Ecological Archives A016-070-A2

Noah G. Perlut, Allan M. Strong, Therese M. Donovan, and Neil J. Buckley. 2006. Grassland songbirds in a dynamic management landscape: behaviorial responses and management strategies. Ecological Applications 16:2235–2247.

Appendix B. Comparison between logistic-exposure and apparent nest success; discussion and interpretation of nest survival as a function of haying data model.

TABLE B1. Apparent nest success (ANS) was generally noticeably higher than logistic-exposure (L-E) nest success in the Champlain Valley (New York, Vermont) 2002–2005. The potential number of failed nests missed is the number of nests needed to attain the model nest success results given that the number of fledged nests was the true number of fledged nests.

SPECIES	TREATMENT	YEAR	FAILED	FLEDGED	TOTAL	ANS	ANS AVG	L-E nest success	L-E nest success AVG
Bobolink	GRAZED	2003	15	8	23	0.35	0.34	0.20	0.21
		2004	16	10	26	0.38		0.19
		2005	3	4	7	0.57		0.22
	EARLY	2002	8	0	8	0.00	0.02	0.06	0.05
		2003	10	0	10	0.00		0.06
		2004	20	0	20	0.00		0.04
		2005	2	1	3	0.33		0.04
	MIDDLE	2003	27	2	29	0.07	0.39	0.31	0.32
		2004	16	7	23	0.30		0.29
		2005	16	25	41	0.61		0.36
	LATE	2002	32	35	67	0.52	0.64	0.45	0.46
		2003	7	19	26	0.73		0.48
		2004	12	22	34	0.65		0.45
		2005	5	22	27	0.81		0.47
Savannah Sparrow	GRAZED	2002	4	2	6	0.33	0.46	0.22	0.25
		2003	13	15	28	0.54		0.26
		2004	29	20	49	0.41		0.24
		2005	15	16	31	0.52		0.28
	EARLY	2002	22	18	40	0.45	0.21	0.09	0.08
		2003	33	15	48	0.31		0.10
		2004	102	10	112	0.09		0.07
		2005	54	7	61	0.11		0.06
	MIDDLE	2003	6	21	27	0.78	0.66	0.44	0.45
		2004	11	14	25	0.56		0.41
		2005	8	12	20	0.60		0.50
	LATE	2002	9	17	26	0.65	0.55	0.39	0.40
		2003	16	25	41	0.61		0.41
		2004	20	28	48	0.58		0.39
		2005	16	6	22	0.27		0.41

Nest survival as a function of haying date

Nest survival was modeled as a function of the Z-transformed haying dates:

log-odds (y) = 2.6933+ 0.1782 (Z-transformed haying date)

(SE = 0.0248, 95% CI = 0.1297 – 0.2267).

This model fit much better than the intercept (null) model (Likelihood-ratio test: = 53.5, df = 1, P < 0.00001). The positive slope of this equation illustrated that each delayed day of hay harvest increased log-odds (nest success) by 0.1782. Each delayed day of hay harvest also increased daily nest survival with a slight curvilinear trend. Nest success increased by 7% when hay harvest was delayed from the earliest observed haying date, 27 May, to the average haying date, 15 June. Daily nest success increased 0.00037% /day during this period. Within the ‘early’ and ‘middle’ groups, the average ‘early’ cut occurred on 3 June and the average ‘middle’ cut occurred on 27 June. Nest success increased 9.5% between these dates. Delayed hay harvest continued to have a positive effect on the rate in which daily nest success increased until 20 July.

FIG. B1. Increase in nest survival and nest success with each daily delay in hay harvest for savannah sparrows and Bobolinks in the Champlain Valley (New York, Vermont) 2002–2005. Nest survival was modeled as a function of the Z-transformed haying dates: log-odds (y) = 2.7064 + 0.1763 (Z-transformed haying date) (SE = 0.0248, 95% CI = 0.1276 – 0.2250). The greatest increases in daily nest survival occurred from 28 May to 6 June (+0.0009). These data should be interpreted cautiously, however, because delaying the first haying data from 26 May to 6 June is a statistical construct that has no biological value; such a delay will still result in zero reproduction. Within this early-haying period, 100% of all observed nests failed due to hayfield management.

[Back to A016-070]