Sample size and duration of time-activity budgets

The total number of time-activity budgets included in the analysis was 292 (Wood Stork, n = 29; White Ibis, n = 41; Great Blue Heron, n = 5; Great Egret, n = 217). Mean duration for all time-activity budgets was 11.1 min (SD = 7.9, Range 0.2 - 35.4) but there were differences among species. Great Blue Herons averaged 6.8 min (SD = 2.0, Range 4.4 - 9.2), Great Egrets averaged 12.8 min (SD = 8.3, Range 0.2 - 35.4), White Ibis averaged 5.4 min (SD = 3.5, Range 1.8 - 21.2), and Wood Storks averaged 7.5 min (SD = 4.4, Range 2.5 - 22.9). Plots of prey intake rate versus duration of time-activity budgets (Fig. 1) showed no obvious correlation between the two parameters for the Wood Stork ( r = 0.01, P = 0.96) and White Ibis ( r = -0.26, P = 0.10). There was a strong negative relationship for the Great Blue Heron ( r = -0.98, P < 0.01) and a weaker negative relationship for the Great Egret ( r = -0.32, P < 0.01; (Fig. 1).

Fig. 1. Relationship between duration of time-activity-budgets and intake rates for the Wood Stork, White Ibis, Great Blue Heron, and Great Egret during foraging experiments, 1996.

Giving-up density

When pooled across all ponds, fish densities decreased until about day 12 then remained relatively constant for the remainder of the experiment. However, there were noticeable differences among treatments in the rate of decrease in fish abundance and in the lower limit of fish depletion (Fig. 2). Fish densities decreased fastest in the shallow water treatments followed by the medium and deep water treatments respectively, suggesting fish were more vulnerable in the latter. Note that fish density estimates and GUDs were not corrected for sampling efficiency (see above), thus resulting in an underestimate of true density.

Fig. 2. Estimated fish depletion rates in 10cm, 19cm, and 28 cm depth ponds during foraging experiments, 1996. Symbols indicate means and standard errors.

For all species but the Great Blue Heron, GUDs were lowest in the shallow water treatments followed by the medium and deep-water treatments respectively (Table 1). Also, in the deepest treatment, Wood Stork and White Ibis had higher GUDs than did the Great Egret and Great Blue Heron. The Great Blue Heron was unusual in that GUD was lowest in the shallow treatment but similar between the two deeper treatments.

Table 1. Giving-up densities of fish (fish/m2) for the Wood Stork, White Ibis, Great Blue Heron, and Great Egret at three water depths during foraging experiments, 1996. GUD50%max was calculated as the fish density at a given pond the day bird abundance for a given species dropped below 50% of the max for that pond. GUDmin was calculated as the fish density at a given pond on the day the last bird was seen in that pond. Means and standard errors reflect variability among ponds within a water depth treatment (n = 4).
	10 cm water depth				19 cm water depth				28 cm water depth
	GUD50%max		GUDmin		GUD50%max		GUDmin		GUD50%max		GUDmin
Species		SE		SE		SE		SE		SE		SE
Wood Stork	0.04	0.04	0.04	0.04	0.59	0.12	0.55	0.13	1.40	0.43	1.40	0.43
White Ibis	0.09	0.07	0.04	0.03	0.74	0.22	0.47	0.12	1.67	0.56	1.67	0.55
Great Blue Heron	0.15	0.12	0.15	0.12	0.98	0.30	0.98	0.30	0.81	0.29	0.81	0.29
Great Egret	0.15	0.12	0.01	0.01	0.43	0.11	0.25	0.11	0.95	0.36	0.68	0.28

Intake rates

Wood Stork

Wood Storks did not appear in the experiment until 9 days after fish were stocked. By the time they arrived, shallow water ponds were largely depleted of fishes and the medium depth ponds were partially depleted. Thus, no Wood Storks were observed in the 10-cm water depth treatment nor did they occur when fish densities were high. Storks were filmed only in the medium and deep-water treatments and only when fish densities were moderately low. During the time storks were filmed, fish densities in the medium depth ponds = 0.25 fish/m², SD = 0.09, Range 0.19 - 0.39, n = 8) were always less than those in the deep water ponds ( = 0.69 fish/m², SD = 0.27, Range 0.44 - 1.53, n = 21), precluding a powerful statistical test of depth and fish density effects.

Intake rates averaged 0.56 fish/min (Table 2). Intake rates were similar between the medium and deep-water ponds despite the shallower treatment having lower fish densities (Fig. 3). Pooling between depths, intake rates were higher when fish densities were higher than when they were lower (Table 2). The same trend was evident within the deep-water treatment.

Fig. 3. Intake rate versus fish density for Wood Storks at 2 water depths during foraging experiments, 1996.

The full statistical model was not significant and explained very little of the variation in intake rates (F = 0.87, P = 0.47, df = 29, R² = 0.09). The reduced main-effects model fit the data similarly (F = 1.32, P = 0.29, df = 29, R² = 0.09). It is important to note that for Wood Storks the range of depths and fish densities was not as great as with the other species and may have contributed to the non-significant models.

Table 2. Prey intake rates (fish/min) for the Wood Stork, White Ibis, Great Blue Heron, and Great Egret during foraging experiments, 1996. Fish density quantile 1 was 0 - 0.79 fish m2 and fish density quantile 2 was 0.8 - 1.59 fish m2. Fish densities are numbers of fish captured in throw traps uncorrected for sampling bias.
Water Depth Quantile	Fish density	Parameter	Wood Stork	White Ibis	Great Blue Heron	Great Egret
10 cm	1	fish/min	0	0.42	0.13	0.25
		SD		0.25	0.04	0.23
		N		15	2	22
		Range		0.13 - 0.97	0.11 - 0.16	0.03 - 0.88
	2	fish/min	0	0.56	0	0.15
		SD		0.39
		N		2		1
		Range		0.29 - 0.84
	All densities	fish/min	0	0.43	0.13	0.25
		SD		0.26	0.04	0.23
		N		17	2	23
		Range		0.13 - 0.97	0.11 - 0.16	0.03 - 0.88
19 cm	1	fish/min	0.51	0.68	0.15	0.19
		SD	0.17	0.48	0.05	0.13
		N	8	12	2	78
		Range	0.26 - 0.78	0.05 - 1.45	0.12 - 0.18	0.05 - 0.96
	2	fish/min	0	0.85	0	0.25
		SD		0.48		0.11
		N		8		13
		Range		0.31 - 1.57		0.05 - 0.40
	All densities	fish/min	0.51	0.75	0.15	0.20
		SD	0.17	0.48	0.05	0.13
		N	8	20	2	91
		Range	0.26 - 0.78	0.05 - 1.57	0.12 - 0.18	0.05 - 0.96
28 cm	1	fish/min	0.32	0	0.23	0.34
		SD	0.12			0.38
		N	9		1	35
		Range	0.13 - 0.50			0.07 - 2.14
	2	fish/min	0.77	0.58	0	0.22
		SD	0.33	0.52		0.12
		N	12	4		15
		Range	0.14 - 1.18	0.20 - 1.34		0.10 - 0.52
	All densities	fish/min	0.57	0.58	0.23	0.30
		SD	0.34	0.52		0.33
		N	21	4	1	50
		Range	0.13 - 1.18	0.20 - 1.34		0.07 - 2.14
All depths	1	fish/min	0.41	0.54	0.16	0.24
		SD	0.17	0.39	0.05	0.24
		N	17	27	5	135
		Range	0.13 - 0.78	0.05 - 1.45	0.11 - 0.23	0.03 - 2.14
	2	fish/min	0.77	0.73	0	0.23
		SD	0.33	0.47		0.11
		N	12	14		29
		Range	0.14 - 1.18	0.20 - 1.57		0.05 - 0.52
	All densities	fish/min	0.56	0.60	0.16	0.24
		SD	0.30	0.42	0.05	0.22
		N	29	41	5	164
		Range	0.13 - 1.18	0.05 - 1.57	0.11 - 0.23	0.03 - 2.14

White Ibis

White Ibises were filmed in all 3 depth treatments. During the time White Ibises were filmed, fish densities in the shallow water depth ponds ( = 0.41 fish/m², SD = 0.37, Range 0.03 - 1.37, n = 17) were less than the medium depth ponds ( = 0.90 fish/m², SD = 0.37, Range 0.04 - 1.32, n = 20), which were less than the deep water depth ponds ( = 1.37 fish/m², SD = 0.31, Range 0.90 - 1.52, n = 4).

Intake rates averaged 0.60 fish/min (Table 2). Intake rates in the shallow water treatments were less than in both the medium and deep-water depths (Fig. 4). Averaged across all depths, intake rates were higher when fish densities were high than when they were low (Table 2). The same trend was evident in the shallow and medium depths.

Fig. 4. Intake rate versus fish density for White Ibises at 3 water depths during foraging experiments, 1996.

The full statistical model was not significant (F = 1.94, P = 0.11, df = 40, R² = 0.22) but fit the data better than the Wood Stork model. The reduced model was significant (F = 3.28, P = 0.03, df = 40, R² = 0.21) and fit the data almost as well as the full model. A test of the least square means using the type III sums of squares indicated that fish density had a marginally significant (F = 3.76, P = 0.06) effect on intake rates whereas water depth was clearly not significant (F = 1.54, P = 0.23).

Great Blue Heron

Great Blue Herons were filmed in all three depth treatments but only when fish densities were low (fish quantile 1). Fish densities were lowest in the medium depth ponds ( = 0.22 fish/m², SD = 0.25, Range 0.04 - 0.39, n = 2) followed by the shallow depth ponds ( = 0.43 fish/m², SD = 0.35, Range 0.19 - 0.68, n = 2), and deep-water ponds, respectively ( = 0.73 fish/m², n = 1).

Mean intake rate was 0.16 fish/min (Table 2). Mean intake rate in the shallow ponds was similar to the medium depth ponds but lower than the deep-water ponds (Fig. 5).

Statistical tests were not conducted because of the small sample sizes for this species.

Fig. 5. Intake rate versus fish density at 3 water depths for Great Blue Herons during foraging experiments, 1996.

Great Egret

Great Egrets were filmed in all three depth treatments. Like the White Ibis, when these birds were filmed, fish densities in the shallow water ponds ( = 0.22 fish/m², SD = 0.31, Range 0 - 1.38, n = 32) were less than the medium depth ponds ( = 0.50 fish/m², SD = 0.37, Range 0.04 - 1.32, n = 116), which were less than the deep water ponds ( = 0.69 fish/m², SD = 0.31, Range 0.20 - 1.53, n = 69). Average fish densities at each depth were lower for Great Egrets than White Ibis.

Intake rates averaged 0.24 fish/min (Table 2). Average intake rates were highest in the deep-water ponds, followed by the shallow and medium depth ponds, respectively (Fig. 6). When averaged across all depths there was no difference in intake rates when fish densities were either low or high (Table 2). However, this trend was not consistent across depths. In the shallowest and deepest ponds, the highest intake rates occurred when fish densities were low rather than high.

Fig. 6. Intake rate versus fish density for Great Egrets at 3 water depths during foraging experiments, 1996.

The full statistical model was significant (F = 2.97, P = 0.01, df = 163, R² = 0.09) and contained a significant interaction between depth and fish density (P = 0.03). Thus a reduced model was not warranted. A plot of the interaction terms indicated that the relationship between intake rates and fish density was strongly positive in the shallow treatment, slightly positive in the medium depth treatment, and negative in the deep water treatment.

Next: Discussion & Literature Cited