USGS - science for a changing world

Northern Prairie Wildlife Research Center

  Home About NPWRC Our Science Staff Employment Contacts Common Questions About the Site

Habitat Distribution and Territoriality In the Dickcissel and Red-Winged Blackbird

Results and Discussion

The Red-Winged Blackbird

Red-winged blackbirds (Agelaius phoeniceus) are quite common in the U.S., breeding from coast to coast, north well into Canada, and south into Central America. Redwings are most commonly associated with marshland habitats but are also known to frequently occupy upland habitats, especially in the Midwest where the amount of marshland is considerably less than elsewhere. The redwing was of interest in this study because it occupied some of the same habitat as the dickcissel populations that were studied in addition to Tuttle Creek marsh. I was therefore curious as to what contrasts, if any, exist between the habitat distributions and territorial systems of the two species.

None of the literature that I encountered on the red-winged blackbird gives a definite hypothesis for the role of territorial behavior but all acknowledge the fact that the redwing is a territorial species. Many note that any adult male redwing will exhibit very aggressive behavior (even interspecific) toward any intruder on his territory (Emlen and Nero, 1951; Nero, 1956; Orians, 1961; Orians and Collier, 1963; Case and Hewitt, 1963). Aggressive behavior such as this would be expected to discourage the settling of any unsettled males. Thus, most of the literature seems to imply that territorial behavior probably limits density in the redwing. Orians (1961) and Dolbeer (1976) especially make note of this.

Only two studies were encountered in the literature that dealt with the redwing in upland as well as marsh habitats (Case and Hewitt, 1963; Dolbeer, 1976), but neither made an attempt to explain the habitat distribution observed.

Density Dynamics

If territorial behavior truly does limit density as implied by several authors, one could make the same predictions as with the dickcissel regarding rate and sequence of fillage of the different habitats. That is, the favored habitat should fill first to a maximum density followed by increases in densities in the less suitable habitats. I was not able to observe the filling of habitats over time since most spring migrations of males are finished long before the time when this study was begun. Also, apparently little movement of birds occurs once spring migration is complete. Some studies encountered in the literature dealt with populations which experienced virtually no change in density throughout the breeding season (Emlen and Nero, 1951; Dolbeer, 1976).

Figure 4 is a graph of densities in the three study areas over the time-span of this study. The marsh is evidently a favorite habitat of the redwing, which is no new discovery. The marsh was the only area to show any real changes in density. No ready explanation is available for this except that the marsh covered quite a small area so that the departure of just one bird or even the movement of one bird off the flagged grid resulted in quite a change in density as computed per 100 acres. The drop in density in all study areas towards the end of the time period reflects the nearing of the end of the breeding season, when most nesting is finished and territorial behavior of males gives way to flocking behavior. The relatively small difference in density between oldfield and prairie suggests that there may not be much difference between the two for redwings. All redwings encountered on the prairie occurred in bottom habitats in which the vegetation was tall, dense, and composed mainly of forbs, much like the oldfield. No conclusions on type of habitat distribution or territorial function can be drawn from this data.


As with the dickcissel, to identify the habitat distribution of the red winged blackbird, one must compare suitabilities as well as densities in the study areas. In choosing indices of suitability we must be sure that our assumptions of what constitutes suitability are sound since any conclusions reached will depend on these assumptions. Suitability of a habitat, it will be remembered, is related to the genetic contribution of resident adults to the next generation. Therefore, a measure of breeding success is again assumed to be a valid index of suitability

The red-winged blackbird is also a polygynous species. This implies that females have a free choice of mates and will settle in the habitat which affords the best chance of success. Therefore, the female/male sex ratio should also be a valid suitability index, as with the dickcissel. Some problems were encountered with these assumptions and will be discussed under the heading of sex ratios.

Breeding Success

The same components that were used to measure breeding success in the dickcissel were also used to give an assessment of success in the red-winged blackbird. One additional component, mean clutch size, was added since this data was available (see Table 5). It was calculated as the mean number of eggs laid per nest. Only nests with full clutches were used in the calculation. All other components were calculated in exactly the same manner as for the dickcissel.

Table 5. Breeding Success of Red-Winged Blackbirds.
  Oldfield Marsh Prairie
Egg Failure 0.105 0.067 0.091
Mean Clutch Size 3.60 3.60 4.00
Fledging Success (F/N) 1.41 1.14 1.20
F/H 0.73 1.00 0.64
Parasitism (%) 29.4 57.1 40.0
Predation (%) 55.6 56.3 64.7
See text for explanations of the components and their calculations.

Some differences exist but were insignificant as tested with a chi-square. Mean clutch sizes exhibited no significant differences as tested with a z-score.

It is evident that no one habitat is clearly favored over any of the others. Comparing marsh and oldfield, oldfield is favored in three components while the marsh is favored in two. Oldfield is favored in two components over prairie while prairie is favored in four. Marsh is favored over prairie in three components but prairie is favored in the other three components over the marsh. No consistent trend is apparent and the differences appear to be truly random. In view of the statistical insignificance of all differences and the lack of any consistent trend in the data, we can only conclude that there is no difference in suitability among the habitats as measured by breeding success. Dolbeer (1976) studied redwings in an oldfield habitat and found that the number of young fledged per territory in this habitat was as high as or higher than all but one of the estimates for marsh habitats reported by other investigators. Dolbeer observed sex ratios that were no higher than those of marsh habitats. This implies that there would not be more nests per territory in his oldfield habitat than in marsh habitats. Thus, the high fledgling per territory figure could not be a result of more nests per territory in oldfield than in marsh.

Sex Ratios

Earlier it was assumed that a female/male sex ratio would be a good index of habitat suitability for redwings since the redwing is a polygynous species. In reviewing the literature on redwings, however, frequent references to female territorial behavior were encountered. Nero (1956) noted that females show aggressive defense of a small area around their nests against other females, especially during the early part of the breeding cycle. Case and Hewitt (1963) observed the same behavior. I also observed aggressive displays by females in the vicinity of their nests. Holm (1973), Orians (1961), and Nero (1956) all contend that female territorial behavior plays an active role in determining the number of females able to nest in one male territory. In addition, Weatherhead and Robertson (1977) found that female choice did not correlate with territory quality but that female choice was strongly influenced by male behavior. In light of these observations, the female redwing cannot be assumed to have a free choice of nesting sites and suitability will not be properly indexed by the sex ratio.

The sex ratios observed in this study are presented in Table 6. None of the differences are statistically significant. On the basis of the preceding discussion, no conclusions as to habitat suitability as indexed by sex ratio will be made. Of course, the effects of female territoriality would be negated if male territory sizes were the same in all habitats. Then all male territories could theoretically hold the same number of females regardless of habitat. However, overall territory sizes in oldfield were significantly greater than marsh territories (z = 8.612, P < .01; see Table 7). Prairie territories averaged much larger than marsh territories (z = 51.890, P < .01), and also a bit larger than oldfield territories (z = 3.488, P < .01). Correspond these differences to the sex ratios and see that oldfield ratios were consistently a little higher than marsh ratios, prairie ratios were a bit higher than oldfield ratios, and higher yet than marsh ratios. Thus, there may indeed be a positive correlation between size of territory and number of females per territory.

Table 6. Red-Wing Sex Ratios (Female/Male)
Time Period Oldfield Marsh Prairie
6/13-6/19 1.38 1.50 1.67
6/22-6/26 1.83 1.60 1.67
6/30-7/1 2.00 1.40 2.33
7/6-7/11 2.17 1.40 ----
7/15-7/18 1.83 0.80 ----
All Dates 1.81 1.35 1.89
Sex ratios for all dates were obtained by summing the total number of males
and females present during each time period and calculating the ratio. Dotted
dark line indicates no sex ratio determined.

Figure 5 presents a plot of sex ratios vs. density. The negative correlation is not strong but is significant (t = -2.667, df = 11, P < .05). Nero (1956) observed no correlation between sex ratio and density. I believe the constant decrease in sex ratio with increasing density reflects a decrease in females per territory as territory sizes decrease, since territory sizes most likely decrease with increased density. Smaller male territories would presumably allow fewer females to settle in each one. This is consistent with the findings regarding female territoriality and aggressive female behavior.

Since habitat suitability for redwings cannot be validly indexed by the sex ratio, we can make no statements concerning the relationship between suitability and male density on the basis of the graph in Figure 5. Caccamise (1977) reported that nests incurring in-nest losses were located in areas of significantly greater nest density than those nests which did not incur such losses. This suggests an inverse relationship between suitability and density in the redwing. However, in the same study, Caccamise found no correlation between hatch/egg and fledgling/hatch ratios and nest density.


As noted earlier, aggressive territorial behavior is a well - documented fact for the male red-winged blackbird. As with the dickcissel, the data and discussions regarding territoriality will be used to find clues as to the function of redwing territoriality and as to the type of habitat distribution exhibited by the redwing. Data is limited to male territoriality, so the discussion will also concern only male territoriality unless otherwise noted.

Much conflict exists among investigators regarding stability of territories and relation of territory size to male density and/or vegetation. Emlen and Nero (1951) observed absolutely no shifting of territory boundaries and no relation of territory boundaries to recognizable features of vegetation or terrain in a marsh habitat. Nero (1956) made the same observations in a later study but also noted that territory size generally varied inversely with population density. Orians (1961) noted a dependence of territory size on type of vegetation but also observed an inverse correlation between territory size and density. Case and Hewitt (1963) also observed the territory size - density inverse correlation. In contrast to the 1951 study of Emlen and Nero, however, they noted great variation in size and shape of territories and a dependence of size and shape on the pattern of vegetation and presence of song perches. Holm (1973) also observed a strong relation between territory size and type and density of vegetation. The redwings of my study were observed to have territories whose positions shifted very little in relation to each other but whose boundaries fluctuated quite often. Only boundaries of those territories in the marsh showed any relation to structure of vegetation. One or two territories were observed whose boundaries seemed to follow the edges of cattail clumps.

The fact that redwings can exhibit very aggressive behavior and the fact that very stable territory boundaries were observed in some studies suggest that territoriality may have a density - limiting function in the redwing. The observations on inverse correlation between territory size and density do not offer much evidence in support of any territorial hypothesis. None of the studies indicated whether a minimum territory size was reached or not.

Figure 6 shows a definite inverse correlation between territory size and male density for this study. Several things are implied by this graph. First of all, the regression curve is not parabolic, it is exponential. The curve fits with a high degree of correlation and is significant (t=-12.79, df=11, P<.01). An exponential curve has no minimum. This corresponds well with the data points which show that no definite minimum territory size was reached. A parabolic curve would have implied a minimum territory size as with the dickcissel. Since there was no apparent minimum territory size, the density would not reach a limit except at a ridiculously high level. Of course, redwing territories will not be infinitely compressable, but the data does indicate that unsettled redwings would be able to settle in habitats of very high density if they so chose.

Since no reasonable limit is imposed on density and no minimum territory size is apparent, we cannot say that territorial behavior is functioning to limit density. The function of territoriality is then either to serve as a density cue or only to space the individuals. Which of these is true will depend on what we decipher the actual habitat distribution to be.

Table 7 presents mean territory sizes as broken down into habitat and mated status of males. The overall sizes are not entirely consistent with the inverse correlation observed between territory size and density. Oldfield territories were significantly larger than marsh territories (see discussion of sex ratios for test statistics). Oldfield densities averaged significantly lower than marsh densities (x2 = 29.71, df = 5, P < .01). Prairie territories were also much larger than marsh territories while marsh density was much higher than prairie density (x2 = 43.76, df = 5, P < .01). The one inconsistency arises in a comparison of oldfield and prairie. Prairie territory sizes were significantly larger yet there was no significant difference in density between the two areas. No explanation for this discrepancy was available other than that perhaps there was a greater dependence of territory sizes on vegetation in the prairie than in the oldfield.

Table 7. Territory Sizes (Mean Acres + S. E.) of Male Redwings
Dates Mateless Monogamous Polygynous All Males
6/13 -------- 0.365 .058 (5) 0.844 .122 (3) 0.545 .101 (8)
6/22-6/23 -------- 1.265 .067 (2) 1.152 .218 (4) 1.189 .149 (6)
6/30 -------- 0.798 .000 (1) 1.079 .121 (5) 1.032 .110 (6)
7/6 -------- 1.134 .000 (1) 1.371 .297 (5) 1.331 .250 (6)
7/15 -------- 0.982 .065 (2) 1.448 .155 (3) 1.262 .140 (5)
All Dates -------- 0.750 .122 (11) 1.187 .109 (20) 1.032 .090 (31)
6/15 -------- 0.061 .014 (3) 0.196 .038 (3) 0.129 .034 (6)
6/23 -------- 0.142 .000 (2) 0.213 .050 (3) 0.185 .034 (5)
6/30 0.165 .005 (2) -------- 0.253 .021 (3) 0.218 .023 (5)
7/10 0.517 .262 (2) -------- 0.234 .008 (3) 0.347 .122 (5)
7/17 0.218 .024 (2) 0.215 .024 (2) 0.245 .000 (1) 0.222 .015 (5)
All Dates 0.300 .118 (6) 0.128 .028 (7) 0.226 .017 (13) 0.217 .030 (26)
6/19 -------- 1.797 .000 (1) 1.139 .283 (2) 1.358 .260 (3)
6/26 -------- 1.571 .000 (1) 1.518 .178 (2) 1.536 .120 (3)
7/1 -------- -------- 2.174 .121 (3) 2.174 .121 (3)
7/11 -------- - - - - - - - - - - - - - - - - - - - - -
7/17-7/18 -------- - - - - - - - -------- - - - - - - -
All Dates -------- 1.684 .113 (2) 1.691 .215 (7) 1.689 .166 (9)
Numbers in parentheses are sample sizes. Dotted line indicates no males of that mated status
present. Dashed line indicates male(s) of that mated status present but not mapped.

Comparison of territory size between birds of different mated status also produced some interesting results. Polygynous males had significantly larger territories than monogamous males in both oldfield (t = 2.526, df = 29, P < .01) and marsh (t = 3.167, df = 18, P < .01). The difference on the prairie is insignificant. The marsh was the only area to ever harbor any mateless males. The differences between mateless and monogamous and mateless and polygynous males' territories are insignificant there. The overall trend of the data seems to indicate that males with larger territories acquire more mates. It is quite possible that this is a reflection of female territorial behavior which limits the number of females nesting in any one territory. This was also suggested earlier as a result of sex ratio data. The fact that mateless males in the marsh had slightly larger (but insignificantly so) territories than mated birds is because of the presence of one male on the area during the second week of July (July 10) who entered the marsh and occupied a space vacated by another male. He extended the boundaries of this territory to include an area almost three times the size of the next largest territory. It is not surprising that he was mateless. A male entering an area at such a late stage in the breeding season cannot reasonably be expected to acquire a mate.

Other investigators have obtained conflicting results with similar types of data. Linford (1935), as cited by Orians (1961), found that territories of polygynous males were twice the size of monogamous males'. Orians (1961), Case and Hewitt (1963), Nero (1956), and Holm (1973) found no correlation between number of females and territory size.

Habitat Selection

The same vegetation height measurements that were used for dickcissels were also assigned to male redwing territories. This data is shown in Table 8. Utilized vegetation was seen to be significantly taller than non-utilized vegetation (z = 2.543, P < .01). This implies that males are responsive at least to the height of vegetation, favoring taller vegetation. The difference in vegetation height between polygynous and monogamous males' territories is not significant. Since polygynous males' territories are larger than monogamous males', this lends strength to the possibility that some males acquire more mates simply because their territories are larger and will hold more females rather than that their territories contain more attractive vegetation. It must be remembered, however, that vegetation height is only a crude index of the total cover of the vegetation.

Table 8. July 14 Oldfield Vegetation Heights (Mean cm. + S.E.)
Territory Height
Utilized vegetation 103.59 5.67
Non-utilized vegetation 87.40 2.90
Monogamous male territories 100.85 21.09
Polygynous male territories 104.69 7.66
Vegetation heights as they pertain to male redwing territories.


To conclude the discussion of the red-winged blackbird, the results of each set of data must be pooled.

Observations on density dynamics offered no real clues as to the type of habitat distribution achieved by the redwing except that the marsh is evidently preferred over oldfield and prairie. Observations on types of vegetation in oldfield and prairie redwing territories, plus the fact that no significant difference existed in density between the two plots suggested that the two plots may actually be the same habitat type as far as the redwing is concerned.

In light of the data on breeding success, it is concluded that no real differences in suitability exist among any of the three areas. On the basis of evidence of female territorial behavior, sex ratio was rejected as a legitimate index of habitat suitability for the redwing. Since suitabilities of oldfield and prairie are evidently the same, and since no substantial difference in density exists, I contend that the two areas are indeed the same habitat type for the redwing.

To apply the theory developed in the introduction, suitabilities and densities must be compared in marsh and upland (oldfield and prairie) areas. Suitability was seen to be essentially the same in both areas. But the marsh supported a significantly higher density of birds than the upland habitats. Therefore, the actual habitat distribution of the redwing is most nearly ideal free, and the function of territoriality in the redwing is evidently to serve as a density cue so that unsettled birds can avoid heavily populated areas and settle where their chances of success will be just as good. No other apparent density cue was observed.

The data on territory sizes lend rather strong support to this conclusion. The data do not fit an ideal dominance type of distribution. An ideal dominance distribution implies that a limit is imposed on the density and that once this density is reached, unsettled males must be forced into less suitable habitat. No reasonable limit to density was foreseen for the redwing and territory sizes reached no minimum. Thus, the data support the ideal free distribution.

No other studies that I know of have been done for the purpose of defining the habitat distribution and function of territoriality in the redwing. Thus, it would not be safe to generalize the conclusions of this local study for all redwings unless similar studies done over different parts of the redwing's breeding range produced similar results.

Previous Section -- Dickcissel
Return to Contents
Next Section -- Conclusion

Accessibility FOIA Privacy Policies and Notices

Take Pride in America logo logo U.S. Department of the Interior | U.S. Geological Survey
Page Contact Information: Webmaster
Page Last Modified: Friday, 01-Feb-2013 19:11:22 EST
Sioux Falls, SD [sdww54]