging substrate was assessed referring to the bird itself (about 5.5 cm width). Additionally, we randomly placed 100 plots of 10 m × 10 m throughout the foraging area (i.e. home-range) of the male and female to assess the amount of snags (i.e. standing dead trees) in the foraging territory.

After recording a foraging bout, the nearest tree of random direction was selected as a reference. Tree species and diameter at breast height (DBH) were recorded for both foraging and reference tree, and then the feeding preference index (PI) was calculated to express the foraging preference, using the equation from Kells et al. (2001):

where V_k is the number of foraging visits of kth site, V_t the total number of visits to all sites, A_k the total number of kth reference, and A_t the total number of all references.

Analyses were conducted in SPSS 13.0. Wilcoxon Signed Ranks Test was used to test the difference between DBH of foraging and reference trees. Mann-Whitney U-Test was used to test for differences in foraging behavior between the sexes. Results were considered significant at p < 0.05. Standard deviations (SD) are given with means.

Results

We collected 117 observations on a foraging pair of P. t. funebris during the breeding season between April and August 2007. Of them, 89 accounted for the female and 28 for the male. Both partners were found foraging close to each other at 56 (47.8%) out of 117 occasions.

Amount of dead trees in foraging territory

The volume of snags in the foraging territory amounted 4.3 m³·ha^–1 (6 and 2.1 m³·ha^–1 for the male's and female's home ranges, respectively).

Foraging trees and behavior

P. t. funebris preferred live spruces and snags over other available sites (Table 1). Foraging trees had by 26% bigger DBH than the reference trees, whereas the mean DBH of the female's foraging trees was larger than that of reference trees (32.6 ± 8.8 vs 22.2 ± 9.2), and the DBH of the male's foraging trees was similar to the reference ones (33.1 ± 10.4 vs 30.2 ± 10.6) (Table 2). The main foraging technique of P. t. funebris was pecking (39.8 ± 27.0% of the foraging time) followed by the peeling (13.2 ± 15.0%).

Table  1.  Preferences for foraging trees in a pair of P. t. funebris during the breeding season in Kache Forest Farm, Zhuoni County, Gansu Province, 2007. A reference tree was selected as the nearest tree of random direction to the foraging tree.

Site	Foraging tree (%)	Reference tree (%)	Preference index
Fir	65.8	83.2	0.791
Spruce	18.8	10.3	1.825
Snag a	14.5	3.4	4.265
Other	0.9	3.4	0.265
Total	100	100	–
n	117	117	–
a Unidentified species.

 | Show Table

DownLoad: CSV

Table  2.  Diameter of foraging trees used by a pair of P. t. funebris during the breeding season in Kache Forest Farm, Zhuoni County, Gansu Province, 2007, including diameter of associated reference trees. A reference tree was selected as the nearest tree of random direction to the foraging tree.

	Foraging tree (cm)	Reference tree (cm)	Z a	p a	n
Male	33.1 ± 10.4	30.2 ± 10.6	–0.1	0.996	28
Female	32.6 ± 8.8	22.2 ± 9.2	–6.4	0.000	89
Z b	–0.2	–3.7
p b	0.858	0.000
Total	32.7 ± 9.2	24.1 ± 10.1	–5.8	0.000	117
a Results derived from Wilcoxon Signed Ranks Test; b Results derived from Mann-Whitney U-Test.

 | Show Table

DownLoad: CSV

Differences in foraging between the sexes

The mean DBH of the foraging trees showed no difference between male and female (p = 0.858). However, we observed the following differences between the sexes with regard to parameters of foraging substrates (Table 3): the female foraged more often in the upper third of a tree (p = 0.010), while the male was more frequently seen foraging in the lower third (p = 0.003). The thickness of the male foraging substrate was significantly larger than that of the female's (p = 0.013). The male foraged more often on the dead substrates covered with bark (p = 0.021). On the other hand, the female spent more time on living parts of trees (p = 0.017). Additionally, male used pecking more often than the female (p = 0.035; Table 4).

Table  3.  Parameters of foraging sites used by a pair of P. t. funebris during the breeding season in Kache Forest Farm, Zhuoni County, Gansu Province, 2007. The numbers reflect substrate use as a percentage of foraging bouts, i.e. percentage of time spent using a particular substrate parameter (except of thickness which is expressed in cm).

	Male	Female	Z	p	All
Foraging zone
Base of the trunk	14.3 ± 35.6	3.8 ± 18.3	–1.5	0.141	6.3 ± 23.8
Trunk	46.4 ± 50.8	37.4 ± 45.2	–0.96	0.926	39.6 ± 46.5
Branches	39.2 ± 49.7	57.7 ± 46.3	–1.2	0.248	53.3 ± 47.6
Foraging height
Lower third	28.6 ± 46.0	5.8 ± 23.2	–3.1	0.003	11.3 ± 31.6
Central third	33.8 ± 46.6	28.6 ± 44.4	–0.4	0.651	29.8 ± 44.8
Upper third	33.7 ± 47.8	65.5 ± 47.0	–2.6	0.010	58.9 ± 48.5
Condition
Alive	15.1 ± 35.6	28.9 ± 42.3	–2.4	0.017	25.6 ± 41.1
Dead with bark	78.9 ± 38.1	63.9 ± 41.2	–2.3	0.021	67.5 ± 40.9
Dead without bark	5.9 ± 15.3	7.3 ± 16.5	–0.1	0.893	6.9 ± 16.2
Total	100	100			100
Mean thickness of foraging substrates (cm)	17.0 ± 11.6	11.0 ± 8.2	–2.4	0.013	12.6 ± 9.9
n	28	89			117

 | Show Table

DownLoad: CSV

Table  4.  Foraging techniques and behavior associated with foraging observed in a pair of P. t. funebris during the breeding season in Kache Forest Farm, Zhuoni County, Gansu Province, 2007. The numbers reflect behavior as a percentage of foraging bouts, i.e. percentage of time spent using a particular foraging technique.

Foraging behavior	Male	Female	Z	p	All
Pecking	46.2 ± 29.5	37.8 ± 26.0	–1.8	0.035	39.8 ± 27.0
Peeling	13.0 ± 14.5	13.2 ± 15.3	–0.2	0.430	13.2 ± 15.0
Sap-sucking	3.4 ± 18.0	4.3 ± 17.0	–0.6	0.474	4.1 ± 17.2
Preening	6.1 ± 17.7	5.1 ± 16.5	–1.1	0.142	5.3 ± 16.8
Territory defense	0.4 ± 1.9	1.1 ± 6.0	–0.4	0.370	0.9 ± 5.3
Climbing	8.9 ± 13.3	9.7 ± 12.7	–0.3	0.374	9.5 ± 12.8
Freezing	3.8 ± 10.8	9.5 ± 18.9	–0.8	0.208	8.1 ± 17.4
Head swinging	7.5 ± 8.2	10.8 ± 11.9	–1.1	0.139	10.0 ± 11.2
Other	10.5 ± 13.7	6.2 ± 6.7	–1.4	0.079	7.2 ± 9.0
Total	100	100			100
n	28	89			117

 | Show Table

DownLoad: CSV

Discussion

P. t. funebris preferred foraging on live spruces and snags that were bigger than surrounding reference trees with an average DBH of 32.7 ± 9.2 cm. We found differences between the sexes with respect to foraging height, substrate thickness and condition of the substrate. The most frequent foraging technique was pecking (39.8% of foraging time) followed by the peeling (13.2%). The male pecked more often than the female, and the female preferred foraging on trees that were larger than those available in female's foraging territory. Although the fir was the dominant tree species in our study area by amounting for 83.2% of the randomly selected reference trees, P. t. funebris preferred spruces and snags which had higher population density of bark beetles than firs (Liu et al., 1994). The observed foraging on trees with relatively large DBH was likely also related to the more abundant beetle prey that typically occurs in large substrates (Hanula et al., 2000).

We found that P. t. funebris extensively foraged on dead branches (53.3% of the foraging time), while P. t. alpinus rarely did so (13%, Pechacek, 2006). This may have been adaptation to the small amount of snags available in the foraging territory in our study area (4.3 m³·ha^–1). In contrast, snag amount was much higher in habitats of P. t. alpinus in the European Alps (30 m³·ha^–1 of combined volume of snags and downed logs, Konnert, 2000). Woodpeckers may have therefore compensated for the lack of snags to obtain prey on another form of dead wood, the dead branches. We noted that sap-sucking was observed more often in P. t. funebris than in P. t. alpinus (4.1% vs 1.2% of the foraging time) (Pechacek, 2006), suggesting that P. t. funebris was more dependent on the tree sap than the other subspecies.

P. t. funebris showed pronounced vertical niche partitioning between the sexes. This was consistent with observations on Three-Toed Woodpeckers elsewhere (Hogstad 1977, 1991; Pechacek, 2006). We found distinct differences between the sexes with respect to use of three out of four investigated parameters of the foraging substrates. The male occupied for most of the foraging time the lower third of coniferous trees and thicker substrates, where beetle prey is typically more abundant than in the upper third and on branches (Chen et al., 1999; Liu et al., 2007). Excavation of prey from the substrates composed of large DBH and thick bark required more time investment by extensive pecking. This may have explained the higher proportion of the male's pecking behavior compared to female. Conversely, the female used more often dead substrates with bark already peeled off, and living substrates. These sites presumably contained less abundant food. We also found that the volume of snags was higher within the male's home range compared to that of the female (6 vs 2.1 m³·ha^–1). Moreover, we detected the male driving away the female for five times, which accounted for 8.9% of all records when both partners were found travelling together. We therefore concluded that the male dominated over the female by occupying prey-richer resources also in P. t. funebris despite of less obvious competitive advantages as demonstrated by little differences in the bill length between the sexes in comparison with P. t. alpinus (Pechacek, 2006).

Conclusion

We confirmed that P. t. funebris displayed foraging behavior and niche partitioning that showed similar patterns to those reported for other studied subspecies of the Three-toed Woodpecker. The observed niche partitioning, however, did not reflect well the expected strong competition for the best foraging sites based on less pronounced sexual dimorphism (here represented by bill length) in P. t. funebris.