<%BANNER%>

Investigation of Tufted Titmouse (Baeolophus bicolor) Anti-Predator Vocalizations

xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101205_AAAAAS INGEST_TIME 2010-12-05T07:48:43Z PACKAGE UFE0017822_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 52388 DFID F20101205_AAALPD ORIGIN DEPOSITOR PATH hetrick_s_Page_020.pro GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
1a58194bf96ab277864e433d8cdf57a2
SHA-1
28f476447ae537c16d44346f8656ee82024e7514
74583 F20101205_AAALOO hetrick_s_Page_004.pro
5dcac111b7827c0acbd24ee01550b58c
a5635e92f39cab28f13a6c1719ff21e7a964c899
50431 F20101205_AAALPE hetrick_s_Page_021.pro
54ca3bc53e36acd97edd957d8adba8f9
0bd5624fdd263b635664b8ff1e06dbb8c11abd45
56629 F20101205_AAALOP hetrick_s_Page_005.pro
80a2de98dfcbfd321286036c05ad0acd
3b718db6aa752f003ca9db628e2b7d46f64578ed
53849 F20101205_AAALPF hetrick_s_Page_022.pro
53653ecbe273988c2bbc5461241cc13b
39fabe33e0059069f75e8f5772a870f4d190d4c3
51238 F20101205_AAALOQ hetrick_s_Page_006.pro
dfa7e2124bb1fa6502aef957c5d3e6d7
d3ef71d2da3cf9e26259caa7e51d3bd2f8102cf8
54508 F20101205_AAALPG hetrick_s_Page_023.pro
88177e7717c7d31478f42785d64f6bc9
99d16a3397d15147630a8a9081d9f9232334f95f
62234 F20101205_AAALOR hetrick_s_Page_007.pro
3f6e647b49f1e85ea5df21712b0be4e9
62e65fc6a990f0d82fb3188b08b89ab86bf364ce
53698 F20101205_AAALPH hetrick_s_Page_025.pro
ee8d9b232702a1e55328a3c669e19629
8759fbae2d3f4cffff24cae618e6e97c6e132542
27635 F20101205_AAALOS hetrick_s_Page_008.pro
cce111dd89755b0812dc4a1ae8f1ca71
804582e8684265c6f88d0805230f96dcea0242b1
57328 F20101205_AAALPI hetrick_s_Page_026.pro
8de13691bb02f389b60762183fee7973
ec7cf7d712fa5eb5e85e004ab7091cb0c9011faa
50460 F20101205_AAALOT hetrick_s_Page_009.pro
abe8468de1c2559eeb973248e4697846
08cd40c2e16c40370afee7eb901d8e28f963aeb7
57808 F20101205_AAALPJ hetrick_s_Page_028.pro
24f31de52bf42aa0f69986d183b9bf90
317b644babfdc2aeda45c8f610fb34a1117a8811
36843 F20101205_AAALOU hetrick_s_Page_010.pro
8cf5214a3094e093c135d0e98276d287
7d98a62d63f9e195be4acfaf04b74fecdb83c107
57904 F20101205_AAALPK hetrick_s_Page_029.pro
f432f8a26caf95ba63e3fa9dd6b547e7
cf29f3a84396e7a1d39b6a99a2e7b668b0875b4d
53087 F20101205_AAALOV hetrick_s_Page_011.pro
6f9b828912f6003c6d5f640fcb3daab3
e3274c9eebfaa62d680c7bc9027e1658c95cf3d3
55991 F20101205_AAALPL hetrick_s_Page_031.pro
ae8784cd92d5fc5dd11364ac6e5b6d03
c15edd8c0804b4d3511f57343f58ee519b31d0b5
47828 F20101205_AAALOW hetrick_s_Page_012.pro
c1bef86e4074a427df0502ff8ce0180f
2628e6c2557555d7f8b7bcea084895d42a083293
18407 F20101205_AAALPM hetrick_s_Page_032.pro
0a5e8b631985c2ca36321b28dae9eeec
e74a53c96eaee6c8e56041b02102fc9a589d5eb8
47203 F20101205_AAALOX hetrick_s_Page_013.pro
54fc9c6e3137957073b3ec569e94b5b3
f495c8e5f4d0b9b9989c789ff25ea22312d49c5a
55714 F20101205_AAALQA hetrick_s_Page_053.pro
6e0b83642d8e7e7ccec2acf85b72b998
a17c6226f768507bd9b28713e9b7b4794d083a9c
14315 F20101205_AAALPN hetrick_s_Page_035.pro
9c884e478302e19ef62ed60abb00a7d2
520dc9e9e47884e6d4ce70f77d9001270c0e01ce
50171 F20101205_AAALOY hetrick_s_Page_014.pro
2e264cd6ca0f026707dfc84b04901fc0
18f175778771a9989c53bc3a163aef4a42645638
53198 F20101205_AAALQB hetrick_s_Page_054.pro
be8ac02f67f100213118bcd968c81f70
bef13ebc56678932deef309dc2dac8bea31925f6
58563 F20101205_AAALOZ hetrick_s_Page_015.pro
ef8925383e9cf6822133a17325471274
f6a9830754e49b2ab65dd56f358dcc9c689007f3
54103 F20101205_AAALQC hetrick_s_Page_055.pro
7eabfaddc82fa7090f93d85c9e5ebf5f
c6eceeae9b5705da1ea59900193e6164d51eb568
8457 F20101205_AAALPO hetrick_s_Page_036.pro
5810da6fdb427935f57f277a15b9b261
0f7f882432b4dfceaa26932fb4eeda3f739d65d1
54526 F20101205_AAALQD hetrick_s_Page_056.pro
89ae36e8b3e1f380bd310ca7030b704a
4d9ea1628a98b77bd28c1086f34ea2d69fc7ef31
5942 F20101205_AAALPP hetrick_s_Page_037.pro
21a43801af6c31808787d05c63598a52
e5a014dbc2cdca43ca0e4f1200adf5cb9ee70979
51534 F20101205_AAALQE hetrick_s_Page_057.pro
36706b6588b71df20b21404343e39885
03732288b16c59fc6cc53936637f336f0793101e
14522 F20101205_AAALPQ hetrick_s_Page_038.pro
58027bdab245534e0d9e40dcabb839a1
65ad0f6ba012b281da3c289f5a40bc74924c7df6
53632 F20101205_AAALQF hetrick_s_Page_059.pro
a85be55e0e477c14be456026ac8a509a
476216127c7df2f9144da3a01f6fb0f547a3b6c3
9846 F20101205_AAALPR hetrick_s_Page_039.pro
f48a5b4dd7602ff30cc38708b6916a66
decb401dab1502bb0b95aaf1036a4c356a86cf73
54272 F20101205_AAALQG hetrick_s_Page_060.pro
423d45fa593e33d2f929cac181541ad4
7760b09c668ecd9c356dd942805b39e8a5b5cb2b
11383 F20101205_AAALPS hetrick_s_Page_040.pro
f3af5ba86b4bfc8b210df6ea6c6330bb
dd055ce23f38a796931880651a0e538cf449f551
46390 F20101205_AAALQH hetrick_s_Page_061.pro
5d19ddb7ea94947127aa175d322762e4
7393aed8d21db1aead27affaca5cdf1c7fbcf353
54187 F20101205_AAALPT hetrick_s_Page_042.pro
dadef561bd7c64f05d235fe67fae1942
6a2c77a0b35104bb2eb6156e2926fb6487d553c3
56311 F20101205_AAALQI hetrick_s_Page_062.pro
c50fa4ae9ee4e65cf023010bfa6c95cf
96bbaee57d82d520d478dfcc11c220f173b2f373
53774 F20101205_AAALPU hetrick_s_Page_043.pro
4dd870b8bec7c0b20e9762aa9e209f49
5a3e78c52e416cc30aa09e79fe7d9021bd260efb
52538 F20101205_AAALQJ hetrick_s_Page_063.pro
fd91c1166ab4f893d15ab633d690155e
f6e72930313ba59650ea05fa2aa6e6167ecf6613
54453 F20101205_AAALPV hetrick_s_Page_044.pro
1a5436c70ac06ae8a868510619f9bbf2
435c990303a089389d6cc1caace7d43d266214b3
57789 F20101205_AAALQK hetrick_s_Page_064.pro
93b72c396a854553eaaf6f3b73551644
da2b3d34e702a6f43c8e370b71350dc4ec713dd6
53501 F20101205_AAALPW hetrick_s_Page_046.pro
127f3bb94e9776f556b602fc86d3a8c3
0f9efd322536eae809165f3c05cd7aa5086f02b2
57065 F20101205_AAALQL hetrick_s_Page_065.pro
b250fcc8b83d2fc2eaf707450dd45325
b875f8b9aafe16594b64d558a5d8903022598f45
56818 F20101205_AAALPX hetrick_s_Page_049.pro
9f70766f08dbc85364a1a04cc19c1b78
8383740daff273254a56d879e2c4c33d685e4eb3
10988 F20101205_AAALRA hetrick_s_Page_084.pro
213867654ad7677651cca4e90d874f1c
1aa1d1c32eb6ee4defb699fdf458c4cc123e14b8
54346 F20101205_AAALQM hetrick_s_Page_067.pro
ed7a5952fca0350b43265f6b2c23e5de
31b69057916d3acec070b92f887865bc868e0b92
54247 F20101205_AAALPY hetrick_s_Page_051.pro
12471ef6dce60b4a35269c6bf6bf18d7
af9e8ef8f55f93eafe373247cd5e9234728f52ec
12903 F20101205_AAALRB hetrick_s_Page_085.pro
035b22bf920fb9b330717054dd136d00
de0395eef4daed4535f4c073fe3de8e13ba98bd9
59062 F20101205_AAALQN hetrick_s_Page_068.pro
c4eb5bdca75bf73f0bebf43a390e33b5
7696a51afa627b023ea852d504e4a778e6378382
54821 F20101205_AAALPZ hetrick_s_Page_052.pro
53b336fc046aee7f6f1e3f1c13aa9f11
20a56e22110746fe60ee83a420e500adb6c5d4ad
14175 F20101205_AAALRC hetrick_s_Page_086.pro
965c1928e2bbdb72f87c4aa67ec71057
e45ac1eeeee8afc6c17da2b5829d62ee4c37f3c2
53203 F20101205_AAALQO hetrick_s_Page_069.pro
c8f1d481f62e6d9d2d93a445aa8003f3
8968d8832a42a43880328605d365978890f89738
35410 F20101205_AAALRD hetrick_s_Page_087.pro
1d8d922672b086d55c4d105b4b1213ab
036a63c0e86a37106c0d4e9963eae1b7307cd3f6
35060 F20101205_AAALRE hetrick_s_Page_088.pro
6abbc4a95714bd5c950139c2280f8cd6
2dcf6b6fc704fde74e0f6bab0d68db14a3ac188a
58149 F20101205_AAALQP hetrick_s_Page_070.pro
79bff7c6087d8a1ab5757d6f16182c57
e8853c24e8d2f13379e8d2ae8464f5923ffe5aa2
50039 F20101205_AAALRF hetrick_s_Page_089.pro
6464ab665e7a6e9fa2a11723c5fab59d
2b5e806be389c74195acf2ca3c1e3aa8698d7622
55732 F20101205_AAALQQ hetrick_s_Page_071.pro
0a577ebf40e65b177ec4153736e23870
d2767caa7dbe6dedeb9f3cbfa80e0ce017a20838
51804 F20101205_AAALRG hetrick_s_Page_090.pro
df157616a473db0b6ae21876542fbfa6
8210747c68b03ec4edbf2b82e2f33c9bd736ef57
57300 F20101205_AAALQR hetrick_s_Page_072.pro
9978efe8b7372e635e12b8e9ddd04559
718937b7deff7d85e6ba9a52708cc43d1868d8ed
51729 F20101205_AAALRH hetrick_s_Page_091.pro
1af762b243fa871868837798e0a2b38e
73004cc88a4578800a14f2a14f65facf4e25b74c
13950 F20101205_AAALQS hetrick_s_Page_074.pro
9571385e5d1f7304662b317b56efd64b
240734f03522812384da5461697bd06fc93a8692
47609 F20101205_AAALRI hetrick_s_Page_092.pro
922d7a19049951026df58be3e549879b
d4acd6ebbad793b2fcd4ccc7f34ef0eefadcc344
10687 F20101205_AAALQT hetrick_s_Page_075.pro
1cd792c206ddbecebad64af53359638c
63af2a50a44bd4ae02ddc8fd685d86933f94d44f
20227 F20101205_AAALRJ hetrick_s_Page_095.pro
e543ff4e9b07e09eb04385057ae61d2d
113425d6a956b6877e0302bf2c13ffdb1be5e75b
9794 F20101205_AAALQU hetrick_s_Page_077.pro
a02d6d564359bdb5c1e3d702b2feb932
140f55eea412fe3ef6c44dc874627cb87abf729e
3981 F20101205_AAALRK hetrick_s_Page_096.pro
dc36c81eabbfe8eb11e693fce3e46a92
7e57fa44c94158b0c0d3a56190e496ac543247ba
15253 F20101205_AAALQV hetrick_s_Page_078.pro
ba799416df23fca696b7c2bd41e4e138
ff232f33d0f6dae79e86435a4c340b8d5395ef3c
19862 F20101205_AAALRL hetrick_s_Page_097.pro
816f98db2fc37efff7af1d1ef5e86091
77bfd36e44336b492125db5ae5f2c736466057d9
10079 F20101205_AAALQW hetrick_s_Page_079.pro
73b544f5bd0d4ee24770c81f4792a3c8
5a8a2ca88b2cd9d4d53e338cabd9b841e140ec47
60430 F20101205_AAALSA hetrick_s_Page_115.pro
06ac9607d4255ef27207333048efdb44
801e0f39ea3828fe60fbec7c3c27c5cfae26a304
63659 F20101205_AAALRM hetrick_s_Page_098.pro
4c58062b5ab5fbda9106baff63388cc0
46a0a32b8dc929df42155fac5244c0b785d49b97
10589 F20101205_AAALQX hetrick_s_Page_080.pro
92f5eab9e813643014c70716914c5b7c
652865ee0614f2bfcbd1f070593f357c6cef17de
44322 F20101205_AAALSB hetrick_s_Page_117.pro
934f80d79cec12cdd33f6e7adf83d475
aa9fd2c39f3da493e89d5b55b243b7f646b3927c
11242 F20101205_AAALRN hetrick_s_Page_099.pro
09f1bfb445c141cab261909bd72518a9
ff6af233117f29a8134ff8b86cb046f4072e90fc
7912 F20101205_AAALQY hetrick_s_Page_081.pro
1f88c0fbb6f302c0c62f5c4d0307a793
f5942d44beb2daead901013efb43660b303c5870
1469 F20101205_AAALSC hetrick_s_Page_003.txt
de5d1d559af5e3409988c807aed347a7
df8d553cc693294a2109f3d2484c1d233e4a7c84
27124 F20101205_AAALRO hetrick_s_Page_100.pro
70dc96c8ec6a1d69871cab92177f2821
895676ad148ad3d3780c336441192794dac8912d
14090 F20101205_AAALQZ hetrick_s_Page_082.pro
bc15445062bcdeb0f56b141a48dcc6f8
0b6a95280a1c52ac35b1074a67ac4778c053d6f3
3240 F20101205_AAALSD hetrick_s_Page_004.txt
39724b7ccad975e0fb8129883424d13f
1bc47755d3e270d3e31188c6b9d9be7d449e63d2
38756 F20101205_AAALRP hetrick_s_Page_101.pro
6e95aed985fed6cd395d19df6cc88a36
22b9a15a2d2d1b98045c1e8e61afcea61fe112f4
2168 F20101205_AAALSE hetrick_s_Page_006.txt
551121684bb15a4fca7e67a65e8b22d3
fc46dced1b66a6850a48ffccea29a22900641bef
24978 F20101205_AAALRQ hetrick_s_Page_102.pro
f3d3826be46ac8363c81424dc6451456
7474611f9c499cf33cbc17371a0365b323bf971b
2581 F20101205_AAALSF hetrick_s_Page_007.txt
72b9d54acc7561309142cf52a74a3929
53ca32a2745d61d60968c2efb3b828c3a7c65c66
25503 F20101205_AAALRR hetrick_s_Page_103.pro
96be14ad06d85bfabca75fa4f3e5d396
f27071516ff0f47a8fdf2de342cc3a2bc9544b17
1158 F20101205_AAALSG hetrick_s_Page_008.txt
b75448a549e574ce7d0b2d329ab04180
7077275efcb1015411ea5a7d71ddfb8b76e0566c
25372 F20101205_AAALRS hetrick_s_Page_104.pro
0ee3ba006c1c973da46bcb510b234851
d89e339b222c5a4739b4a5b4621009ca1219cbef
2195 F20101205_AAALSH hetrick_s_Page_009.txt
707fe8a0b4a12c08b4b416c2598ad062
34c7ca0cdbc53daf04bbc361a3230480dfee5619
22619 F20101205_AAALRT hetrick_s_Page_105.pro
6f197774cc73a6bb29f46460b66d105c
f2e3238460c3dbb18d0db7e29fac556fe34f4ed0
1457 F20101205_AAALSI hetrick_s_Page_010.txt
6f60d6bf2236ee7fb2e1a457cbddafc1
c1b0fa24b3fc22c931dbfed5e6dca0640c47370f
23992 F20101205_AAALRU hetrick_s_Page_107.pro
cd4e4670aa83dad7846f31a2ea3ed586
baf96e32c1fd6a6a921997f5cc19c1a89c35e967
2167 F20101205_AAALSJ hetrick_s_Page_011.txt
fab73c0c8793d9999f8689fa4bb61965
148094b194957b2839cc1c793327d99d4490dc1d
9540 F20101205_AAALRV hetrick_s_Page_108.pro
9efa9656d2a844b876ca331b90f8351e
5f736f3c25c0c7ec09ac73148349b1d3723ee0d7
1897 F20101205_AAALSK hetrick_s_Page_012.txt
64272808fc735a483caacb58ebdf4eee
ac80de08c065d2bcd3ef6fb1c73a352e76e1b8f3
57139 F20101205_AAALRW hetrick_s_Page_109.pro
c9549a96c32010822a13521e196f4c5a
02df970092a537e3dcc3ae833f04a63cb2ac33fb
1923 F20101205_AAALSL hetrick_s_Page_013.txt
0022c2df33abf6e49d770957321fcbd3
0560670507019cab7e53a6b7dd648e02096e2652
61231 F20101205_AAALRX hetrick_s_Page_111.pro
b1a2b60fe46092c7932bdf6d34dd41f8
3fff6a74c48cfa6023ee097b5f959422a6c65f2f
2265 F20101205_AAALTA hetrick_s_Page_029.txt
b03539e7d7a5d6330015fdfa39679bd6
95478ca808d1f170f8c654fc03db507d2df88530
2035 F20101205_AAALSM hetrick_s_Page_014.txt
a8ef8285495b76bc88b12e5766e267c0
d390fe2c2ee997d77bd0bb46565af938868ebf91
57540 F20101205_AAALRY hetrick_s_Page_113.pro
3c352f4ae44359d7b3564f6fe2ce37f1
86155f592c96126fff3b033b0ff3be88ca8e9651
733 F20101205_AAALTB hetrick_s_Page_032.txt
e7553f161c4239114c02c0351a5906fa
13f42ec014ae2cc00d9d09e64a5b93e807827f44
2293 F20101205_AAALSN hetrick_s_Page_015.txt
ee252bfc2e0838bcabd3ad62a1430f3f
53821757de4215bb5e6839a93ab300b9fb2883d2
57643 F20101205_AAALRZ hetrick_s_Page_114.pro
1c0c936dd00a34a8bb6e6ef2e5858ce9
c17d7ac6256bf15a1fffc5e6cdff38377c58d663
253 F20101205_AAALTC hetrick_s_Page_033.txt
88eb044f00b2fdaafc202b38ba477c20
ab199a648b9768a985a4a5908fd2051d2d666b6a
2259 F20101205_AAALSO hetrick_s_Page_016.txt
547a4fd52de842bc682057fc6a3169e0
095295cc11c6d7f13b8a81ea942732fefd1926a0
780 F20101205_AAALTD hetrick_s_Page_035.txt
7edaf62d49503d1fbec3072e8e2ae17c
fcbd454b1b76fc05703be6a1bd5268dd2cc790a6
2042 F20101205_AAALSP hetrick_s_Page_017.txt
191fcad6987e1b3af34a0129e8ce739c
711bf0819f05a3eec7c93757e685935e1b60980c
501 F20101205_AAALTE hetrick_s_Page_036.txt
fe63fa77a920eb0e219655a6964e2e63
d3854a8341eeb487029dfaea0cc1d55c2a55fda2
F20101205_AAALSQ hetrick_s_Page_018.txt
9eaeac98a6e58c9bcba6c2e6e273d717
a06ca28b993777fbd31de8c5f58bc79efaa7922b
532 F20101205_AAALTF hetrick_s_Page_037.txt
7d94a25e5e9e48383b17a198afaf5534
5461fe8816f208bd56e118a4e572c02b3a2eb6d7
900 F20101205_AAALTG hetrick_s_Page_038.txt
7381fe7063a0972a2d6825e8b9b5178b
f44dcc9b6687a427beb0998d934b3b4a09dc208c
2150 F20101205_AAALSR hetrick_s_Page_019.txt
307e3f3971aadce7038f768d4d72214b
e2e7100a1fdb7009a8cb41f3255cd5063adc5ab1
649 F20101205_AAALTH hetrick_s_Page_040.txt
78e0a23fad01ccc9dfb1637a9222333e
3f0a853c4c5fd141a422af57899478823073f3ec
2113 F20101205_AAALSS hetrick_s_Page_020.txt
43144aac6cb7998d0b8923c20e248acc
1727c4adf211aedf38ca816c6b26fad13793ec7f
1204 F20101205_AAALTI hetrick_s_Page_041.txt
11c60936769b407fb7dbb8eee8eeb00f
6b4cd2bb9caf65d9f560931a34eedcd11149a2a5
1992 F20101205_AAALST hetrick_s_Page_021.txt
499b9e272fa05b51f6950be394497fc9
46977f3223e25ea3073e9c12c697cfa0df047b12
2177 F20101205_AAALTJ hetrick_s_Page_042.txt
f47c7b8e474f56cbef1bd0a64f9faaf2
9b4c428a77f1f511bae016ab0aa9b62d18c18d86
2114 F20101205_AAALSU hetrick_s_Page_022.txt
864f113a81f2f75a0f95bef2fc750259
5a9be8c6ae9e120a5ddd172b6d511945bad6e8df
2134 F20101205_AAALTK hetrick_s_Page_044.txt
35f96854a3882c8339cabda11d1be9fb
499c99656d1633523d68108a3964b1c40a4e64cc
2149 F20101205_AAALSV hetrick_s_Page_023.txt
27510f5a15258164fcdf129d61e0bd75
6832b8c0c2d4f6a385ed2ded4137163343ebf610
2155 F20101205_AAALTL hetrick_s_Page_045.txt
7c3e79251daade8e333489b0249a9c83
65701e3de94637a84f6a08633def8bad8af2a591
2120 F20101205_AAALSW hetrick_s_Page_025.txt
da53cd2e5062f52a244f40867934264a
1061f5d0b488fe9a1c84bf03f261bd8ee703870d
2198 F20101205_AAALUA hetrick_s_Page_066.txt
6aeefef7ea8eab2d1239cdbc5340bed0
c3fa7873dd6c8bdca6ae970370174e85cb48cf28
2148 F20101205_AAALTM hetrick_s_Page_048.txt
dd8589361417bbffc54369fc1fe9e936
95a6514048b3eee195b78f7a75225eb51bc8c37e
2284 F20101205_AAALSX hetrick_s_Page_026.txt
b83ed67addcbdd4db8397df39f8e5ce6
bd0546383a587fb4a34380dc5b5d61a4ce84235b
F20101205_AAALUB hetrick_s_Page_067.txt
c151b85a7c02e6460d7b8aa048e959b0
a7007f3b3fb54734f0f94fe9218f28741749577c
2223 F20101205_AAALTN hetrick_s_Page_049.txt
839ecdfa2bbead6a5a3e94052410fe0e
cf303faf27f9ee2fcd3cf95dfa7c2b396ac1d1bd
2214 F20101205_AAALSY hetrick_s_Page_027.txt
3d73c7aafb51d7c21480b9b4b7e362b4
c3a81a2286cab1636cdf9cacdc36c5be5bb4f48c
2278 F20101205_AAALUC hetrick_s_Page_070.txt
014b7619328fb451fe51ff21b6a8e197
9fa9a7ee1e4fb4e7221539c9a216c99f3f471455
2010 F20101205_AAALTO hetrick_s_Page_050.txt
781d0c34344eee5245b1faf2c9548bf7
b802a5c7b1895e9df43f863fb74a3e639714b2d2
F20101205_AAALSZ hetrick_s_Page_028.txt
ce157a27f15a196726c81d107fedd5ef
2e5d614b6bf4c90c20a13a6b07747a37fda572c2
2181 F20101205_AAALUD hetrick_s_Page_071.txt
27cd745290f4834897e0e98b47b9201d
2f5f2cfdc572c362c244e05f041b94dd77fe12f8
2159 F20101205_AAALTP hetrick_s_Page_051.txt
e8fa0a7f583c2b202c3bf282a875e31f
478029eca87b2d8ee8ea8fc9a96ac26a44002166
441 F20101205_AAALUE hetrick_s_Page_073.txt
8785dbc3bc7ed0859c1fc503c207ecb6
9b3e93e7eb158d987244689dfcf41977fb41dbd8
F20101205_AAALTQ hetrick_s_Page_052.txt
3d8f8cc749b804d17a818ded5f6f90d1
06bed50bcb5fff4318a8fac500b56abff1e673cc
709 F20101205_AAALUF hetrick_s_Page_074.txt
8d70a7c6cf4f8a2c9599946a23d4a6a6
52cd53e072b8881448dad0739484a02cb83dbe26
2187 F20101205_AAALTR hetrick_s_Page_053.txt
f23153a3de1e2aa6a2e3f436625487b7
6f428ddd1988baf8ba18ed1548a7dd2ed48c8b73
615 F20101205_AAALUG hetrick_s_Page_075.txt
2cef13596fa1b3647597283e4fcda7fa
38b4b912f8a54e8b52c6faf26d9b38b7fc5727d7
6343 F20101205_AAAMAA hetrick_s_Page_090thm.jpg
233565765d7e9bb9203c9712efb0f215
29acaf13284b89ad46552566f01c8739a1540612
463 F20101205_AAALUH hetrick_s_Page_076.txt
6bf35ca5a7357752fda3bab0c02c3095
a9819fb908be0cff0f38f5a5011148b545bb43bd
2128 F20101205_AAALTS hetrick_s_Page_055.txt
f859a9bf3c5103d918f201da7020f22d
70e91aa38aa38ef05e6a41f39491e5877856016e
25156 F20101205_AAAMAB hetrick_s_Page_009.QC.jpg
2c670ba52249b663d00c3cd6ec04d1fc
f2f7a6664d736fe2b1c6f342ba2b759ceddc81c7
526 F20101205_AAALUI hetrick_s_Page_077.txt
95c5affabe7b2314ef3aee7907478153
1d447c0c24289d9fd661b5e8958402d18affa112
2194 F20101205_AAALTT hetrick_s_Page_056.txt
890029bae878ddf3c1311a0e0ce02c7c
33f732c049effbc6b517af4eae6e8c5e9ff719f1
24542 F20101205_AAAMAC hetrick_s_Page_012.QC.jpg
97ad08582f547a0b8eba1195eb62028e
336b8047e09df260e705bdd9cea984d117b1b487
544 F20101205_AAALUJ hetrick_s_Page_079.txt
1d1c989477a57695044fbbb9f3cb089d
4a99a480c2637ccf6e9a12e06213d1b44acc5033
2040 F20101205_AAALTU hetrick_s_Page_057.txt
708209a97960cd2f5ac4d0960938ff0d
07dc3b5c9618c6814532a57e25828ccc92e0ce60
2372 F20101205_AAAMAD hetrick_s_Page_083thm.jpg
e5e1f12044b6ac186f384f60ec71a186
21fc758c49db5daff9892ce1434920c12db825eb
494 F20101205_AAALUK hetrick_s_Page_080.txt
25e48bb3e5975e9e6651151b7255a4b8
cf02beb9e4ac5c76d1df6080176601a483c3bc6e
2106 F20101205_AAALTV hetrick_s_Page_059.txt
f0c9bdec8023a05ab9f2b0f4f499bbe8
53a93f740fb96442a4193f422db186213cc0c20a
5242 F20101205_AAAMAE hetrick_s_Page_034thm.jpg
7bb37d6a9005e973134174b66c7b7394
8ff961799fffcb35e3a7e713550beaaa21005771
442 F20101205_AAALUL hetrick_s_Page_081.txt
69b81780f6a848bd919873bb03b7934a
b4ba1b23720b7ef1899529d0037d986d8d10946a
F20101205_AAALTW hetrick_s_Page_060.txt
12b8dd0a128b1248b88a6a79a237a217
abb65fd6ea7ebcb74a0c2040be677e1feb93f508
27038 F20101205_AAAMAF hetrick_s_Page_113.QC.jpg
c83c75bbc0c1b9937e4241e58b029b53
4b72aed88fae62d89959ea08046eb047f56ffede
875 F20101205_AAALUM hetrick_s_Page_082.txt
f11b0a613a8d79bb3b39dc99c9e5d83f
aceb050b9fcaebc0c4a6d2c2ddfe22b0952069e6
1845 F20101205_AAALTX hetrick_s_Page_061.txt
a1cc0ced472a59db63de219777c1a602
dd82c8cd46ad5630e24a549b677f69dd1c2ead8f
10651 F20101205_AAAMAG hetrick_s_Page_032.QC.jpg
386e22a3bdff168351868ee79e37301b
659792ada64a552499c129d14a63ee888330f729
2942 F20101205_AAALVA hetrick_s_Page_098.txt
b6ef145c6614870a283f6c6cd9fc4455
7cdd77bdc98e95a633affbc253d57da9ef2641b0
745 F20101205_AAALUN hetrick_s_Page_083.txt
93ab0e89d7cb8c323a9ab2a565440b80
36f830cbf630c7249e9fa90a86812bbebf76dae1
2247 F20101205_AAALTY hetrick_s_Page_062.txt
c50fbe97adbde4d90b45514ba6c508dc
8e97ab44ee50a35f76faae76598c5acccf3ce6f9
27262 F20101205_AAAMAH hetrick_s_Page_019.QC.jpg
b6fdd8f03e8704b369c10f973ba2a58c
a331b1e2363f2cc96aba366404fcb12ef55ac175
747 F20101205_AAALVB hetrick_s_Page_099.txt
2ffb2d97e0a2695ae2368f8748521464
689648a56e40f90ff2b10262b51ed6e930630df1
579 F20101205_AAALUO hetrick_s_Page_084.txt
fc0a757e50b62783067a213f4eb60f9a
665ca899e73356297859c83bb1c8c4085de41177
2071 F20101205_AAALTZ hetrick_s_Page_063.txt
8e87c9d5a4b2f58368d41b3a05833171
1cc414d69669230f29327825453e4f290050dec3
5841 F20101205_AAAMAI hetrick_s_Page_103thm.jpg
560322d69b41118a44daf6381a45ed25
f79a9ceb8ee3bdf4abedf484d164b68f5aa0b9cb
1824 F20101205_AAALVC hetrick_s_Page_100.txt
7e9b5c97cd8cb2f358b1cb0481bb341e
74bec4de0b010ec20957187279bf2cd578f7bfc7
941 F20101205_AAALUP hetrick_s_Page_085.txt
c629d4a9f4a901d79bc8a2addf945db2
96935afc955974563c0d1e31839943effc0f2f48
2793 F20101205_AAAMAJ hetrick_s_Page_084thm.jpg
931761c27251fcbd73b2015a1849c325
838dfa264678b740d888d2fbdffd079896e44be7
2590 F20101205_AAALVD hetrick_s_Page_101.txt
d3623661e2eba49d02ce2a4c2e924280
cc975624aa916a3554fb2abbdccd88f2da8e5486
1502 F20101205_AAALUQ hetrick_s_Page_087.txt
fc44f118a35496845d202483b1401452
c85c26b36509c591b735f8f9f941bce18f8f12a0
1748 F20101205_AAAMAK hetrick_s_Page_037thm.jpg
f07b47686011d92208bf8ab9018152e5
a95423a0589522b3bf3f504f25f2852ffe294d3d
1319 F20101205_AAALVE hetrick_s_Page_102.txt
259e1172954e3421b324bbd887754dc5
0e64aebedd82ddeb17481690b4c8ed0edc4f7049
2039 F20101205_AAALUR hetrick_s_Page_089.txt
5fe77a216fab99ddb486f6c97ee6d9b0
05f183fcdb957ff125db52c61f88ce982cc9dcb2
7028 F20101205_AAAMAL hetrick_s_Page_065thm.jpg
26e017eedaa3541254cd84b89a377ffe
ef83211a150835c70c2c4e5322ec5c0edde8c794
1311 F20101205_AAALVF hetrick_s_Page_103.txt
b7bf47e5b6bd3e017b26f262b42b7d83
ea9922ece871c3afa961c5b19d5f0690996a0574
F20101205_AAALUS hetrick_s_Page_090.txt
0563adbcc0f4457e5d8ec78273c8657b
f06566e26ef4d6f86b7533870d3da6b7e39b8a4f
7166 F20101205_AAAMBA hetrick_s_Page_081.QC.jpg
83aab29dc07070f22c39fb41e2e56aa0
2665cda171a17886cca3d1baa662ef51a9de0448
28848 F20101205_AAAMAM hetrick_s_Page_066.QC.jpg
c8c0b3c459176c67facb22c085b2b6c5
1d07f1d4e2a028e799e5b233cb1e3f396ad46ad0
1269 F20101205_AAALVG hetrick_s_Page_104.txt
df0385a8d0c8614efe3bc2aabbb707f4
8c3db63a1153e978ac03440c523e8f8e3926c72f
29746 F20101205_AAAMBB hetrick_s_Page_115.QC.jpg
1cd2ad5c2e9e893b3ccd74dc1c5b2816
04f88760a75c03ee0b5c1c195a3780a588d64ef7
25093 F20101205_AAAMAN hetrick_s_Page_013.QC.jpg
c4fef32ad544a171244c1fffc25f5278
31bb84b2279f65c49236d654b407accf12670e0a
1159 F20101205_AAALVH hetrick_s_Page_105.txt
068ff46b89b4a10b735ac4f2b2db2e28
50f45097604743d476d1a9380a4ddbbaadc44be0
2030 F20101205_AAALUT hetrick_s_Page_091.txt
9fc18e10e35d5c7d132386b55a9e4b4d
8d47e232e75c3333ebf45f08d3f37f03eb4096af
3144 F20101205_AAAMBC hetrick_s_Page_008thm.jpg
2c320ba9814317130d93e12558024726
30b94eadfa917cec316262a9f9266f55a94344f1
28809 F20101205_AAAMAO hetrick_s_Page_051.QC.jpg
2395f472777bc0c542bd94de44ecc088
02763315b552c4ed1a53199ea3ca493c51fdc01c
1477 F20101205_AAALVI hetrick_s_Page_106.txt
92e928dc2fde60a29d093b5eb33a2ec0
bccb00e1d06c63073a3a4dbbbd7f7644146b1ad4
1881 F20101205_AAALUU hetrick_s_Page_092.txt
ddb7e200ecc05e7cf378b8e7f2f2d4d3
14f30b26545941ec01e6a616567341f6320b7733
29236 F20101205_AAAMBD hetrick_s_Page_052.QC.jpg
e60195e72ea9a4697507a1e61cb5c49e
4d28076f9e7cba6b8f58abb964e2315aea957aaa
28983 F20101205_AAAMAP hetrick_s_Page_049.QC.jpg
0de9e5d822e7532a2958a883ca47c4ec
88ca7f93faa52fe232b0f9d6f71bf3da6ac385dd
1432 F20101205_AAALVJ hetrick_s_Page_107.txt
c1e22cf24e9eddac7bd98bbf82b6fcef
b9054e25fc78bc19706cda7c09453e56bded73ae
25578 F20101205_AAAMBE hetrick_s_Page_061.QC.jpg
0545a9870121307f71e4e0c32a5a8107
ead6676e6cca513c8ea6f4c12bc97544066bdb32
2096 F20101205_AAALUV hetrick_s_Page_093.txt
247721773f0e76e968549dea3b171aba
bca4d266277edd249fda99761991b9cd2557053e
27166 F20101205_AAAMAQ hetrick_s_Page_069.QC.jpg
ed0e75613a770f511f51413622fc8541
7775f8a77780902a7ff31f1b710a72efcb943e91
F20101205_AAALVK hetrick_s_Page_108.txt
218f2bb03d120f769d97f7c221a800d3
e236c288f3ae7dfd550a355970f1f50d13766af7
28584 F20101205_AAAMBF hetrick_s_Page_109.QC.jpg
016108ed3be593c260c3255680892cab
bd679771373a2f83210df2e5ddbc896b233c1cbb
244 F20101205_AAALUW hetrick_s_Page_094.txt
185a3068e7bfce46314cc7b39293bc39
e5baa07c85976ca063455abb02d07748b8e9c572
6072 F20101205_AAAMAR hetrick_s_Page_092thm.jpg
0cdfcaee7100a5f419f44ab0f5deb1dc
96ddf76311ec53982967c909ab69bcb20893d096
2351 F20101205_AAALVL hetrick_s_Page_110.txt
bf8e9f4988c226f17d09dc66cc2aa6e6
eb50c5d294da673dbcf1a11f55e0002c6fbb3c0e
6593 F20101205_AAAMBG hetrick_s_Page_113thm.jpg
ab0b5ab0a52b7f2e53689be9b660e3bf
95ccae5fe9dff6b8f11e54e9732fbf093167c450
930 F20101205_AAALUX hetrick_s_Page_095.txt
77145a1bfbc523a8f9108b6418981af2
730eb377588f9bdc78a919e25dc08973f853d2ce
22006 F20101205_AAALWA hetrick_s_Page_107.QC.jpg
b90775ee6760bdd9bc47ec5793b62eb8
8c59fab5145333782b193e1a0537c8b1be09ba5e
5480 F20101205_AAAMAS hetrick_s_Page_104thm.jpg
d7aa45ce9d562dcc8c41217a7abf68d6
005326d8b5cac89006023767a66ad4aafeabfb43
2468 F20101205_AAALVM hetrick_s_Page_111.txt
6f466743a7d3b8d13142176f6d75073f
369a09e2a8c33eeff11ffd453e1cc1eb1dffd339
23533 F20101205_AAAMBH hetrick_s_Page_117.QC.jpg
76e716582555abd49a1580c3bf66b69d
7183330dccae277a0c81fa3a3e1375d5b979e94e
186 F20101205_AAALUY hetrick_s_Page_096.txt
868dbbb9ebda79128362e2a78ca7922d
e8ddff3791bcae5e533399f461a9089ddaf45874
27187 F20101205_AAALWB hetrick_s_Page_063.QC.jpg
945840b2450fac3a68d016f37fc15366
5c9420cdbe71439d14edaa242bd4b29b54515090
6918 F20101205_AAAMAT hetrick_s_Page_064thm.jpg
853e0455784a74a0565732eef07706f3
89d7a333818220c07f1fd5154741d617967a6dc0
2204 F20101205_AAALVN hetrick_s_Page_112.txt
d2003fa4457693ec509279c8a9c43e7b
0b225ffa09f06b16e183221bf5833c2d630b9367
1190 F20101205_AAAMBI hetrick_s_Page_094thm.jpg
bd290dbd9d4b31966db976db8f643cbf
9f7e65e9ff0ad98f1dd09eb13fc0c80ba1048203
1051 F20101205_AAALUZ hetrick_s_Page_097.txt
0750b985a112cfd483b0db45283f9156
28dbfa14932d1babf1075e3aedb91886551b0eeb
3858 F20101205_AAALWC hetrick_s_Page_035thm.jpg
57b13ae0af71fedfa42db0bd90b10138
f00786f0a57712fc9cdff5d3e2c902c4f578c0f3
10786 F20101205_AAAMAU hetrick_s_Page_038.QC.jpg
5a591f162c02a25b9e7e504df1f9cea0
1d13f75caced68e8cb95658d36ef95ee36150cda
2333 F20101205_AAALVO hetrick_s_Page_113.txt
960a0131decf974cbce69f25006e58c2
915bb47a7bc297b37439a2cf13cb664c52415c8c
4013 F20101205_AAAMBJ hetrick_s_Page_096.QC.jpg
7bf9cca044f20c4c473cc2e0c1cf5caa
849ba136b124ad78bd8c257fb0ba87633a3aed9e
26563 F20101205_AAALWD hetrick_s_Page_011.QC.jpg
cffb94d5891444d6078550ec49c59b36
50e1a2ae9235ec36d3b742e018889c60c7218c9b
6937 F20101205_AAAMAV hetrick_s_Page_029thm.jpg
083c9fbb2f36dc2516869d22c7cb2b31
15d3f85bb056a00a29f473e33bd213f9178b8656
2328 F20101205_AAALVP hetrick_s_Page_114.txt
ac7217fe245e05ff36fd2cbbf7699188
fb6654f0bddad1462d96aae13c3a3478157223f2
20885 F20101205_AAAMBK hetrick_s_Page_005.QC.jpg
62831ff3a5ce6b57d1b2b3131c7eb759
5b5cf74a4d810437e6ca883e30639bde0b59093a
6912 F20101205_AAALWE hetrick_s_Page_111thm.jpg
838ab7b9a7d5ee43dc62a9b299b2aaae
9249da5901b5c481c51a64aef3810ca204e5025f
6207 F20101205_AAAMAW hetrick_s_Page_089thm.jpg
35395f7b081b19363d47337079134bdf
7c3b74fda366ced425d8d82b0f8caeb4202699ab
2438 F20101205_AAALVQ hetrick_s_Page_115.txt
5a1a4b2ba74d15ba0ec849f79859c5fd
64a56c0f9fb7af6c520edb9fd2d8ee7380dadada
29245 F20101205_AAAMBL hetrick_s_Page_029.QC.jpg
c27ac724c8ab9957251df8f7cbfa9fef
476ae6a72739c03811e9cfb438ca1539b4316042
7134 F20101205_AAALWF hetrick_s_Page_075thm.jpg
7b501d63601adf30aeded0cbaf309ecc
10d65a13db53e7e90004dfd9174cb3cbd411629c
6404 F20101205_AAAMAX hetrick_s_Page_091thm.jpg
d6c44d1a064717edb157ff0bbd7171ef
a4dcd61bbae2971138d623846a8c9601c6189803
907 F20101205_AAALVR hetrick_s_Page_116.txt
adaee03b9519524e31a7e60d8596a4ce
77a57c7a43d5cdefb440e365a1cfc263777eae13
6451 F20101205_AAAMCA hetrick_s_Page_057thm.jpg
5e8c3162d2398976757d2e9dd96d3a5f
eaaf484d98301a7e92cc1f8a9ce810ac1474d542
28835 F20101205_AAAMBM hetrick_s_Page_023.QC.jpg
b73e6e4c81ad323b64e749aa78618b4f
96039955240518d08e831ace139a4854b16f0407
21373 F20101205_AAALWG hetrick_s_Page_004.QC.jpg
587b151ec4219de6fe7acc798ace9b90
67810a50847cc7422e9837da0db0a5e400d8eee4
6223 F20101205_AAAMAY hetrick_s_Page_014thm.jpg
80ba8d763286058c3193ad7ae6253aee
d59dd360495c2354e147a200302425354471563e
1800 F20101205_AAALVS hetrick_s_Page_117.txt
03f0e9e8106ccd5d64bbbac4247372c2
379baac4168134cf5ae299baca6261384938dab4
28476 F20101205_AAAMCB hetrick_s_Page_075.QC.jpg
2a68d886982fc46d83778c83cd2cc8c1
35e4c8c07b821a44d49d78662f9b7ee99eec1475
6849 F20101205_AAAMBN hetrick_s_Page_054thm.jpg
eac24b6c4aca80512aa08ba5a052004c
bcde328681a20b402dd349cd83ef1cf840cf5b7e
10160 F20101205_AAALWH hetrick_s_Page_078.QC.jpg
71bbf47f44e27ff0a1dd9ee7121ce468
47fee8733e23aec4a986182c9ae67acdfb98c02e
2454 F20101205_AAAMAZ hetrick_s_Page_080thm.jpg
77d4d1e662e32ac75ea168794164cabc
74e98d2bcc46f454b93e5f3cc624c9a4a68f7b92
911178 F20101205_AAALVT hetrick_s.pdf
bd26315b90194622cadcfb237fe10b9a
6a0310aa0b60c5385ac457536fd99872fc5da5aa
6512 F20101205_AAAMCC hetrick_s_Page_073.QC.jpg
3965c60ca93aaec746257f297ea06725
bbf5f713c57aa81c1ee6bc22a113402d9a746ee0
6567 F20101205_AAAMBO hetrick_s_Page_051thm.jpg
6bfd2e5146e95c58d1c1d7f8d38c2b44
0e0657e8b813e756dfde7594f11d34a169a63616
2940 F20101205_AAALWI hetrick_s_Page_086thm.jpg
6e7f0a92c8b8531f7841d20295dea8f8
9b1a61d16c8703f2fd433d2a1353d004cb1b2302
176383 F20101205_AAAMCD UFE0017822_00001.xml FULL
184f9d3efbd84d7e0f76dbe6be1181df
21c91e0082effe1551a1e5edd7470a647835ba2d
5264 F20101205_AAAMBP hetrick_s_Page_004thm.jpg
4dabbceab3bcd1fd17596c7a6444377c
0a279c2fcdd5c3fccfb7269a9e21b655a7305bf1
6925 F20101205_AAALWJ hetrick_s_Page_110thm.jpg
dc7835d41ec09dd098c253f6cd4b308d
c6ee2026fec9f78b20c03d12d425f19f26d90499
1588 F20101205_AAALVU hetrick_s_Page_001thm.jpg
467c0be1ebda97ba085ccbc8f0e21018
94413b7bd87507f63e6eacc26cddaeec801c1f90
27241 F20101205_AAAMCE hetrick_s_Page_017.QC.jpg
6eaa5cdbd312031bd0cc48e0ba91fef1
84b61598dec5eed2b263ea8f3fa77189211d9d45
6583 F20101205_AAAMBQ hetrick_s_Page_053thm.jpg
69ceaaad56faddf613e0d2afec134f66
f7101cfc45df08963b38c7479405adab402dac1c
6062 F20101205_AAALWK hetrick_s_Page_048thm.jpg
a57fac5f174ff825333b3000cd966f62
2e59e188d1f298bdd638cce9198dff07423f8651
6566 F20101205_AAALVV hetrick_s_Page_066thm.jpg
7dd46d7479f83f6e254d8708baec461e
3377ec5bf388ae2b04b78d93b7f4d66472da8e29
F20101205_AAAMCF hetrick_s_Page_025.QC.jpg
e619c11c7df7d87aaf3c60305aa9e3f7
c0278484a16ec4eb4937cc4dc509c82a763607af
26416 F20101205_AAAMBR hetrick_s_Page_020.QC.jpg
d17e8353b6540021b2a1eca443798fd4
c803df49cb76c9801424bb91e60caa6aa43ee25c
28658 F20101205_AAALWL hetrick_s_Page_027.QC.jpg
f342a9f6c59ccbc58d5762dafa830653
97b7c831a98e95416fe1ecb9ade4577e5ed78e3c
6988 F20101205_AAALVW hetrick_s_Page_018thm.jpg
c8b1a4bc41608f3e0721b70034673ad9
ba25a11efd6cc55e5d36ec6785915b1bdf6e5cc2
6796 F20101205_AAAMCG hetrick_s_Page_026thm.jpg
91a9322b3e0db39ad08761624aff87b5
c38bfdff7b3c119dd5f83b0e18f3c5cb1f436bce
7819 F20101205_AAAMBS hetrick_s_Page_076.QC.jpg
8fc7cdb0162ee9b8957373869c7f129c
9e7a26c2e7693293047c67e1d9cb27cd433f29b5
F20101205_AAALXA hetrick_s_Page_052thm.jpg
420f1f2cd4fe0b7af450bc1a753daa98
30140f3e65dbfd83560bd7835cbb8639281663a7
28033 F20101205_AAALWM hetrick_s_Page_024.QC.jpg
d809bf1ac05381d87ee19b661a1a8417
b9ff7a6094fbcb8c613641d8719886cda6ea4dc6
28960 F20101205_AAALVX hetrick_s_Page_056.QC.jpg
9771877c61c6a0548692e851e3853c50
f8ff41df14056e3664c95036cf42dc7685b478d9
9142 F20101205_AAAMCH hetrick_s_Page_039.QC.jpg
28eb9966266778af00681b4acde53894
1df2b858bf47fdd7499269049af79467ecfc3c83
6741 F20101205_AAAMBT hetrick_s_Page_025thm.jpg
1242cd5a67abf497120436abbf9db126
b3176e7467361a6a1d726522129352d034abd516
13348 F20101205_AAALXB hetrick_s_Page_008.QC.jpg
a508005c315e242da9ab643665583ece
330926f823832bd580a8e128dc1ebadf1b843ca0
6090 F20101205_AAALWN hetrick_s_Page_009thm.jpg
66da1730c8934d4b44c6255e13215971
297fcb07bda7330196d3cfcbe19a00cc0642ed4f
30179 F20101205_AAALVY hetrick_s_Page_030.QC.jpg
80076648055daf43f5a5c3dc3bd1dba6
f2a5a789e2d41b01d2fb949a5488554a7bc265cf
6893 F20101205_AAAMCI hetrick_s_Page_049thm.jpg
60c646168cfadb1c6825992a5c7d680d
006baabcb84fefd70f93218a2cf0a18f29e6c06b
5577 F20101205_AAAMBU hetrick_s_Page_106thm.jpg
4d768c9e7a7ff86b3adc4d0f0700d56d
eba135958f1eaf2a7654558a6cef4b124e75267b
6617 F20101205_AAALWO hetrick_s_Page_022thm.jpg
ee2fa1bf5d8fba613df2babd03f3ea67
e840fc3974136e0170a3d8cb122ffed7e35e44d4
F20101205_AAALVZ hetrick_s_Page_085.QC.jpg
a343172f58be04df968bf81bcd9137ab
ce018756982f0fbfa140ad7c41a94112f245a72d
29690 F20101205_AAALXC hetrick_s_Page_018.QC.jpg
aba2f71a0a4d1f9989f19f3bc1fe5fac
47a15e0fe563b79eb7ec4e5df8fc1d5393aaf0ac
27953 F20101205_AAAMCJ hetrick_s_Page_059.QC.jpg
f57ff31ece364a2c800fd56550237e2b
7a11eac1f6602888fe69c66ac64d501474abd2f1
6957 F20101205_AAAMBV hetrick_s_Page_068thm.jpg
706e095d07c3159496a97feb891a2236
99a89ba93e256bf67e1e077f313eb19067dd5ba7
3782 F20101205_AAALWP hetrick_s_Page_074thm.jpg
308e91f3438ee255eb1ec61b251a4db2
2f138b83124b2b46817c54446dc85b7fec003dea
6948 F20101205_AAALXD hetrick_s_Page_115thm.jpg
4c1d33c358668cb3a6e2a521b26bf2fe
923ed4c3c69ca24167636b0e683b77a820ecf52b
6278 F20101205_AAAMCK hetrick_s_Page_063thm.jpg
e709dd55fa2023b888e9338d662e49b4
c7a110ba524e903edd1af0ae56e8bfde0d752b2b
11696 F20101205_AAAMBW hetrick_s_Page_116.QC.jpg
6c2514026935919c1d8bae8299ab2b9e
d2b5cc199b3b02ed7a86a19ede8b4ae3566b7f47
6671 F20101205_AAALWQ hetrick_s_Page_097.QC.jpg
1b13dd6ad13c1948c05cd8ce025a76ab
601e95ee4dda26ba5f2fef458a4ed8741f155e0a
28059 F20101205_AAALXE hetrick_s_Page_045.QC.jpg
c84c5983fcd1849f2af2f66517c9005e
22da051738f29fd2caae77d08d879fb6eb8d9094
6694 F20101205_AAAMCL hetrick_s_Page_067thm.jpg
598e84d8ae719ce4ad527656b25cc36d
e30d15e8a67ff05c8b2c901045dcfd658b812d69
9119 F20101205_AAAMBX hetrick_s_Page_084.QC.jpg
1895ac45810dcc9836ff3ee74f114885
ba424f54e9e620d5c4c098aa8ec15e7b5e1f1995
11215 F20101205_AAALWR hetrick_s_Page_095.QC.jpg
960779b6f7d63cfbf0dcc07e17311caa
681b17f80642acd13cb9ee8d92e38f1ba2e0d8c1
29614 F20101205_AAALXF hetrick_s_Page_058.QC.jpg
bbb482d168ec994aafa6e9834903fae6
2320fed08687f5dd6bd85ebba2e8f142a3073cb5
2325 F20101205_AAAMCM hetrick_s_Page_081thm.jpg
23e1cd17e9985a86eaae744e52f1b497
2d0f1884fbb05d1be92d883c9af49c2fbe33b784
5215 F20101205_AAAMBY hetrick_s_Page_105thm.jpg
bda976d944cc8ca0753ae648d40cfaf7
9ceca57ca8c83d3bda17a8c9fa19913a74a97e64
27269 F20101205_AAALWS hetrick_s_Page_093.QC.jpg
55a5feca04d7a7d92fdb2bbde72f47c6
1c05778d8161857579360098f865dc2e444952eb
29143 F20101205_AAALXG hetrick_s_Page_016.QC.jpg
c4de97865e47f7180d5c12e31e21e09c
5a7464dcda76a32d3c3cd5f0defe506ad27057db
4343 F20101205_AAAMCN hetrick_s_Page_094.QC.jpg
b4471435a91b2d256b250dfb88bc1115
168349ca8993ba4831c4f159a5b863955fe738b9
6771 F20101205_AAAMBZ hetrick_s_Page_058thm.jpg
8f679deb05902b78dd15e46b0185830f
5f7685ee5f7debdb7890403306fc3359c25cfeb1
1405 F20101205_AAALWT hetrick_s_Page_002.QC.jpg
392de93ea46dbc2251a7e4cd357138e8
16d27cffba8e6f84f5e33083d705c9e08caa6886
28923 F20101205_AAALXH hetrick_s_Page_060.QC.jpg
6dd5ba5336aa471555b612370cb11d34
fa354ee1e6a2030a46e8ca343d8eafbfccd78177
2318 F20101205_AAAMCO hetrick_s_Page_100thm.jpg
d78fd50cfe17d841eb85a95f050a2c0e
0257769a5dde65d78279d7ce5c83ed20ac96d461
4556 F20101205_AAALWU hetrick_s_Page_041thm.jpg
5a5d9540295e9fc1a17061b93a5463b6
6996a94508994093bba6afec08e817040ac3db84
6765 F20101205_AAALAA hetrick_s_Page_114thm.jpg
26534c39134695ed5f815e19d8f4f06e
7c79b827271fe6a43b538d8e9e5edc553254e88a
6139 F20101205_AAALXI hetrick_s_Page_112thm.jpg
1dc79f72ab29da71800893052a762d22
178d9aca80f3aee789ec64ef8e62e42e6baa6d5a
11350 F20101205_AAAMCP hetrick_s_Page_101.QC.jpg
3ece22c34853c70040d471974a2ef477
203859559e0f9e3feb1c4e01fb1fb9c874621fb1
2090 F20101205_AAALAB hetrick_s_Page_054.txt
73de1b84688ea8410e0c6c662329fb8e
d7097866c54826fffb3fa6e63c8190d90aef4584
28014 F20101205_AAALXJ hetrick_s_Page_046.QC.jpg
9457403882b7cf62a648d850c63e7f86
9d527cc55a780756135f0869e056c68ba084f7e7
28931 F20101205_AAALWV hetrick_s_Page_065.QC.jpg
04311230552b52108b51d5610378fb82
9313b003b42b56b203ee6b173ff3a1bcaab8f6cf
54225 F20101205_AAALAC hetrick_s_Page_112.pro
adbbd1af40c161da406cc9fff3ca19f8
c2cbc1b0cab624a975f3bc10a0918913e67fbb6d
29229 F20101205_AAALXK hetrick_s_Page_072.QC.jpg
12c64bdd07e9d1f98c3ed3660c6816f0
f7b85f0e13a7a3098108ede9fb32cc39801e2d99
25861 F20101205_AAALWW hetrick_s_Page_014.QC.jpg
9853a8bb1c14220898120515ed833e54
6c24cca167f15f3d1df920550ce4ab70384ecda5
12018 F20101205_AAALAD hetrick_s_Page_035.QC.jpg
16e36d82bd33748c821397f1624bcc00
edf69b77a75bc222f5e38c67d9218a6eeab314d5
2436 F20101205_AAALXL hetrick_s_Page_076thm.jpg
72d199998de09f07d7209fd2c316e526
e3c2ceeb01cfd47cbf3bfa8946f2c152f6ddf6cd
21374 F20101205_AAALWX hetrick_s_Page_106.QC.jpg
028f4a14c58a1f1a326c16c614f0076a
2081db6b73e70805e64042c05ca64aab40ce946f
66918 F20101205_AAALAE hetrick_s_Page_107.jp2
7a4bee74a3b91df7ccd64255dc7adda3
627663a092a0ca15e9f8e06e6fd9b852e9a7b6ba
4759 F20101205_AAALYA hetrick_s_Page_005thm.jpg
94af5cd5b72265a2d32b2a7579d382f2
cc9198efb64c08a6a77a2f6ddc551850c7b12164
24990 F20101205_AAALXM hetrick_s_Page_092.QC.jpg
45db811a463420c89289cfe00ea77706
c98a8d62b91e078b8cf8e5604dd781908f500388
2634 F20101205_AAALWY hetrick_s_Page_040thm.jpg
b0040a76337cb4e8390727a02dce3c9a
2871c9c88cb26a548f81b02d44db8b5172d02fe4
2138 F20101205_AAALAF hetrick_s_Page_046.txt
7a03c4da86340737a22a7181ca1ac432
afb501e6c9e29428ca78cf3e025ba076786be305
22154 F20101205_AAALYB hetrick_s_Page_006.QC.jpg
8375d531dd3aa95027fa494ef2c55e79
88fcdc3b91710bb7a06ef7c45d47ee74b22ed41c
9316 F20101205_AAALXN hetrick_s_Page_108.QC.jpg
9bf3696b84dc91981bde5bc61683ad96
7f8a6cc5f59ee28f24247cfe0e9a618ac3363e31
2842 F20101205_AAALWZ hetrick_s_Page_116thm.jpg
e4e4233ce54a4f303222a30e1166d1b5
106f9fda202c3f318ac28a28ac35e98b8a216e15
1053954 F20101205_AAALAG hetrick_s_Page_072.tif
8d61ed1440ebbe6712f6f0fe516b64ae
38aae65127f788883a698f06046f52c4e54436e1
4775 F20101205_AAALYC hetrick_s_Page_087thm.jpg
f2d30521b9c49aa444859225265d3211
3b467f6f1e3fa4184c28ae28dabe15040b6969b0
4114 F20101205_AAALXO hetrick_s_Page_098thm.jpg
af54e8d7273b0792c1d073ba9c48973d
3e9b053f7304fd22a2005361d174c95e916aa2d3
64212 F20101205_AAALAH hetrick_s_Page_088.jpg
c3657dc72aa3723d133ba6b9121c9cd5
2deda1fb77d6d1874578e95016acc60b11381e16
3061 F20101205_AAALYD hetrick_s_Page_101thm.jpg
d70e82d74a3572cd4ab422b94d70954b
58e79efa6fc52d4ca0f801cb1d9a0d09452c02a2
29268 F20101205_AAALXP hetrick_s_Page_062.QC.jpg
a46f289db7a6afd070cc0014b639a2b2
0549048f518e1839d69ef1807a357c588df97dcd
2170 F20101205_AAALAI hetrick_s_Page_043.txt
0ff8e5ecad8d3906de097a677821e682
6d79307a18faa1c648c0a1b787c80c02aa4644f4
6843 F20101205_AAALYE hetrick_s_Page_069thm.jpg
a4485faa1baa7eae42831d2fe845362c
00797c31a011ca4ebd4ceec4ec91519102e2f869
2923 F20101205_AAALXQ hetrick_s_Page_078thm.jpg
9535fed8ac546e3f5223cc13885b4893
a4c53a6e09f49221448008839cb08f059144ed5e
2375 F20101205_AAALAJ hetrick_s_Page_109.txt
e512ffbd446eb9833e5ca08217567406
060d94ddb458b5a60e4ef56fc29a5754c66613ac
5527 F20101205_AAALYF hetrick_s_Page_102thm.jpg
30516112bff09e86135945a0f8ffce92
535ffe4bd44e367bec3c82e1cdadedafb0682e6c
2455 F20101205_AAALXR hetrick_s_Page_108thm.jpg
634a63fd0e8a9f288b3bd3d8ca550062
43ea34871d19e63a39b7fe5b893988ef00f191fe
F20101205_AAALAK hetrick_s_Page_009.tif
5b6c7f2938094e75e2603172fca0392c
12d7cda158fea4a9acbe8f3363bcf1b1a729ea07
28944 F20101205_AAALYG hetrick_s_Page_047.QC.jpg
92609297418f110faa4ceecb84c27881
1fdb9e14974245b6acb415d4f6910b9581525160
5391 F20101205_AAALXS hetrick_s_Page_117thm.jpg
f16308e8ea31bb3589fe6662ac2daa17
abbbf19236c3e2693b2a795f4ae977f00ae62e98
29673 F20101205_AAALAL hetrick_s_Page_070.QC.jpg
2fe403478a048c7fd7df6c8bfc3c9801
50a17ebd3794e6cf334ad7966892a46a70f071f5
1375 F20101205_AAALYH hetrick_s_Page_099thm.jpg
6ba0e0dfa3c57e0356c80580a154c1ac
df2262f70c71eec015b2eb9f8838d63fa45daeee
6791 F20101205_AAALXT hetrick_s_Page_045thm.jpg
0d0afe6f62530dc71bb832b091edc7d4
0b93c031c88ce010d7b39eb4d382db10fd8e9ad4
2238 F20101205_AAALAM hetrick_s_Page_072.txt
a6e994b2d1d9e0c80881396aea9ecf8b
bfd7ebf7162f016f78d8680f05648bb0da346f6e
4912 F20101205_AAALYI hetrick_s_Page_033thm.jpg
43d5d049c7123c24ce3ad0237eeb8acf
2f0c8f076e563c17a5d3e1c58f944f723941a228
9692 F20101205_AAALXU hetrick_s_Page_036.QC.jpg
fc8bd3e90d5197ecf08b7f5bc82f6829
ead8449a76c36861d5f9e8eb1ac0c900583dcdd8
121133 F20101205_AAALBA hetrick_s_Page_049.jp2
e904198ac07a3e980457af88576df9ab
42414940841595dc6f10dbd933ddbffdcc7df808
28882 F20101205_AAALAN hetrick_s_Page_053.QC.jpg
43b2986976dd0569cb18c216ae7c14f5
a7b9ee355c076fbecaaf248f0ed7827b76cc3cfd
2982 F20101205_AAALYJ hetrick_s_Page_039thm.jpg
1f659a8ebddd04122b51aa37245580b7
d23159e33d0a0da9e495c203ac0e52649e11eba7
11982 F20101205_AAALXV hetrick_s_Page_074.QC.jpg
82af2ab2f8f2506db8e0f049cc0402e4
b18169d07f21ce22218b6ecccc84a217e48a8466
24601 F20101205_AAALBB hetrick_s_Page_079.jpg
a5836d116ca6183420a97ece2f70bb37
976b7f5df99910504e1a7fd85bf67b64a138240e
128482 F20101205_AAALAO hetrick_s_Page_068.jp2
2306bcdc3b8b0b324871a0350808ed4d
5cde4de18f0c50e0606459e9285ec4984813db50
22625 F20101205_AAALYK hetrick_s_Page_103.QC.jpg
56b38345fa3204c547d579426af7b6a8
a6bdb9831d5078bcccd83ed292a431cf31d872aa
3149 F20101205_AAALBC hetrick_s_Page_036thm.jpg
6d27af6106c57bdfdcb90d6492acae13
ed396f4221123ad7d0dcbe533652d592ec6b61f3
F20101205_AAALAP hetrick_s_Page_117.tif
51da32e6f157f5933223bca348aee41c
486035d21e7dc85bddd1451a9717498522f65d5c
6442 F20101205_AAALYL hetrick_s_Page_050thm.jpg
d227df1c60e569d1d980514996d4d1d6
fb8d544b1935881ab1e23c4da336875eb684077c
F20101205_AAALXW hetrick_s_Page_072thm.jpg
cb9a998372f2a6c0c847b020293eb7a8
d89bb5a494fd9e4d40b96c3985b7f34bb11fefcf
24226 F20101205_AAALBD hetrick_s_Page_001.jp2
2d23dc64ef7e1042e07f74bc38d8089e
f773e21b10b0dd39b903fd7089170428851ab925
41982 F20101205_AAALAQ hetrick_s_Page_032.jp2
20af54a68385f08f985ed200f04d2a41
a2f155bac763191dd9b64f496e5273d969c3672c
26776 F20101205_AAALZA hetrick_s_Page_007.QC.jpg
64118d5ba8d69df7e64814d7c6c6f8d5
5c5ad87926fda85bf8d2083ab87b6af8f6fa84eb
17666 F20101205_AAALYM hetrick_s_Page_098.QC.jpg
cbf461c277cd3ce2aec7caa6b3aa1af3
76083a5a487cc057f5fbef6ecec4ceddb3e63a99
8641 F20101205_AAALXX hetrick_s_Page_040.QC.jpg
c437099f8ba06dfb78c5b5303dc4796d
a1dd921c323142ccea9c191207c6d1a8cf0afc06
58089 F20101205_AAALBE hetrick_s_Page_110.pro
1ed9540da5f01d0e292f2879d8a16e32
65ff43fe8ecc2b2d9f56ca1edb1afee60d810d48
6230 F20101205_AAALAR hetrick_s_Page_001.QC.jpg
e76d26f4fee43413288b665e425a6014
08186a0e8b41629c62c4fac2842b6738888a0d63
4514 F20101205_AAALZB hetrick_s_Page_010thm.jpg
566d20f07e5213874bd1b98d6a2bf156
e64c409f1d264d03be43f0bef8471d7ad3e964b0
19578 F20101205_AAALYN hetrick_s_Page_088.QC.jpg
65651fde749fcd038881b28830ca5cb0
93491553a7278bc18fd7cb8ba8311fcab29fcf89
6785 F20101205_AAALXY hetrick_s_Page_055thm.jpg
b18912bc2a8a1d52212e5590d0e95aa6
1887490b83d30a4d9e3bcc221ea2c2bb110be508
F20101205_AAALBF hetrick_s_Page_073.tif
35e6ab99a77aff4463a1b1c505c2b091
fe4c5a6f6e629c35d653065f4ef62631e95bdee9
2089 F20101205_AAALAS hetrick_s_Page_069.txt
09b995c2304bffe9c6ebf41ad1921120
19cf34a5bb5bc021a79e608335c037a6052c0f32
29396 F20101205_AAALZC hetrick_s_Page_064.QC.jpg
1a2846c404d3d315dc7ae179bc3493bf
22983b1e74f60b5e5dc545ab4581f79aa770a9e3
1100 F20101205_AAALYO hetrick_s_Page_096thm.jpg
2295ed917315728f05365fc86fc5c7c2
a32e20c9db665a8fc9869b291e6e7503202a8a66
29885 F20101205_AAALXZ hetrick_s_Page_015.QC.jpg
47b75935e53ddf33da66de00e2c5e826
3d51986bbb122eb65d5c25d0008200ff4b54c08c
376771 F20101205_AAALBG hetrick_s_Page_038.jp2
8656e46052afd901d4d52ea9805109e4
0eae628f1ccf6abcbe474c97b0e0e80b7a85b136
6522 F20101205_AAALAT hetrick_s_Page_059thm.jpg
fb31f6a3726796bb691fcac200a19d02
f2f3996d12935ad557970ac0355343eb912dca7c
2690 F20101205_AAALZD hetrick_s_Page_082thm.jpg
6a44ea56760abb6cf97def165a32eff3
6874c867d5ebc1ffb9c1e1b9b9d8f7e844742006
6845 F20101205_AAALYP hetrick_s_Page_015thm.jpg
801c6df475b5282cef565a6afceff7ab
692e4ebb3951441f3a062d4f7dce4214f2f3900d
2266 F20101205_AAALBH hetrick_s_Page_064.txt
d81c3242fab44097f02151b040a25e74
5c500395f7f91ceefc2078c5b501c545b624edd8
2262 F20101205_AAALAU hetrick_s_Page_030.txt
80d60fc1fe6e39a9cadd4df1dd367546
279e0d67e28571da9f0d85666dc753d63065fc18
6151 F20101205_AAALZE hetrick_s_Page_021thm.jpg
f675a027930f663c38ed9e5723c0b4c9
08bc5662b396de1a2f4968fe8516a64f988f57e9
6616 F20101205_AAALYQ hetrick_s_Page_044thm.jpg
7c16255c0bf7c9f36d8fcd2e33306207
8c34536a2457396909a0eda1635be88127817453
101643 F20101205_AAALBI hetrick_s_Page_007.jpg
670bbc758469e99de44e4e39d33954ab
77d9febd846348cbc3d2a048d9ff71054e8206d5
89347 F20101205_AAALAV hetrick_s_Page_023.jpg
44e99d3f87abe26bfcc81c91c86664cc
4c342845419520f1fa55665288a1bde6f9801d84
4894 F20101205_AAALZF hetrick_s_Page_003thm.jpg
41ec8e32985846d85acae7905f51475e
58f9ef80db3568f07fbfafe696d44a360406ab67
6726 F20101205_AAALYR hetrick_s_Page_109thm.jpg
25bea7be365b052d8da339793641a853
8afdb1a692c600fd30bda1785997099ab054ca55
81099 F20101205_AAALBJ hetrick_s_Page_010.jp2
35a9c1bee60b62fd440dd4fdb2e12780
071bf7235121c5fc7d6cb2ad0f21a61dc10b388b
56023 F20101205_AAALAW hetrick_s_Page_058.pro
353aed441beb8d4333451f0e6ee08db1
626c1c0a3898e46a1224ac6f0b3c614473327437
18826 F20101205_AAALZG hetrick_s_Page_033.QC.jpg
db6aac2feb52c178810c912f54194e05
6605eb0b12c102ea1936b61af1996723c4276b2a
6382 F20101205_AAALYS hetrick_s_Page_017thm.jpg
69a0fed4f8bf789e84c28c799461ba78
a387955850b1d32dfb173e78feeae7dc818bc89f
14457 F20101205_AAALBK hetrick_s_Page_099.jpg
2d65752a2ffff9b485c1615b8efc4cf3
07784e8a7302546e4569697631e6f8cf589ad384
27836 F20101205_AAALAX hetrick_s_Page_114.QC.jpg
1cf44dfc503b082fd820862288e93fc4
d291cc989bf2294a177ce3fabe6c37179b8c8b15
6661 F20101205_AAALZH hetrick_s_Page_043thm.jpg
e165e54bf92177d5de63842377ccd7ee
fc55b57a8a873b9a66140ceb1b40ffea91b80514
29367 F20101205_AAALYT hetrick_s_Page_031.QC.jpg
a5923948ecf8eb343b2d325e3f66878a
7e31674cf72e9bfaca8e707d37f795585fc84140
7219 F20101205_AAALCA hetrick_s_Page_083.QC.jpg
ac7d1424fba2774dda810b6ef4d637a2
cf86590a6dc20a807af5ddc3d28080fafb7d6a1d
50874 F20101205_AAALBL hetrick_s_Page_050.pro
9b31ad08722800b6336569eae4bf40cd
a8e50031edbf83a730dc7f63934b7f2d6aa47156
1691 F20101205_AAALAY hetrick_s_Page_073thm.jpg
a3d03414c371daea07fcb67d1e4ff6dc
08b2224a7b7972fc075db23a9897d710f538d596
28526 F20101205_AAALZI hetrick_s_Page_042.QC.jpg
e5a1ef8ab418e71761dfb19fda418d98
88d675d212fd2e4426c7193281dc13eb208dc8fc
26411 F20101205_AAALYU hetrick_s_Page_090.QC.jpg
ff028c53fa9c18698bbeb6a8b32278d3
c32b2769feacf4a581694f2125e614be647ce04b
3228 F20101205_AAALBM hetrick_s_Page_095thm.jpg
67ae5e42aced5461dc6c43d854029a87
94d63dde397e180551c5eab6b643cd92629941fe
93652 F20101205_AAALAZ hetrick_s_Page_064.jpg
947329d9faf90c658571f07cfd8df32e
0c3b87c31656dc5e327e198630a0c8a4c4b03d1b
9019 F20101205_AAALZJ hetrick_s_Page_100.QC.jpg
344e2eadc52694c25a6e3e95930d24d3
ea272e188b86045d739dec0d2294ee714be70b30
6483 F20101205_AAALYV hetrick_s_Page_056thm.jpg
21a8d077b2185777c7caa396c2b1395a
2ca290a2bd0a987fdcee82931fb83624a6ae7146
53567 F20101205_AAALCB hetrick_s_Page_024.pro
82989b9d1e348edf8523c78cc4d3f314
60ea09b6b2e0c9b65d8d3eb16f6848a4b7db81c4
5262 F20101205_AAALBN hetrick_s_Page_037.QC.jpg
9b2472730e58324d3c92551789684d8e
90b58fe7357d8af8f84798bc4227ae5618a47097
6462 F20101205_AAALZK hetrick_s_Page_046thm.jpg
b470706c244d3b78c15426bdd562ab02
7823e80554bf3ce9a55f04d0cb468d81e2a3818e
3019 F20101205_AAALYW hetrick_s_Page_038thm.jpg
5755923a1cbadd5f73b7d4232b8c8ca4
2072e2449afe30e8010ad374ea28c5f7e9658e87
2602 F20101205_AAALCC hetrick_s_Page_077thm.jpg
b797add9d6044a142bce90111246b64f
cebc0a156fdec85c20f336fb483b73064b3cc27f
86938 F20101205_AAALBO hetrick_s_Page_019.jpg
bc510c5a47753864e86cf3fe0bb13804
3c05015d7cc603bf8431ab87847731138da5350f
19456 F20101205_AAALZL hetrick_s_Page_010.QC.jpg
8f976ba8f874e734607942b46194b7c5
f27b9adf83b6c5bca75f1bd167290b5ad426898e
474 F20101205_AAALCD hetrick_s_Page_039.txt
668e15f8c7ee4e2407c1cf3920cd0aa1
54945f774b8b37229d29d01c184967fe92b41856
4192 F20101205_AAALBP hetrick_s_Page_099.QC.jpg
67a6c499828d4f2d59f555ce7a138360
3cb78c3b2a97db01f7621d63dd5c41b855d9fb72
6870 F20101205_AAALZM hetrick_s_Page_028thm.jpg
94fb16e61f476e31a728472df6feea13
0aef62e69575c5b963064745371d1b316a966246
8276 F20101205_AAALYX hetrick_s_Page_080.QC.jpg
e2458bbbb60057ca266f62bf3568c441
4d0b775bb4d6449c2c0c67426d207bd10efb2bfd
26483 F20101205_AAALCE hetrick_s_Page_050.QC.jpg
65eb0048d024d36d53d830960ecffe02
896a95e5a0fc225df83e02770c6869a17a4cd0b6
241594 F20101205_AAALBQ hetrick_s_Page_036.jp2
3224ec5e635574e6f8a8fdd4427c8338
1ee1deab4c723533d78650ddca7f8b7b9f28e7c3
28908 F20101205_AAALZN hetrick_s_Page_071.QC.jpg
bfd10bb36e9b87a946b8a8986673f531
c2cea8c041eef7b4be1493aa29cbd641073b52d8
28004 F20101205_AAALYY hetrick_s_Page_048.QC.jpg
548ae93dfa1e9eafcc2c7f644bdd02ec
72f750fe33c59d8f1b38f0bbf3ddc7955497a634
2371 F20101205_AAALCF hetrick_s_Page_079thm.jpg
4935fd859a6e5f197e39f911fe3f6c78
d56b7f09bc443af21611c49224b1c5ec5f406daf
8595 F20101205_AAALBR hetrick_s_Page_077.QC.jpg
2743101daed22d9b5b5e4f1391dcd0d4
9147a9cd626a6402ab3b05b997330bc6be386183
6256 F20101205_AAALZO hetrick_s_Page_011thm.jpg
e5a3bc8e81a3e855a03d9888ae4fd4dd
0ad2737d626fe438248c0842317a884c22a9eb0c
6805 F20101205_AAALYZ hetrick_s_Page_027thm.jpg
d1d9d42a6e799d496ab14bfb837d57af
54a11f17896bb67aae68fe1944b9d6497557c26b
F20101205_AAALCG hetrick_s_Page_045.tif
ace39ba8e609f7f313e5a043260ff0c7
c0b905f17379447c4d12353e8e0d74ff3b8772ba
2202 F20101205_AAAKXA hetrick_s_Page_031.txt
4f8cc1421ab90e89b2fc8136ca57e343
1f911021aa67f74d96dee175bfaa0b8eb101de7c
66672 F20101205_AAALBS hetrick_s_Page_106.jpg
c480b09b08ea9225024983982e22a5aa
00f164b8e9e16270c17bfe692213822d75304027
6279 F20101205_AAALZP hetrick_s_Page_024thm.jpg
a5a9168490ce6014b1486e64cceba8fb
f8eb7352909afe0246793e8d89184b0d3a4cde99
92350 F20101205_AAALCH hetrick_s_Page_026.jpg
32204b3c858ac5b2af6a9dd10f1c6b33
bb0ef4326e1e6df83d60fab5cc1ea2a17d3278ff
8072 F20101205_AAAKXB hetrick_s_Page_079.QC.jpg
3288204fd78c8d9ee2c51909e11fae46
d5c10226e9dfd539a901f506174c0170ca7a1248
98756 F20101205_AAALBT hetrick_s_Page_004.jpg
cfecf3224f4a0ad262cddd94de66bf56
308f5ce50c5f558ad5918bb779df3386acd60aca
6709 F20101205_AAALZQ hetrick_s_Page_047thm.jpg
00def3e4f5797db85d0e381507edbc01
81e89a0db7576f23e299effcd062508d5b01769a
5792 F20101205_AAALCI hetrick_s_Page_107thm.jpg
8db8063b48c6bc5796cafe3a727f11b3
b46e612ff087cdcf51dba481b2dd2bd27aa365a2
6501 F20101205_AAAKXC hetrick_s_Page_093thm.jpg
a49242997d69387be1c3ba271294a66d
4a354ad6e4810a582e2788482c5ad7185045b828
393 F20101205_AAALBU hetrick_s_Page_001.txt
f9c2cba06c7f649c41818eeb60a30a71
fa94d54eaa948b8e15109a4cccfa450c077766de
5786 F20101205_AAALZR hetrick_s_Page_013thm.jpg
57e219ae3f6aad3ec72f1f3552057361
b569ac27f49aaf97584e3033dd963578ce9e0e69
57693 F20101205_AAALCJ hetrick_s_Page_030.pro
23c39724cfd7fa80596940bddbb7c55a
556143a5179f90e9cdc7be8f4827d2f8fa1b1740
32910 F20101205_AAAKXD hetrick_s_Page_078.jpg
327861a6e9884046b80334651b8f925a
9b29f9f48d0c10c1effba901ea9549905e300ca9
109933 F20101205_AAALBV hetrick_s_Page_090.jp2
3f0668dd698aa3f4313de9378a24ca31
f81ee8f1347b7735a653ac96b727ef9a68b7ea68
F20101205_AAALZS hetrick_s_Page_007thm.jpg
f8260a4ac073801435ed71a9ac41f446
489fee5bdb93f83c78974f2c23fa6e3a43fe38ee
2220 F20101205_AAALCK hetrick_s_Page_058.txt
70214d0a3bdcf9a0d99d03bf184242d6
50f7ac5958680e10a06ea4c9f7f242090624ca1d
1051911 F20101205_AAAKXE hetrick_s_Page_004.jp2
2dd5dec55405591580a15fcf973dafa8
8e26c044a7d103ffaa6fdf187ca3e5cfb1e78ad2
2729 F20101205_AAALBW hetrick_s_Page_085thm.jpg
b7b090770512d63ec4eac8e7ac8943cc
a02b45951807083b7abead1b8245d54a21decd77
9939 F20101205_AAALZT hetrick_s_Page_086.QC.jpg
86ce570d3627a88a32db2ce685c59af3
0bdcb53e51fe318790e79239a1f458d14cfe43a4
2144 F20101205_AAALDA hetrick_s_Page_024.txt
3efb7591bdb1f688ddfddda2f9969f0c
820ec3eac07092ede271f881ee17aee7f8449bfe
25480 F20101205_AAALCL hetrick_s_Page_112.QC.jpg
afc9820edea6c51ddba5660b14b450d3
adfad86ffcf045993aa9d41be4e412a7670b3c9c
22261 F20101205_AAAKXF hetrick_s_Page_116.pro
42fb4826fac9f9e1a20d17611a8be403
d3adc7833b62e4fae945097a1af200e26655c846
356001 F20101205_AAALBX hetrick_s_Page_086.jp2
d5aecdc3fef1b61c18fd345c340db3bd
d676e41774f1295da4a805f827257e69880a1bd3
5913 F20101205_AAALZU hetrick_s_Page_061thm.jpg
57d87d86704362a595e0a1b85c253c3e
d1f87d923b3d0b03e610e053c3c0ed29392bbfa4
6715 F20101205_AAALDB hetrick_s_Page_016thm.jpg
afaa3c69bd9a9bb2b7f0b4b7802a77cf
bb4b6507e2ad2463a7ef5837fdef2945514112d6
29799 F20101205_AAALCM hetrick_s_Page_111.QC.jpg
c979a90fc0434e088a29173c10887664
2c5b1b5350b7fcfa989a95b4fd0665c8916b9ba9
574772 F20101205_AAAKXG hetrick_s_Page_074.jp2
ae326e17f9901740c954fc1749d241d9
10bdad7974758fbbb77518ab81cecc0e18aeb7fe
19802 F20101205_AAALBY hetrick_s_Page_034.pro
eba42480952925b4c6573905d3619111
3fb7bc278ea42eedf8b3ee5341e7c7a8eded14a8
16917 F20101205_AAALZV hetrick_s_Page_034.QC.jpg
b152f11c7dab8126f2f66870a6b3d5ab
e88a5154f540f140651875ed4a6ed80c77d476e7
28831 F20101205_AAALCN hetrick_s_Page_110.QC.jpg
0f0f51f9c8a8e045f5a63871cced8624
a652d51f5738172e1712e9760c26416cd9881f13
4834 F20101205_AAAKXH hetrick_s_Page_088thm.jpg
77a3ccb0a9f9e0df28d6ca7a78b32753
6572e2dde20993b2f25f9ddb66afb147d4db80e3
24548 F20101205_AAALBZ hetrick_s_Page_106.pro
abfbab8698189b3a9762b8ee443fc491
e9b606fc857a1d477e0f9f231e9a7610dc3f1f3a
6927 F20101205_AAALZW hetrick_s_Page_030thm.jpg
c9c45a6509edb079f4ee4c7bbf9c5026
d39a8c217587353bd0ab8d9561a3fc865d64842c
6781 F20101205_AAALDC hetrick_s_Page_060thm.jpg
dbcf3cc4155116fa33167b2fdbdee438
251deedaac1ef6a107eebf7d32515bdd8abbabba
4798 F20101205_AAALCO hetrick_s_Page_033.pro
35a4e0aefa3d8e03b1dc071c4e4979dc
7df07db7294d016d3625417a20160c5cd40a79bb
26530 F20101205_AAAKXI hetrick_s_Page_021.QC.jpg
598dbcb6be8d1291c977592aae6edc3e
1368b2b692513d9bce7341e91ff10f96a2e2df81
28435 F20101205_AAALZX hetrick_s_Page_022.QC.jpg
003871406a3056195d7a721fd4ac76e8
0c756d7473c5bfc2cd222b3e0ecba05a6997330e
88614 F20101205_AAALDD hetrick_s_Page_043.jpg
d6adb4a358c87eeb62bd67a0071a5fb9
e1c250945ee4bdbe79d913e3a025f0a319679684
F20101205_AAALCP hetrick_s_Page_065.tif
f214e55bb4aecc0553193f3cdfa6998a
eb9807b9325ca5332a275e0132b2b3b330169333
6061 F20101205_AAAKXJ hetrick_s_Page_094.pro
442ee52cb4abf97c03a167ad139d4ad7
44f527acb2d18bebf319ff5b675f73e9ba64ce63
2518 F20101205_AAALDE hetrick_s_Page_032thm.jpg
f9da52a6bd8edec922a47ddcd0e34143
4afb325ebbddbeb8e6ddae91efe0184262ffd6ce
25714 F20101205_AAALCQ hetrick_s_Page_089.QC.jpg
94e2f6e591e52b0ac702e02b33bb4b5e
e0d13e06fd3b9715bb41ff5c148dc7df97171226
55949 F20101205_AAAKXK hetrick_s_Page_066.pro
d99d1dffaec8c47a0b906cd95f62bdee
adaa4e771d7db6f5e90fd91526afb59d019c5f9a
6659 F20101205_AAALZY hetrick_s_Page_062thm.jpg
34152f920823459a5a205bee8e4b302a
0c5dd1c830a335be7cd08a0a13a4be4fcc6c1e1b
28669 F20101205_AAALDF hetrick_s_Page_055.QC.jpg
e6f4a87631e4bef423bb2b50839075f9
556af03fbda3ebe5820b1daef37e54847b2f2997
49990 F20101205_AAALCR hetrick_s_Page_034.jpg
40ca4f938da16dc78807d3627dfb60b1
bff59748d25c91bee70f225c5c40a7fd0f563b08
6500 F20101205_AAAKXL hetrick_s_Page_042thm.jpg
186b9fdca14e2f1ab92cf883c8be149b
9d47cad521654bf3a846f96f393a3b0711e276a1
29235 F20101205_AAALZZ hetrick_s_Page_026.QC.jpg
9af823e08f214722dc6793e8e47013fe
234e9e87b3a548e8505cde1e43fbe9225b6b75c2
6862 F20101205_AAALDG hetrick_s_Page_031thm.jpg
6960485adc13a9414710a6f9ed3daef0
69ab6a730b9c503253d6ab3da5f324968b8a8b2e
112196 F20101205_AAAKYA hetrick_s_Page_011.jp2
3b6913b4137474a0b76e1ca82651ff31
5eb0f41b582a9502f58d7ab469b9983dc0e27f03
6302 F20101205_AAALCS hetrick_s_Page_020thm.jpg
2986e09af86826cd456e5d0bec01f399
12ef074cbdab8e9754ffacbfc287a809c46c7096
22324 F20101205_AAAKXM hetrick_s_Page_041.pro
753142c6c901f0deb769f285aa949c84
a1c725da4cb027438b4ddf68727211969fb9e043
1051956 F20101205_AAALDH hetrick_s_Page_115.jp2
29e231309060597c1eaa699cb2f122f9
ae488794ab5a66687550fa6983b0c263d00b7ce1
78224 F20101205_AAAKYB hetrick_s_Page_012.jpg
9c2bc9086898f5c5cf20cf64f5c5ca6b
f63f7f0dd44d09461af5f007f45e10e6e4c56f3a
94247 F20101205_AAALCT hetrick_s_Page_015.jpg
fb8e8a7f0f5f8e8d69a7c1f16f468bf1
0f5a4d43321547261d64c95ceff6164bcabe59de
19384 F20101205_AAAKXN hetrick_s_Page_105.QC.jpg
2d9adf308fc73c78c7b4e573b74bb305
8acc0b4956cc85e2e5e3e33ec5bc02c4efcc4d64
28272 F20101205_AAAKWY hetrick_s_Page_044.QC.jpg
6ce4d0bf5f423cc04e4430ae2ef00e97
0e40d44c6cf8fe469eba1ee8d017f8a520bf2fa0
20408 F20101205_AAALDI hetrick_s_Page_003.QC.jpg
3b853dea0e1901055091028b9e754771
000a22efccf98b8d60313dda3c7996df909fae89
F20101205_AAAKYC hetrick_s_Page_040.tif
3e347e9edb9e622efcb7cb95d6ba009b
4f2b7cc7bc8a0925755bb006e9c3f0de2625d63a
26623 F20101205_AAALCU hetrick_s_Page_057.QC.jpg
e6fcfbc47dba25f7a6c292d1bf0bef8b
858f42559bc22aabbb49956b7e19701b5bbba80c
53545 F20101205_AAAKXO hetrick_s_Page_048.pro
a848ec05e9cac05c7b0335bfc1cfa860
e16696029c4f8ceaebcee856c796893acc8d8f09
F20101205_AAAKWZ hetrick_s_Page_049.tif
94a01ccdda36136582e65be1ad94ff2e
8b6c93e62d366f941c55ce5b844c9e71c3ae3015
54835 F20101205_AAALDJ hetrick_s_Page_045.pro
598b27bbb8d5c493803af1c66b68c671
d73505e7d58f9bf8d0e05fa114bebb15d7cdc87f
8686 F20101205_AAAKYD hetrick_s_Page_082.QC.jpg
4df88863621284bf484f37b2ee3a3c15
de08f5a1ef0508fd1dd295cc9383f0f3ae594391
114282 F20101205_AAALCV hetrick_s_Page_019.jp2
d1331dd33531de7cd96b83655519e19a
b4870f518347677dee335e3ca328fe4cfe224cdf
1538 F20101205_AAAKXP hetrick_s_Page_088.txt
bb2b3ea661db897d5e6ae8beaa222a82
1e17f7505c2a699fe727c646a4c3f8b0a357f7f4
F20101205_AAALDK hetrick_s_Page_048.tif
2eaa94bd143fc4a296340722972b19b7
01e2fadfbc068331c0093bf1413f0016a03c9e05
85050 F20101205_AAAKYE hetrick_s_Page_090.jpg
2ef2c5c381a2749e63e2f42107608ade
e016560b076ee1f0e28462b6926c2974031b0d0b
126754 F20101205_AAALCW hetrick_s_Page_098.jp2
cf9fc045896926be8e3cb8103f895893
60139b0d6c8dcdee131db62ceba22c74304a0713
26428 F20101205_AAAKXQ hetrick_s_Page_091.QC.jpg
5fd28382ba3efd01b364a6b03955f7ac
926c32153aa498b166bc0238d7aff878fe9795be
1164 F20101205_AAALDL hetrick_s_Page_034.txt
c9e9075ea970b1a1b0b9168bd678f038
4a187c69a8129d10384de2881df4a419451753bd
53483 F20101205_AAAKYF hetrick_s_Page_093.pro
cf5d3b648db7fa31035f337cc9c9a59a
965135a7802d46fd873c81bbcce176685bb10f69
8423998 F20101205_AAALCX hetrick_s_Page_078.tif
a69cb5827689f88a9497887577ff1993
ac2c4ff1bc9c79d1ed66ee6f94f0c11054755d47
5367 F20101205_AAAKXR hetrick_s_Page_006thm.jpg
2e32908d9b3798166bd432d09a852423
9d6e050ce17b977eea036394503f3abfcc7818c3
91039 F20101205_AAALEA hetrick_s_Page_016.jpg
931c8806f65c2b027ed830be719a53ea
04b8cebce85d4e726365c3556624b5c5b9b04c94
55956 F20101205_AAALDM hetrick_s_Page_047.pro
4bfcf812e2a1a576fb1578033fe354e2
9664982e6bddb699c373cb71ca0f3d069fa058bb
25271604 F20101205_AAAKYG hetrick_s_Page_005.tif
66c99db24fb270a63c105686a59c2db2
439467721ac7dc510d05b591128135d39a53ce1f
21603 F20101205_AAALCY hetrick_s_Page_104.QC.jpg
e802f90e2b8bb508bb1b1eb58cf0f72c
caa710dabdc4738d1a0f4d087be6c3ac788f7848
21708 F20101205_AAAKXS hetrick_s_Page_102.QC.jpg
509d960c60246557805446d19664ea89
0a71fc88607962776641e577b60a613de3b48504
87012 F20101205_AAALEB hetrick_s_Page_017.jpg
ff7f4447e555c41488faaaffc819e5b9
ef27b1bcc2cc3c84543d9a65ebf227c3c21d8b88
28424 F20101205_AAAKYH hetrick_s_Page_043.QC.jpg
fd17181f5d7cd660077a93ad0b3f5639
33eb8b6397e350307c9520cd15ecc5b18932ef19
5708 F20101205_AAALCZ hetrick_s_Page_012thm.jpg
1f849f24e87e671b7e7086540dfbb1a3
7dfac571d0747c23ca59da6e5a42a698f4e12686
28092 F20101205_AAAKXT hetrick_s_Page_054.QC.jpg
67710d0066f7073893ad764d4a9cf40f
e821ec05df4766b2626615054e65835c68ce8393
94243 F20101205_AAALEC hetrick_s_Page_018.jpg
b0dd228eaa4aa8020dc5e306d3fa1e5a
0f04b0daca3a5398f800adeaedf73967b1f8d500
136215 F20101205_AAALDN UFE0017822_00001.mets
c05942cb26bac27ce789b6f187abb598
46a5e7c06ba3279f225fdea7055c789dbbbd0a27
F20101205_AAAKYI hetrick_s_Page_019.tif
11a59a9365be5033a4db23632c67fe6d
f95f70230f430841097ecd3183f2b8c8ad77015a
14083 F20101205_AAAKXU hetrick_s_Page_037.jpg
fb541ff8b8bce64b243c826f33729bf9
3c82f36db848e57829025534aa3c2c3f4ddbab3b
57628 F20101205_AAAKYJ hetrick_s_Page_018.pro
69d82a6fdc03b9bc1fab08f7fda95812
4dd321769a87e2a0505db4a1256a931b10998161
6669 F20101205_AAAKXV hetrick_s_Page_023thm.jpg
e5cb41e7908081226aaa9f5e6e62c898
259d3d09ea5a0ad8f183f38b3eb1f7b88d13acc9
84671 F20101205_AAALED hetrick_s_Page_020.jpg
92ef0c310a8dbbed55a3388ec062212f
30e552a6f1856a64ef6f1c60004bff5ff2395e71
1134 F20101205_AAAKYK hetrick_s_Page_078.txt
adbb9357a9000140674e2ea9edb9eeb2
80779986e53102a3e3d5476b6daae89335dd88d5
14665 F20101205_AAAKXW hetrick_s_Page_041.QC.jpg
4dc24c834182a31e9777b57ca3af1a54
60e735a2f1ad7e949a36ee0873ac4e41a707cea9
84235 F20101205_AAALEE hetrick_s_Page_021.jpg
8ffb6c0ed3de66feb71091899206b1f0
c95feca96ff1d679edab4b2a0ff74cf5c97f06ad
20359 F20101205_AAALDQ hetrick_s_Page_001.jpg
6d3f2e772e185a58316e2aa617826fc1
c35c69e97597da185dea020565fc7fae237a2abb
6682 F20101205_AAAKYL hetrick_s_Page_071thm.jpg
f37aeb7222a4c6ce5535f4aecc22b8e3
4c6658473cf944761b10ffebbc0c9714c59acb39
89287 F20101205_AAALEF hetrick_s_Page_022.jpg
2a4db41091336d0af5bf1019e00952b0
1da5c850947be9797f01bd88611de30a85d5b26e
4316 F20101205_AAALDR hetrick_s_Page_002.jpg
3bda6bb42010be849625270a1ea88357
e25ed7f52907200dfd3b650ec93f35d114604f96
F20101205_AAAKYM hetrick_s_Page_062.tif
965589c0d8256754fade3e91897fcf5b
3eae63287531b97c17271e0911164a74017a8ff8
28038 F20101205_AAAKXX hetrick_s_Page_067.QC.jpg
d4b6d919bb73397c219497fd9a2052bc
bd8e0aa1674cfe953e947299f3f3cd2f60d25e78
89066 F20101205_AAALEG hetrick_s_Page_024.jpg
fd15905c83da6c5f1254a311f3a48137
5a403eeb5aedc34a417934c04a541f906779d188
86955 F20101205_AAAKZA hetrick_s_Page_112.jpg
7979d6465d7e788c336460ffad939720
f4388c25cfcc6017bfec9a744916c92230ad3e97
64387 F20101205_AAALDS hetrick_s_Page_003.jpg
76bc6e6c6fe1f4641a558f84dab74820
7bf861690d40a05983c4c96c8b69cab6a098e717
F20101205_AAAKYN hetrick_s_Page_107.tif
577796e86ddf74b3f50be5320403e1ee
eb6835452a5ef3d53b485ebc8c059920ad51db77
6558 F20101205_AAAKXY hetrick_s_Page_019thm.jpg
2a4772f6b286a1a03967c652c4c9a27c
fc4e9934f0f6ca8308e2147bebd6ff0a7ad6f59d
88039 F20101205_AAALEH hetrick_s_Page_025.jpg
efee6bcd153d0cea1ad30657e9c72a4d
233e8d5469628a65d3f8d7199c3d62f456ab6d02
264711 F20101205_AAAKZB hetrick_s_Page_080.jp2
8c637945f6abe20e976406dbd4be7ae9
c9a36fe44dcf3eb99ebc307d62f5aad555bce64c
88671 F20101205_AAALDT hetrick_s_Page_005.jpg
d760512e53df7585bb88580b2c2fa244
6e7738302a7d3952f62a08f13c8b831ac5b414d5
123892 F20101205_AAAKYO hetrick_s_Page_070.jp2
5f5f488dfed45f9f10522bb46d83db03
449cc70707514f001347df01e2e476a986492627
2435 F20101205_AAAKXZ hetrick_s_Page_005.txt
9d35e881f71e87d0f6cae8f207ee3752
d44c0c2417c9a964ba0d3d9e338e2b0c4190f4b9
91184 F20101205_AAALEI hetrick_s_Page_027.jpg
e40aad685b08350a6b9e44a955c4dd82
157596baa5b9e1549aed610d70a1497f1462f465
2232 F20101205_AAAKZC hetrick_s_Page_065.txt
374eacf95114a13cbb341f6ba5453d4f
e49a4096d7edd5b1f1c5ee9ae4cf54feb724e3de
89898 F20101205_AAALDU hetrick_s_Page_006.jpg
9e911163fc2a1ea05f50ed25c6c943be
98549b8be8603c81109850f65df12312cca4e157
309756 F20101205_AAAKYP hetrick_s_Page_082.jp2
a009abb85a8edb82a5118b9323540ffd
439e829d9bc3a01f9b37e6975e9304612341af99
94668 F20101205_AAALEJ hetrick_s_Page_028.jpg
4b889e23a3d15d6c4b875919cc5abdb4
6ac5f2da0af2b4e6f9dc2e3788a1f9ddc5a1d327
18509 F20101205_AAAKZD hetrick_s_Page_087.QC.jpg
c0ebcdf1b1be998e479dc1d69e648ec7
f97500b14c06043918b5dfe74d9e7706ee40e311
52251 F20101205_AAALDV hetrick_s_Page_008.jpg
43aa66156b6fa6a9d5b0a276c88015f2
0c664c6d6f1ceee9e838afcc846345441ceac3ab
120178 F20101205_AAAKYQ hetrick_s_Page_047.jp2
dee4f0327bf7e1073983264470ddfe42
25d879fb16a123d03980f5f45454155621cb10cd
93772 F20101205_AAALEK hetrick_s_Page_029.jpg
ff2081157bd7777452452721376d9a34
77731459411121a2508c8548380cc3ce93f4e652
346928 F20101205_AAAKZE hetrick_s_Page_078.jp2
190322188e7a5d3ff2b9365967c77e0e
a1cb465793d568285c003d71bcd390ddf1cf9afc
84412 F20101205_AAALDW hetrick_s_Page_009.jpg
b6753f39a9774197bee97d1c5f495092
c7421c0768af500a105aa291ae4e384f4455973b
81450 F20101205_AAAKYR hetrick_s_Page_013.jpg
427c57362a752dae089e9f3eaf0163f7
2235ecad61138048efc23b80e0a351c0976af358
87491 F20101205_AAALFA hetrick_s_Page_048.jpg
af681d4b21319e1c42ffba94c2e694e6
3c041fc2b2524926f997a2d94cda3546f7dc9d73
92652 F20101205_AAALEL hetrick_s_Page_030.jpg
61b356d5c68ffd27e8b2387be2928ed6
eb35eea59ee7ac9fbc810bb3e53e315e311daf0a
81813 F20101205_AAAKZF hetrick_s_Page_101.jp2
2d49d72be57810205d4740bc6c9606e8
8bc6647f6b5ab73aebe463dc078690761130b8ee
61959 F20101205_AAALDX hetrick_s_Page_010.jpg
7da14e47c64b069f7474659887814acd
8bed3305b8da01bb21ba908a55d34b4c513f291b
29827 F20101205_AAAKYS hetrick_s_Page_028.QC.jpg
edf571cdf6084d34de8acd8551353cf5
3bbca5bb30d131b18c3f407d63ecd954313fd398
92854 F20101205_AAALFB hetrick_s_Page_049.jpg
90458c8b992b413b5d6c2a32da20036c
d3af3f6609a1cb86302444ef1fd78f5051fae489
91436 F20101205_AAALEM hetrick_s_Page_031.jpg
60cf9a17d21c3d8c03bb8024124a7164
08dc8291a4a87ade2c6e31cfe0bd44270c78e4a8
56533 F20101205_AAAKZG hetrick_s_Page_027.pro
0ac4290ec702853cae5aad1b63dd8daf
ea73d00472f0d65410d7187bffc3fe9ea3d0a637
86240 F20101205_AAALDY hetrick_s_Page_011.jpg
8710980ad9fbb01a5a9a009663398b61
6ee53bf5fa979f8ee4fb9edc05e1da617bd07357
1051983 F20101205_AAAKYT hetrick_s_Page_110.jp2
e6692330d601983636b2903bb6f43420
416c8a5bd986aa2c0007e863814119aac32682a8
84633 F20101205_AAALFC hetrick_s_Page_050.jpg
86e8a7fa09162bc59a39662c54674f88
c944f36923a03c777ff06ce9868078be5a04955e
31880 F20101205_AAALEN hetrick_s_Page_032.jpg
7c8852d71218736f46f24da36f3c6416
9b19d4e70b7d205e50e344c196786d2c8ee3904b
6823 F20101205_AAAKZH hetrick_s_Page_070thm.jpg
a04dcdf647f10a00bc7abc554e25d72c
5a5c4f294a2381e5e843dc3c729e9e731f6b680b
83069 F20101205_AAALDZ hetrick_s_Page_014.jpg
efc2d76609ed5aa4e1176a91a2c4c769
cf5f9e1ec55a72d9d1e75f090c45abb46bc33e81
F20101205_AAAKYU hetrick_s_Page_114.tif
623fbee0ab6a0b780202b21c88484b8e
ad1ec0166c89065e77c4fe85f949fe715b5125a9
89490 F20101205_AAALFD hetrick_s_Page_051.jpg
57f1b1ca02ffe674d2891150b8ade760
b841371b7a99a1ab83198c8af686ee6f3fe1f731
65478 F20101205_AAALEO hetrick_s_Page_033.jpg
176f03294af9727920addf12d933b579
9c0f91407e72f1f80ccb811a31bf0673e8a3540b
986 F20101205_AAAKZI hetrick_s_Page_086.txt
8836402d1e582d85881773b0c93bd90f
1aebc63f7080fddc988665a1a16d2f5d699f0051
1778 F20101205_AAAKYV hetrick_s_Page_097thm.jpg
30b4019b53a76e6411e7e16e6136e212
68be2b7cbf7ea3e3c1000bf7543f03f0732a7776
34116 F20101205_AAALEP hetrick_s_Page_035.jpg
3944b3e9ca3ba8750ca0a1aa06224846
98c240cb2b340e7a99a6935b44774474f214ec09
11056 F20101205_AAAKZJ hetrick_s_Page_073.pro
f93be0f192a2b1b5dd1ba4cbe6b32102
18f89f4edd5e9e2e772c5d9c1bbb661ab4390a48
F20101205_AAAKYW hetrick_s_Page_016.tif
9ce9f99274727a1f6c0363fcf8ca5d68
23891532b7810d8c8432a65a30a8916d036a45f8
92687 F20101205_AAALFE hetrick_s_Page_052.jpg
23a241c70dd4e5a710cef83fbb389c68
c3dae4b24720db08af8b803c615dff3296c69136
25594 F20101205_AAALEQ hetrick_s_Page_036.jpg
c74b67eb1dfab1f8bca459a6e1cb8df6
d947e8220dcd3a3810392b6849478d2d9c35d81a
1051971 F20101205_AAAKZK hetrick_s_Page_005.jp2
a7c668232d07e8208f19f0e779650ff4
5e78a9ba6fe16cb25a5427b87d105830677bebb0
F20101205_AAAKYX hetrick_s_Page_103.tif
b6e91fbe3347bf5f70ed0b62b003e11b
d5d1179c02e3ec3ec5fa7a910bd1060b54dad735
92551 F20101205_AAALFF hetrick_s_Page_053.jpg
d2e18d77227e91dee4e5a1d214009059
fd6b89b20d159226a24cd5c564f83dc15278b5df
34980 F20101205_AAALER hetrick_s_Page_038.jpg
665e50702a4a36a4b68119ebfdfbce44
aa2d891e506a3d47399170b11851024c07f991b0
9733 F20101205_AAAKZL hetrick_s_Page_076.pro
492da49e8fa60c1e36d1ba8d8ad3debb
579ec1e9e86a64776c9a4c3b1adde8448b5f0b3d
87504 F20101205_AAALFG hetrick_s_Page_054.jpg
291491fffe285dfbf35b2654b57205ec
95c0edb188d0c68423d276e7d89fc58bc10e97f1
27333 F20101205_AAALES hetrick_s_Page_039.jpg
1f354d62d2085fe0850b5700ca00a19c
17232b7f7b613ef0c850dd8f464e758496577b72
31180 F20101205_AAAKZM hetrick_s_Page_068.QC.jpg
11ad8e898866fa87e361b92d5ec9ef27
cefe19c161d36c90794a32b049eefc93f69a6dbc
F20101205_AAAKYY hetrick_s_Page_064.tif
400018e2b8b9f247d7f03bc8a0498222
5290b6c1cae6b7305a090706b942f8b85bba123a
90246 F20101205_AAALFH hetrick_s_Page_055.jpg
9ae11a01ed5c69a1c5141c86af5f8741
a2077687fd993d78d7c8aa18d4fe5dd036424670
27255 F20101205_AAALET hetrick_s_Page_040.jpg
9a7dd5cdf8f231eec32e7066c1a94905
f86a1c6be13aaee2d52d4798850c1541dbb8e6c5
9788 F20101205_AAAKZN hetrick_s_Page_083.pro
fedf62f131a4e052a4bff41f0e5ad2c7
a9730244c1a592968f617b4ee99da4c651150e0c
21500 F20101205_AAAKYZ hetrick_s_Page_073.jpg
2141f6cae73b48d8eafe8b7d79afcb9e
e771a9c08a27f3516085cdce76d123ab1f6de102
91240 F20101205_AAALFI hetrick_s_Page_056.jpg
7f1c9ed829d753dace144c8c05c72fac
e3acb5dd6898f95e2c69828be50ef212deface47
47614 F20101205_AAALEU hetrick_s_Page_041.jpg
a4c653fd30dd5b21205c11d6ad5358ed
8e46b3f0a07115770b46e86d6bd1f3934c2de117
F20101205_AAAKZO hetrick_s_Page_021.tif
b9befde4a993e0a7566ca0237a5b2a32
0414ab5dce7e9c3acedc83e3501f2148e6cd07ab
85298 F20101205_AAALFJ hetrick_s_Page_057.jpg
a0ce4b2f0600d1b60391dcd366e33e7c
d8e6d1437047c1f61a83f405946809d15a0bb86c
F20101205_AAALEV hetrick_s_Page_042.jpg
f4ad44778673569597a5dc0d1649788f
8f25cdbcbda1ef55e11aa044672d44825b6c80ee
112 F20101205_AAAKZP hetrick_s_Page_002.txt
b35ef836ef8a261621ec723739eb3626
0549b265c02954e6af78f984954545d1781a38f2
91661 F20101205_AAALFK hetrick_s_Page_058.jpg
57081f8c1d5f326e7961c47e880533c6
2091c83d2a35d4dd4f9d047a4559b18f2e4f8159
89931 F20101205_AAALEW hetrick_s_Page_044.jpg
0ac0e636a4c6a099aaacb5a83b46e28b
c80ba31ff6b44ffb45833c7a81021ef3358f880f
109970 F20101205_AAAKZQ hetrick_s_Page_057.jp2
76cc245b72c3124690475230e9829648
e82d4bdbbaa38f5e9f46bb3ae0d48b596a27cef9
89916 F20101205_AAALFL hetrick_s_Page_059.jpg
fe78154cc01f3ce8e85f93b7946eec23
8cfd47832eae4660065b58c56e548f5ca3073eda
90535 F20101205_AAALEX hetrick_s_Page_045.jpg
19324202c0323801814efda368d5b045
41af1a9553bc274275c864fb03863dfdea099f76
F20101205_AAAKZR hetrick_s_Page_047.txt
a9dc48c5d2130f9c27b6a4095b6151ce
2182e7ca6067a66feef24aa52b84e94572d228e0
24829 F20101205_AAALGA hetrick_s_Page_076.jpg
42ef60ca95a8e343179d965a92939057
570671d5c4aeb0b83fdecdb02828af3ee43b9d32
90479 F20101205_AAALFM hetrick_s_Page_060.jpg
734ba3d04019e7e7f5475df0e276a942
df43842b107846247a865d5f3f71ce8271e362fa
89539 F20101205_AAALEY hetrick_s_Page_046.jpg
a000325e78676b1ea76c83a7dcd96fd8
0834dc59904b0df17a3f78c34dd73d0495b6c89e
F20101205_AAAKZS hetrick_s_Page_111.tif
581b569a89b912e79e06d67f5d2afdf9
3630c1ecda6b3fe94d6fd822da81bcc5de62f6f6
26222 F20101205_AAALGB hetrick_s_Page_077.jpg
5dc7d53543f5ba4c9beb19fa75d94d03
41e5e59064a76fdf5544596747c31ce5b5149791
79833 F20101205_AAALFN hetrick_s_Page_061.jpg
8472797158b182b9aa21c4f875952499
b6aad65823cced18d778ef6440f75cab9d5c3b2c
92434 F20101205_AAALEZ hetrick_s_Page_047.jpg
e942632bc150aae3969eb0e5898fc924
255e0271273129eac3d3dbbf6ca9db484d126d3c
69227 F20101205_AAAKZT hetrick_s_Page_102.jpg
e8a6a5eef94bf559d6063d3095a50d83
3e90992b48930a6fa1b43c524148914aad02a7b9
25870 F20101205_AAALGC hetrick_s_Page_080.jpg
a5a6526cfe53bc2ae4fd258b4eacbc34
76cd77ab0f4136a6b209a383087c5c3525bb795a
92113 F20101205_AAALFO hetrick_s_Page_062.jpg
ac48be435884aea4c70902d193f6806d
f68037e050a4d7846118e9d7279cc89edad1454d
117465 F20101205_AAAKZU hetrick_s_Page_067.jp2
d68bcff1526c7dd739f8e17918a6379e
b3624e2c0d0c5d1bbd13b7b8b3cfc354c17d2882
20979 F20101205_AAALGD hetrick_s_Page_081.jpg
11cc46b38dde2b5d4995d8c9368ea83d
9eeb5b671c059ec85bd06170b564f528953c0c39
86284 F20101205_AAALFP hetrick_s_Page_063.jpg
37b9dd8459197e2e28c46a6000d2bd9f
a17e6c4ec32331329f8e0c4e132750803aede068
123215 F20101205_AAAKZV hetrick_s_Page_064.jp2
476ae823a4f126987dc7354300b2f72d
a01133f441cee98a04a299a0f5f0939fb19870e8
29270 F20101205_AAALGE hetrick_s_Page_082.jpg
840b2594c569af2cc20b035b2c85409b
7d6bbec8128f447e218c412a8146fe23fa74091a
92823 F20101205_AAALFQ hetrick_s_Page_065.jpg
ebfdcd06d643d57d19b710177d5e52d0
154ab33fd5c89283ba7e0d1a7d76deefd45fa7a8
119599 F20101205_AAAKZW hetrick_s_Page_060.jp2
116e75fa13ebe75f5c75a091bbe31bd5
a31c52fc00920572428cafcb5f5db4c00400ba9d
91575 F20101205_AAALFR hetrick_s_Page_066.jpg
df0411b0998609b4d7bbfc6367cff3e3
67aa34f48888cf8f47622e2754017804171e7b60
533 F20101205_AAAKZX hetrick_s_Page_002thm.jpg
a53b894b7ad26cec783cc1554deb713b
7a571c04d68fff3613abeed59f7ce0731a7af442
21359 F20101205_AAALGF hetrick_s_Page_083.jpg
ff4fcf510019ddc22637b621fa7bd8d3
fd3580a17f40d207fc08415138c531903603d1d7
89740 F20101205_AAALFS hetrick_s_Page_067.jpg
25c03201c6d9f6498f239a4faefd9f0f
427c5140446250a7ed923f92145a6dacdfb40024
2307 F20101205_AAAKZY hetrick_s_Page_068.txt
93f271be76a8c52d634beb172aadf8c3
4f09fc8b841d083a30cba8aa087af759c2437b3b
27812 F20101205_AAALGG hetrick_s_Page_084.jpg
056007045c31ac6937329e64674af900
a5099c8e10c802145bb944beabc9f6e51408ceae
98158 F20101205_AAALFT hetrick_s_Page_068.jpg
f7dc43612b6f4b5f744d2576c6e5c235
599ad1177c665a2f0ba7915fea852c2044663e71
29428 F20101205_AAALGH hetrick_s_Page_085.jpg
d0acb539bf5137c3b05a31e8c887a4c7
4ac558cfd33f491f4c38ac10976757248d78d3a7
86980 F20101205_AAALFU hetrick_s_Page_069.jpg
5f0d693716d679ff61ee0856b26d95dd
0679529a903ddc87521393979fbc6cc2a19d3e8e
F20101205_AAAKZZ hetrick_s_Page_112.tif
5b40e97ecc994378dc1c45fab811b78b
7bcb9104db835541661a9ca0ef14bed6877f9434
33569 F20101205_AAALGI hetrick_s_Page_086.jpg
b1d642e771734379c124a43a27a669d1
538876a46d2389e8dfd8a9922c68ea9484bc73cb
94352 F20101205_AAALFV hetrick_s_Page_070.jpg
d87964f2d19178b15704f0141b3fe182
0308103e7a802e0d05dc8cd61c8ae40179559e69
64074 F20101205_AAALGJ hetrick_s_Page_087.jpg
8a971acf182bc8c27ae7dde9b8785ad0
11bcfba5304307d2f138e912ca6d7630083c03b1
91763 F20101205_AAALFW hetrick_s_Page_071.jpg
40e7c2058c07ac56e835ac6d46f2e3e5
c0ad4c8ecf8352768c564cf582cb0b2433e8ac7e
83096 F20101205_AAALGK hetrick_s_Page_089.jpg
ddac3b544e5151d4420ee4f083aa25d1
78d734b0b90fd8c987de1b4d309cc34ac6d99752
94600 F20101205_AAALFX hetrick_s_Page_072.jpg
0c15a679ccfc6cedb1b403d4ddb95005
662d4d50db5b6880aba8a7862c00f5d1add7c678
97456 F20101205_AAALHA hetrick_s_Page_109.jpg
cb4cd54e1466f36f21ad929df636d2cf
ab3dff0f34958613f4a54d718a233eecb36b2e15
85234 F20101205_AAALGL hetrick_s_Page_091.jpg
fbee7e5c4acd520781d98c6b08a83066
4c14f7b69a54720c520e67cb4e269feb00e58708
39959 F20101205_AAALFY hetrick_s_Page_074.jpg
8523775936e75f9475e0116a2e896939
9dceb969bf8ff4b7b25d6c8e37efd6058423b2cb
100312 F20101205_AAALHB hetrick_s_Page_110.jpg
53cfccad50f1c46de1316421c0cd1c31
b5ba66293ca61ad282f3900c014ddc84b5911a40
77846 F20101205_AAALGM hetrick_s_Page_092.jpg
740cd386998f364c9f27af2b5427ae14
5eddbdeb508aff5077399850a6642db919a45c78
105620 F20101205_AAALFZ hetrick_s_Page_075.jpg
468785bc5ee9654fb9c406cae455e7f8
d589aa4128859f6f4539bea82d2e883e4417612f
108205 F20101205_AAALHC hetrick_s_Page_111.jpg
e259895d0b31df7920ef24698f380361
3beb03eecdb596be6f91026e37bfd1f58533df5f
87396 F20101205_AAALGN hetrick_s_Page_093.jpg
15a911e33510ed8e5984f1c4133be6e5
4587a549b6b2489728e4006575cc6e8112d50507
94240 F20101205_AAALHD hetrick_s_Page_113.jpg
437ef31825014c379f402c19c9c12905
bb1b928647f2673c7cd1f4795995d72930718cb0
12807 F20101205_AAALGO hetrick_s_Page_094.jpg
b7eee6f84497618edb43864eaee6897a
007d9e305bd98cde5ee67772922a1c6461f93593
96954 F20101205_AAALHE hetrick_s_Page_114.jpg
e63e190787db637d2fbf058a881d5681
d23342f661956188b63bd5dc2dd9103c34d8b575
35982 F20101205_AAALGP hetrick_s_Page_095.jpg
7b0939925cfd45ee499062932a213eec
6b4a7557b18a48c6e894829cb3b4fc54c842562a
103116 F20101205_AAALHF hetrick_s_Page_115.jpg
03b2e5053ea51dbc882565cd9284027f
c44e358d0eac736aaafa55534232c409530ce778
13040 F20101205_AAALGQ hetrick_s_Page_096.jpg
27982063087dabc09470230d008a8a84
0e045e5c8fc5c93497880b567a18e66eacc6646d
23552 F20101205_AAALGR hetrick_s_Page_097.jpg
c140b6a8dbb8aed1e638101c682fdbe2
35b36deb678d79665c06db04d6b942b65fd48d87
38926 F20101205_AAALHG hetrick_s_Page_116.jpg
636832a0f7a3b731ccf45212b70cd50d
6f4657bc4ac40510722359443285d2f091d5fcdf
67526 F20101205_AAALGS hetrick_s_Page_098.jpg
7736275ebe5e64d59ce6b1436e6acee8
28d93bf68bfc897da5be310d43a08bde4664f6d0
74737 F20101205_AAALHH hetrick_s_Page_117.jpg
8851aebc1e76f9f07512a024c4fd73d3
aafe1cc76d4c10ded39c92f53c3c22b51ba824ad
31964 F20101205_AAALGT hetrick_s_Page_100.jpg
edf9cc98f091efd4b33b42693e2d8c13
d313779e4ea1eb1edd14c18b9b3e546efa4de2ce
5960 F20101205_AAALHI hetrick_s_Page_002.jp2
8aa1e6c1dd94d2f5df1e640c9c713b9a
55848f5f21e312764c1c0f15228220b208270699
43303 F20101205_AAALGU hetrick_s_Page_101.jpg
f34696363bfe03314f054b95ff175f0e
267b3aadc4cb4cc8a9855e6a5251b6681da8b5d9
81801 F20101205_AAALHJ hetrick_s_Page_003.jp2
97f55c80ae819cd7edc3b351dc6f2d9b
7a1952ada4365ee9a8a76a5be2f3ea261e37d9bc
70167 F20101205_AAALGV hetrick_s_Page_103.jpg
8ddbb5526b3548275284397c7216931d
2ddace77f73eae61b87c4b21371ac939ac9bccea
1051982 F20101205_AAALHK hetrick_s_Page_006.jp2
68a419cb00fd9bacd5c441f84103b3da
2e68f6a7630a6862c86f02565de11275635ca16d
68040 F20101205_AAALGW hetrick_s_Page_104.jpg
76479c25eb69d21d473404b4240d8417
65b011b5cd65d272c7e270504312a298286ffe8e
112681 F20101205_AAALIA hetrick_s_Page_025.jp2
0281866a952024555c1b455cb5f72443
399e470629b3db7002a31aeace77c4e5eaeb2e2d
1051986 F20101205_AAALHL hetrick_s_Page_007.jp2
da06b1f68c5f08f15b297655ab059773
fbf9c8179a09abb71487deb0f6d06924f0b9775b
61883 F20101205_AAALGX hetrick_s_Page_105.jpg
b5940029970e0aba12f56eaa193ec12e
00944a45ac8ebfba4d66a6af669867cd936c7819
120798 F20101205_AAALIB hetrick_s_Page_026.jp2
810cfc06f55f7ec20602198496523ddc
56570e4182e8458fb49bbe02f9dea38d0fdf0e1c
1051949 F20101205_AAALHM hetrick_s_Page_008.jp2
9cccd6882217c882ccb9beb29698ed7e
16646be03edc7c8fea2db879585917f52bd66fe6
65471 F20101205_AAALGY hetrick_s_Page_107.jpg
26ff52cf206d955b6155797cc6013f43
6fc83b0f185e1d54a21de8e70bb3493657db7b50
118696 F20101205_AAALIC hetrick_s_Page_027.jp2
6c8994d706ae9abc35c557f5ae593ab4
e0b599c5ffbcd789a2bd80ce22d3cfa50220a528
109370 F20101205_AAALHN hetrick_s_Page_009.jp2
8d9cdfa4ef90193dde496b78cc6763da
da0c85642f5a8910eaa76064651660f732cb58df
28622 F20101205_AAALGZ hetrick_s_Page_108.jpg
61fe61c3c9c7b6641860ef674d36a41e
d44579b67e6e4c0eae7401730ba468bc03870af5
122570 F20101205_AAALID hetrick_s_Page_028.jp2
6698109712dc9aaf4bbb61bc9bc7b8fe
4695b4afd8125e324730b88a56cdf0ec15d20821
102693 F20101205_AAALHO hetrick_s_Page_012.jp2
d645c2d1f30bd4314c1c2e4c66e7c8b0
b043328971b3d98639d80d557f66cf9d463d298e
122498 F20101205_AAALIE hetrick_s_Page_029.jp2
46e2b8719444ab7e243ac174f1bc768a
5835167acb297966406daf0741372498bbcbbc34
103633 F20101205_AAALHP hetrick_s_Page_013.jp2
b996bb4c2f8aa8e74d04da982b063c14
c11f74987a087c3b3e21f488d032a5afefbe0b99
122663 F20101205_AAALIF hetrick_s_Page_030.jp2
d31ad401cc219125dba244b81d5da39b
fa0d129f91ede847a9a51c9bf87ff6f4c0ebb59a
110072 F20101205_AAALHQ hetrick_s_Page_014.jp2
a2ec4ef40112fbccfc01281a27bcca9e
3de884465358025e990edc605560ef4a2a610329
117996 F20101205_AAALIG hetrick_s_Page_031.jp2
8c16707047879964096f0879109d7245
800cefbc86695d41d8da3e6b9cec694708dd4570
124047 F20101205_AAALHR hetrick_s_Page_015.jp2
a2bc0ea3e454673fa295decd6bc88cec
4587012dc8aa6d63d760c47916cccfaabb8ea704
119537 F20101205_AAALHS hetrick_s_Page_016.jp2
9288c9576375375981ac156dc5dffcf9
c03edf63e9909a12a978a7b8c2f7c1352bd2f5df
1051947 F20101205_AAALIH hetrick_s_Page_033.jp2
6143358ae86949e5a3ebbadd8acde7c4
2eb593459ee47d4a117049165785205054aca0c5
114390 F20101205_AAALHT hetrick_s_Page_017.jp2
738e6e30b21e26da11f47091024a33a6
4f492b05379add2e804a46cb857819bbea99d0f5
683980 F20101205_AAALII hetrick_s_Page_034.jp2
2f31bddae9d71dab1157cc03f26c1669
5bf9d102b2a75686633103937dcb77c848192699
123137 F20101205_AAALHU hetrick_s_Page_018.jp2
4931b16b023cb106c080efb9e5f7cb83
14ad4c685b546e6865439e7b41198e4ab3c1e848
355197 F20101205_AAALIJ hetrick_s_Page_035.jp2
9dc163114896d7514225f45c1d480191
fc3933eed3dbf0283e14e209c9df9615a6da50b3
111388 F20101205_AAALHV hetrick_s_Page_020.jp2
a16ea13d3b0ae55dd590e55042a4f381
efecba93f887813e72808d0903da04f6cead1de6
109046 F20101205_AAALHW hetrick_s_Page_021.jp2
85e546521f2c314b4b9fa0b04a2710a3
04d3aa117171b0f0f4c92265a2049e5f4948d141
127826 F20101205_AAALIK hetrick_s_Page_037.jp2
5820d5e369f24459af5b2ff9348674a4
832a528a0eb346b24fb9a27383f03d4bc9670c86
116119 F20101205_AAALHX hetrick_s_Page_022.jp2
8013b84e861b11bed09136d2416f400a
95a4d54aa5ae6779b0642e30a33140c4aca0af7c
117613 F20101205_AAALJA hetrick_s_Page_056.jp2
5d90b33f96ffb4c48bfe4b4c44dc4d33
ad0b310839878093597470305fd676e8349b05a8
280359 F20101205_AAALIL hetrick_s_Page_039.jp2
a10bd326c1cc67e9fdc40da2aede148d
7077c2120e73515f405f44f59b43464386bb2738
116470 F20101205_AAALHY hetrick_s_Page_023.jp2
f795f4b881aadae36c062160e65aa2c2
b83a31bc3115db5a044812e2fb2d0fd685685ade
119702 F20101205_AAALJB hetrick_s_Page_058.jp2
43a4999f8f2e868236b933a42cf2338d
df133086629b454ffb903e453892f65064f25d6d
30585 F20101205_AAALIM hetrick_s_Page_040.jp2
76847e13f9d2efdd3c4ea7f103d531ca
e02b0cda63ea5ef188e6ded6e78230ed16e1c6a6
114957 F20101205_AAALHZ hetrick_s_Page_024.jp2
541b6a973cebf6546aafa9b95d90d276
72523d12e82a0ea6640417d629fe34cb359417fd
116073 F20101205_AAALJC hetrick_s_Page_059.jp2
57c9addcfe1f00b6607ac90388789bef
29e052c2f3edbf54594d41c3879611d6589e0ddd
56453 F20101205_AAALIN hetrick_s_Page_041.jp2
e879dc9ec2ca3121faecd35f478cd15b
9ae6c116ffad69ac8fc347b2a26df864a3c7e1be
103419 F20101205_AAALJD hetrick_s_Page_061.jp2
4579674791715dd7f2d73cc02b58f5ce
5f708f96fd179384f5da19628c4dccadfff7257d
118955 F20101205_AAALIO hetrick_s_Page_042.jp2
06b4b627a98fddf6d741dfbc77339e77
7318dcccdd9fc63d2e74459184da2dde3996d2e1
120017 F20101205_AAALJE hetrick_s_Page_062.jp2
8183673e243ebd8cbe7312aef5dbb95c
67e2eebbfc4588d05e04a1ed7cd01a4d20f4e4b2
116102 F20101205_AAALIP hetrick_s_Page_043.jp2
32112d42a2aace4c7464bf31a32aa812
e625ab089b757d303836342ec06e3fdba4106c58
111809 F20101205_AAALJF hetrick_s_Page_063.jp2
28c64c88299ad1f746425bffa34f1526
7e269ce27003cf5abb7c7eec7da9ebcce7e3cb8d
118545 F20101205_AAALIQ hetrick_s_Page_044.jp2
8ecf7927ff641f94a29a247dbb02b8ec
a25b9fed4a9ffc3a565858d9d642ef988c58e917
121988 F20101205_AAALJG hetrick_s_Page_065.jp2
efbd84a367d916943232d581cafbc490
9766feab8a6ed4576b589ca9a1e16b16f8d10448
120143 F20101205_AAALIR hetrick_s_Page_045.jp2
e7bc6dd6d7858e8663854cf542348e12
b16e74a92cf497deb3a8674f1c0fb55a10277edf
119069 F20101205_AAALJH hetrick_s_Page_066.jp2
0d95f9704927984a7cff7342cb3e8c72
90f53600487e2d273e8d81d1b9bde3f1d4ff4fbd
117451 F20101205_AAALIS hetrick_s_Page_046.jp2
26d4363b7b8778ded58216b50c1ae765
c148ca06ddf698a950c9da331bc65634b9f0fc18
115324 F20101205_AAALIT hetrick_s_Page_048.jp2
3aac553df6907d1d1db85714a973126e
da10b85d848e4fc41690ff133af05ebd6f999f60
115671 F20101205_AAALJI hetrick_s_Page_069.jp2
3628212a65bcc7195dd082716a47f4c7
424223bae9c169a1febbf18e1d30255822f62afa
108142 F20101205_AAALIU hetrick_s_Page_050.jp2
688bc8ba01b736c19d6434a86500def8
0d8581f2c144178b6baea2ab3b32a4e44ccc1169
119064 F20101205_AAALJJ hetrick_s_Page_071.jp2
d6a50dd806c2dfb1072edab594576322
1c8042a31aad2781a03ac155e3b72c563bb2883a
115945 F20101205_AAALIV hetrick_s_Page_051.jp2
0fc816d10bfcb1e0cd8a73f757deb3aa
df976c34ab5502f2e2635ae481388f9f76fbd288
122697 F20101205_AAALJK hetrick_s_Page_072.jp2
8e81e9f3e749d2bdc722fbfe299e1527
b8f7eeaefefd14177bc0c5cb8e435bf40260ee95
118991 F20101205_AAALIW hetrick_s_Page_052.jp2
4d4a8b3ac3451baa314790073820f2df
7464b6a9917443cb86c42d1c6b4f18d2ed80b6bc
16549 F20101205_AAALKA hetrick_s_Page_094.jp2
15de05e4831645f715dceede8c7d8f3b
8884ea2548703534251bd4713925ed2d4b1f619a
27540 F20101205_AAALJL hetrick_s_Page_073.jp2
52f0666985b1e59a3fd80c9ab24475a1
f0635f4db2c96f07dc34cae8053c738d4253036c
117385 F20101205_AAALIX hetrick_s_Page_053.jp2
f6354e33eb1c8a03290a476db11fd203
12afdc00834aa188e2237592f1c5a37e56ab2eab
660273 F20101205_AAALKB hetrick_s_Page_095.jp2
ec120b49b992947850fc5ca2dc05da7f
e5bf750294ebfb58de7c93d2d971da5df0b360b2
1051746 F20101205_AAALJM hetrick_s_Page_075.jp2
2472010beee68bf8f075d0d84b75bb5d
4d7ee34ab4166dd747bcc48eeb9c0925f7ffe151
114329 F20101205_AAALIY hetrick_s_Page_054.jp2
0f891b53f16d4be912bdce4ec592042e
56a42f04ce014d9a1dc0bc1bd7547600495aff5f
13815 F20101205_AAALKC hetrick_s_Page_096.jp2
8c1a15bb8a11146211ca3131a6dbc305
9716a6d492315783515f3a5e40fff58651d850f5
331561 F20101205_AAALJN hetrick_s_Page_076.jp2
1076f7ffcfa463661d145945188dcb81
e5b35554743f5e3fac59d7720375984f134905cd
116194 F20101205_AAALIZ hetrick_s_Page_055.jp2
e4bf066f0da71e86876daf08c8c67d7f
e83dd4bb6f426c7d682b4d8cfbd22500fcd40c3a
42953 F20101205_AAALKD hetrick_s_Page_097.jp2
6508b59dd5bf6d33794fb18e29734109
3fe0173a9ae538d9ac50fab0db3c95147a84eddb
270613 F20101205_AAALJO hetrick_s_Page_077.jp2
2cde3032ec6e31769a598b73de8ef15d
4a636f3967ec988859cb92afbd7aa15024202e36
26061 F20101205_AAALKE hetrick_s_Page_099.jp2
9c2b4192bbc795dbdcdfa33d6ff6cd20
a2dd8ca46efccef07da7b415fdfddddc4486985a
260795 F20101205_AAALJP hetrick_s_Page_079.jp2
d6783c0ff9d3ba357272a498f85e3614
7728259b3b51da32005b05a078e6b74c196decdf
59061 F20101205_AAALKF hetrick_s_Page_100.jp2
0606472948a8bf1579f39c6a822869e6
d09be79c43d87958d650728a36afdd2c585de57e
222066 F20101205_AAALJQ hetrick_s_Page_081.jp2
97160aa5d16157f45e170dd8a19bd14d
19057a3b8465cf8578bb6bf13de5dbaec2615da6
67823 F20101205_AAALKG hetrick_s_Page_102.jp2
40d4b4ba86252ed0374519496d777803
139d9976729b7b13e10c6642abb0410990f22215
230140 F20101205_AAALJR hetrick_s_Page_083.jp2
da5b101949c2d9d063e6bd71391320b4
964eb6ff39c82a254c3f265ddf371dde98f3ead3
71304 F20101205_AAALKH hetrick_s_Page_103.jp2
92b0a50b1dc2b05299ce7513f96a1735
641f537208fcd9d9865f30ccc7f475acf00eb3c5
300346 F20101205_AAALJS hetrick_s_Page_084.jp2
4cd47ec0aeaa210cb194192fbfa103c0
ad066d1b99ddfef58ebd09fe3ab4ab437bbb5837
70067 F20101205_AAALKI hetrick_s_Page_104.jp2
2712bd7080c250da03d8e0e70068c5bb
c494bdef96dea027031ee3543750d0f5a657baac
314192 F20101205_AAALJT hetrick_s_Page_085.jp2
a8bf1005e868bbc95899a95bc3242997
b1008c36770209d5780c3828d859449e5497b301
78345 F20101205_AAALJU hetrick_s_Page_087.jp2
c0411860e8f4e63462aa097eeee01c1c
4a27bd9c0c7451d8ebe1aafad0ab505c405be1c5
63632 F20101205_AAALKJ hetrick_s_Page_105.jp2
ab6fbe46cba01dc72d1d4832604952bf
7c7e984e2220c2583ce622bbcff76537cb28acd1
75739 F20101205_AAALJV hetrick_s_Page_088.jp2
7a24e7a731a23cbb22578d3360839f63
1843ff6f95991a6b218dba78be5df5396ad25889
66331 F20101205_AAALKK hetrick_s_Page_106.jp2
36dc81c6adfb5153a62dfbabd8f14440
2e45e7cfc04c0e75eca376de7ad12970a445ae5f
108956 F20101205_AAALJW hetrick_s_Page_089.jp2
78e98b7ea23b85033d89889c6374ff57
650b4516cc9208fa35137693bb81c5610995ba1b
30510 F20101205_AAALKL hetrick_s_Page_108.jp2
8b99566208a1c6089ede73f2e5f80193
a62d4289896001748dfdffdcd40a8a9a2ce3f54a
110429 F20101205_AAALJX hetrick_s_Page_091.jp2
108d835e44a5faedabb7f4fa8396f99d
b579e6779808b8c6ddd37d48fbd7e5a1214b89dd
F20101205_AAALLA hetrick_s_Page_010.tif
64901ea5cdeffe0c3e1abfaf4af1bcb8
55dc75e8150a845398a54c5810a0d27bf638c180
1051974 F20101205_AAALKM hetrick_s_Page_109.jp2
acd900422bac16215d807b98f7ef98e6
f9105b239d3302e9c3541487b28a2962069614c4
101740 F20101205_AAALJY hetrick_s_Page_092.jp2
9e12c5cd95de818093da043c380dc1a7
f92ad10ccfd8c66d78530737454c9bc56a445582
F20101205_AAALLB hetrick_s_Page_011.tif
96888293e828d44b46ec3922d41d609b
ec399afe40f1ba6d1ee9215f851d2563b9643ff8
1051958 F20101205_AAALKN hetrick_s_Page_111.jp2
bc273c25050033771cce9f7f9bd86ccc
dd5893379f8076b949b01051fd5a146bc7bfde58
113515 F20101205_AAALJZ hetrick_s_Page_093.jp2
f224bc38fff9b155aa5ea5c69720b353
c834ead51ce32692ef6bfa806b1e98326ff098f0
F20101205_AAALLC hetrick_s_Page_012.tif
b5c03ecceeea779826d966fbd46accfc
24a345952525909f1d57adef6cd45293eb560e4d
118263 F20101205_AAALKO hetrick_s_Page_112.jp2
0b57b5a3cc92a0022ed996cbcec91655
7756d8b8bb0eba0358d5f42ae3c18ff7515c8ef9
F20101205_AAALLD hetrick_s_Page_013.tif
e84ddce06bf5f85dc122cdff7d3d3158
83f5f1704b9e4a7b9cfa7de240f32917e1462550
125907 F20101205_AAALKP hetrick_s_Page_113.jp2
71d486652a7764a42fcfb8663519c4ea
0f63aa908388f859db4c550f7e2837512eb8bdc4
F20101205_AAALLE hetrick_s_Page_014.tif
4a97261cdbbbbe170a9ee47e77b00e2e
e959680ec49d0db93a23a12e7e272c567a1b9520
128932 F20101205_AAALKQ hetrick_s_Page_114.jp2
33677c465b0f23588d705f0ee320b076
4d9d45eba1dfdae3de0c26eafeffe585de8afd63
F20101205_AAALLF hetrick_s_Page_015.tif
df98088fe9ec98b655539565eef1769b
6f10a846807fa8acca772b6f1dda759407979156
52927 F20101205_AAALKR hetrick_s_Page_116.jp2
05d95b15281a47e1058e56796bdf0879
6ae75863cb2385aef1056192412694762b066af7
F20101205_AAALLG hetrick_s_Page_017.tif
97f8033d29ad66be80cf21f3be994f81
e69bfba6329b8238fb9e87b92d0bb7a2893b4b65
97891 F20101205_AAALKS hetrick_s_Page_117.jp2
afa5c8da4bf627982062626819650373
7dc7fd90001bb91dcc5d2c64e963cb387316298a
F20101205_AAALLH hetrick_s_Page_018.tif
5fae4a36ebad4f981dc23c8f2d929cbb
9203b9cc2cb364041307a4b8a05944b0d2ebe3a0
F20101205_AAALKT hetrick_s_Page_001.tif
e5350be8f1c58fe1ee5fa42427fd746f
99c906398cae44150b752e66b902cc9f44f6bc5e
F20101205_AAALLI hetrick_s_Page_020.tif
d5c3eb80501c1ac9f1e4863526c095a6
8e34ee160bad0d8f2800fd2faaa4213d656f57bc
F20101205_AAALKU hetrick_s_Page_002.tif
d8dba6fbf9795046c52cd2482b127b92
6244e7a6df0363c488b485b87e1bb0cb362da921
F20101205_AAALLJ hetrick_s_Page_022.tif
61cf9d39148ce1562c42655deb57191b
604b19a616e8f2ef03dd46a43f6f195b891ff1b0
F20101205_AAALKV hetrick_s_Page_003.tif
f2993bd28c02187d4abe6e68bfe49992
d771fb836ed245d5590e6aeddb07bd6175d309cb
F20101205_AAALKW hetrick_s_Page_004.tif
8c4adae05a9d0298a9dbeaab134ac056
ad8c19a540c648f37a57081c11b2137aeeb599e9
F20101205_AAALLK hetrick_s_Page_023.tif
5569b8c1f1d017330396750fb39bc1e2
a75e4a6e51f6c5f6b0c543049b702961c650a3ef
F20101205_AAALKX hetrick_s_Page_006.tif
83599007fa2e1aedd71c951b6bc88a0a
c20bbf51376bee2873eed0f823415d91be842c04
F20101205_AAALMA hetrick_s_Page_039.tif
ad5e55447ee873e4f0033b3ee1de97c6
419c05f5d40d8bc291ddb38a22aee062fc0fe347
F20101205_AAALLL hetrick_s_Page_024.tif
ad9a460a7092c97b7b310d6ba361d19d
48d8a59f00f2efe775bad552ecf716b45b16725b
F20101205_AAALKY hetrick_s_Page_007.tif
66bc021177c604ad2d50d7cfd0f63f28
149ea72536a0f39127349be47e738d41ba80d3be
F20101205_AAALMB hetrick_s_Page_041.tif
a00e50ba77e58f97c2a05a5b4bcdf947
e813231046ee3e5f4098d0c9b87cd67cc5d4f47f
F20101205_AAALLM hetrick_s_Page_025.tif
e450b04b3c91cc42fded083fece02f2c
2ec3f78bac6a86213c12112c481c870a18bae638
F20101205_AAALKZ hetrick_s_Page_008.tif
12ab79d4c2883f6df69a4d41caa4fc49
2e39a4696584cdc026f97ad580458b102d82ac0f
F20101205_AAALMC hetrick_s_Page_042.tif
37c4019be1d4f11ceb805c034f6ce8ef
0e59996b32aff4e649b9c4036d2b272e6e93b4f9
F20101205_AAALLN hetrick_s_Page_026.tif
f8964d072853850fd9d9b3a2eaed264a
0da5d0e25093645753d028ee8830e48f99d9e078
F20101205_AAALMD hetrick_s_Page_043.tif
4b7b98f18ddde6890bd224e023f9e7b5
48878af39e4823149bfb88fdf5b84456789e30fa
F20101205_AAALLO hetrick_s_Page_027.tif
3dc5e3b022601923b68cbdf352d47724
e54da6492250f8eda3b44ba7fda6179a694bf433
F20101205_AAALME hetrick_s_Page_044.tif
3032b0bf0956b38aae6a912a95194950
7f35b7dc4254daa697bcc5d162e58625e9f0257e
F20101205_AAALLP hetrick_s_Page_028.tif
3fb9156aef43edfd45ecb5182edf9f7b
5170c160f1ecff4276ae3edebf5b8e80c43471ca
F20101205_AAALMF hetrick_s_Page_046.tif
1c0762aa6b36400cf2df3768c8fbb12a
865f3eed702e68efdfab00bbf5fe33c617632e85
F20101205_AAALLQ hetrick_s_Page_029.tif
4b5a0c03c55afcd1d18051bb95e376ed
d45762cc847c01276849d2f4bea82fd357106eed
F20101205_AAALMG hetrick_s_Page_047.tif
3e26fb79573b603cfc514b57132f53bb
0387990c440096153bbfaf05091f25eec0c2a61a
F20101205_AAALLR hetrick_s_Page_030.tif
a9588bfe4c8eee174b12bb7932b3f12a
45136dbb0df9ae890ea44012815a7835fe6ae9fa
F20101205_AAALMH hetrick_s_Page_050.tif
4fcb6f7339566f38fa4f33251c977754
9ee18a5b7c4b5bf5473d900a654eb8d2e1edcd3d
F20101205_AAALLS hetrick_s_Page_031.tif
39bf8f7ef62a7aff421d247e1f332e1c
11c45729310450654289db3fb09b1ad18a45fc69
F20101205_AAALMI hetrick_s_Page_051.tif
77a3ce837693987324044d07b7150ed0
0d639e4d4f8ea3b3d4bcb62c1b10d518902a4db5
F20101205_AAALLT hetrick_s_Page_032.tif
e6f0453007d3a2188455129d228494c6
7ffb14840fd065539c214350c424ec86ae3f6ba5
F20101205_AAALMJ hetrick_s_Page_052.tif
6930364da290d350a2b3078285721364
364b875f8d8ba06e2eeaad792cda899da9774c38
F20101205_AAALLU hetrick_s_Page_033.tif
8f637a33c04277340279f8a31fe5b831
8e88cec7270baa9f8c0e2dcff310da2d13445e67
F20101205_AAALMK hetrick_s_Page_053.tif
e6251a8e5361aad0a694a4c0c34d388c
4c4018e70af633721e9afcc767ebbed170194747
F20101205_AAALLV hetrick_s_Page_034.tif
d5b0e235d2f8d4b7fa69a4fdc34566d1
4e342a9d2fefc86e71fa8fa014ef370bd930240c
F20101205_AAALLW hetrick_s_Page_035.tif
309bab87fd7f79617a1983bfe57076e1
431967ff266ac929f4aa7281b622fd03efaa3b36
F20101205_AAALNA hetrick_s_Page_074.tif
a67c3f9844cad28b04ce1722d5183b0d
09ad7f81e04be3ad0b23d604188f5a107d1c6066
F20101205_AAALML hetrick_s_Page_054.tif
6b5d1fdb62e9595691ede5b88eb409d8
d5af065a155a071e75cbeeadcd31524f0ebcd9c0
F20101205_AAALLX hetrick_s_Page_036.tif
a58cbb05c1d71c379304873bdaf13203
a0a54bd2fbdac05539437612c62f60e1637e691d
F20101205_AAALNB hetrick_s_Page_075.tif
7e5fc19541ba4595fa14f6b7ba8b3b04
f0f04e7a95dcd4614d2d3361cf63e4cf4fb65ecc
F20101205_AAALMM hetrick_s_Page_055.tif
cdd87770a8c3c38aacb4d419ceb9ad42
710d442483d53344ab3f95280b0c8506515e1301
F20101205_AAALLY hetrick_s_Page_037.tif
9a9e9ff7498779b777a7dfc843408755
cfd626376fec7d8491df0636035ad550923a3d75
F20101205_AAALNC hetrick_s_Page_076.tif
096e3bc6fc31a07e2b7de7a0b5d1e9c0
2cc9ddc9d69201ec0a41309541da12a653da535c
F20101205_AAALMN hetrick_s_Page_056.tif
2138df450db195dd2dd8c2ee7b8edcde
692d525c6304ca3d9b6a60cc266b80669fa6b128
F20101205_AAALLZ hetrick_s_Page_038.tif
c2e37029dc70018b6a9c9976cf9b0de9
d43cbb3041f3660e1f62aa61c7db2bdbcd3a43a0
F20101205_AAALND hetrick_s_Page_077.tif
1ed0d9bcefe0a5bdbb544e7f7234e1c3
6715f879e816c306940fcb3ea915bf3eed436075
F20101205_AAALMO hetrick_s_Page_057.tif
38744bc5c140ea7af7b6d8c96627fc17
7d4b689e0ca92e37a6d8588e8c9ad944d05f0015
F20101205_AAALNE hetrick_s_Page_079.tif
0ad501e9ffc77d31a823be5db8319b8e
3230f514b79ca5c08a932282f28cef4d47fc83e2
F20101205_AAALMP hetrick_s_Page_058.tif
dc8c280ecee4ecdb2ef2a48d1711bcba
e6db0af7283ea2031743e2952565b115c200e4e2
F20101205_AAALNF hetrick_s_Page_080.tif
5c00ebad42f5c1fa36f529bb338b4028
5343eadea772b6aa2c1ce0b795b7fc4cfeebb10f
F20101205_AAALMQ hetrick_s_Page_059.tif
e5c964bad0ce05dd218d713108abe778
326a045b4209872cf077861882d4f049e88ecc82
F20101205_AAALNG hetrick_s_Page_081.tif
315fa4cde58f3900a07a71d655064c59
b823748127aea2421868486f6ce4d51bf98a1661
F20101205_AAALMR hetrick_s_Page_060.tif
302e96b625d3cd6db6513d2f258ca8b5
7ddc81379dc7ea97458e5c4ba92781dcec58ee67
F20101205_AAALNH hetrick_s_Page_082.tif
d976bedaf8df40d16d8d7f745719f0ef
d54252047bb99b13502dca7922faac8f7b4c8173
F20101205_AAALMS hetrick_s_Page_061.tif
512fe318b5c8a003fa92b8cffe8863b1
456bfda0e70b93c1ab1a573b0c53d485b722f511
F20101205_AAALMT hetrick_s_Page_063.tif
f23c915fc6a7422c2eca53abc48bb66f
02b6928d4403f01a93cbd7a691bdf3c7f411786d
F20101205_AAALNI hetrick_s_Page_083.tif
2044999b9b1cc31326a26757da2c1899
4cb4ff7fd522513124e0f57a20d90cc9a3a1acc5
F20101205_AAALMU hetrick_s_Page_066.tif
1776295e3dd9da71238f9966f2ffcbd4
3dfdcd80a499b211f672c336df64619340bb190b
F20101205_AAALNJ hetrick_s_Page_084.tif
9c5786accfc845fef05b0e504922a87e
ed7fa5a126a9caa89e15881bdee7b37fbf1f1118
F20101205_AAALMV hetrick_s_Page_067.tif
ec5346fa8a3ec2d0797931347c9088a8
585753440379505cf98f0aa627b54bdb4b65deca
F20101205_AAALNK hetrick_s_Page_085.tif
73b19577e277378852a1490406540cb3
eefe8435ed5a1e265c37fe449d248526fd3ec158
F20101205_AAALMW hetrick_s_Page_068.tif
631435ed278877b51173f4867a24ca8f
1c4f8b6e577e3c9fc4eadbdfb2eeee7c5e353e48
F20101205_AAALNL hetrick_s_Page_086.tif
63c9bb0c6ea9451cefacfb635d808e2f
ca5335674d896363120413a188586b42d19a8bf0
F20101205_AAALMX hetrick_s_Page_069.tif
391c253ad5ace0a6edc2c9ca1c42efd6
654901aafe43618927922598936d38b384ceecd9
1054428 F20101205_AAALOA hetrick_s_Page_101.tif
208f44ca38cb4d435c99d2c7876333f0
52d874744d0e9c26c76a22acfeaf35c37d66b583
F20101205_AAALMY hetrick_s_Page_070.tif
f09b60457e7997a37286e3a00f8eb09c
ce8ed27418bdf2f6b1f939e2a405d94a2ad5c5d6
F20101205_AAALOB hetrick_s_Page_102.tif
76d0a8769092a5e752209e6dc32d84fa
07d5cae0b1b9d631e4800d39c0e2956cef03486e
F20101205_AAALNM hetrick_s_Page_087.tif
f99ea4bbe7ff30d063a65c329ce27f9e
62602dec3b78c80fdec352570df30e42e8c4bbfa
F20101205_AAALMZ hetrick_s_Page_071.tif
6c13ac270b184dc607630c3146fff73e
6c7ce9b86d169ec9eef3a8719696997329e87d86
F20101205_AAALOC hetrick_s_Page_104.tif
2bb9cac0d3df3c82970cfb1c178ae73c
18066ae5bdbdd00d14c520385606c28989d6ed55
F20101205_AAALNN hetrick_s_Page_088.tif
8faf42ae1ad05168006b280bb20dfdff
d634f8a4cf8a9e47d1004189281b6c18670767cc
F20101205_AAALOD hetrick_s_Page_105.tif
020ff1fa2269ed69648a90f289711821
574f3bcb295be966e5a46931f814e55ed28ae102
F20101205_AAALNO hetrick_s_Page_089.tif
ca50fbe310fc8e45f7a5f3c25764ee65
4228643f150301b0ea8c33aab456144851e8e337
F20101205_AAALOE hetrick_s_Page_106.tif
853a44cf6a74c7dd85650af5d2961e4c
596d365f9ae8cf187307994a5d9445e8122044c2
F20101205_AAALNP hetrick_s_Page_090.tif
9d5ae23b1a1013c30f045b0da7389c4a
03291620627ca5ed283f4ccc630a22d2f9f1bc48
F20101205_AAALOF hetrick_s_Page_108.tif
8857f6022d682c7b099449eb640dfd01
54c5d9584c815dc159b17ec68b6a3d2e9b5f1f39
F20101205_AAALNQ hetrick_s_Page_091.tif
ad8397b50ec83bf985677531f6a5f1c8
c7467df05c056ed12aa812d10dc838d73765f3ae
F20101205_AAALOG hetrick_s_Page_109.tif
8e4ed2cd2e096b4c07e016bdae62d18f
0f58e5211f598db37f8d8072c4876f79540e0733
F20101205_AAALNR hetrick_s_Page_092.tif
ecbf927fc8608e9bcfeb585b271d2fb0
86416a5bbd0662650fcb3f664fc3a374d9b2ff3b
F20101205_AAALOH hetrick_s_Page_110.tif
629e862554dffea72d135b963024372d
918f7baa5215de0718627cb7b98a131c11f1fe54
F20101205_AAALNS hetrick_s_Page_093.tif
b6a511c7f36b9cae17778e0ba14f2436
795bb15fe83404ea2c9e15b07496573b5a57cc7b
F20101205_AAALOI hetrick_s_Page_113.tif
2f6c80563687968f3eab859598922c06
9396b126ac012347977994a4f03c9db09ba14fbf
F20101205_AAALNT hetrick_s_Page_094.tif
fcb8d31ddfce9b028bc240273000dc6a
96a4e2dc888403de362ee20cf2259ee66c31a4a8
F20101205_AAALOJ hetrick_s_Page_115.tif
cdfd159fc93166caa67e1f6d1b168053
3a5262fcf7769532dcaa9d1627c1f6475e586b02
25265604 F20101205_AAALNU hetrick_s_Page_095.tif
a6093b37792191d200faea724b3de754
355ab260218f441ac2a80da2c84cb2851fac0bc9
F20101205_AAALOK hetrick_s_Page_116.tif
b60abfef2edfa4876cd39b43fac29bc2
231ef0ff5038b882cd4f93903aacb45dfc9d6e72
F20101205_AAALNV hetrick_s_Page_096.tif
898f6fee34de05271631227008143d9c
14e76eb5b2f98f38fd6cd0368b80be6a03c51d2a
8186 F20101205_AAALOL hetrick_s_Page_001.pro
3b9d89775dbd0b59a41ba53550d7a2cd
57a282bfb5be5879556376059eec071353f53550
F20101205_AAALNW hetrick_s_Page_097.tif
96d8b7f5a3093ecc21753fb228b0e977
450394cafa3194b31c87a055133b5fa01cc6d85f
56611 F20101205_AAALPA hetrick_s_Page_016.pro
93df8b9ff9accb2468d4508517b98a8c
ecb8ab6bb539bd2e8da80ba7810afbb2f5bcb1fc
1225 F20101205_AAALOM hetrick_s_Page_002.pro
b066e7d9f8932d213cfaaf726e8ae225
d07b460b6a1065814fbbd241526b57ae973cd788
F20101205_AAALNX hetrick_s_Page_098.tif
13a97cf25dbc32e05595d60e27f54fa5
ffdbdd1e98c8a80412a6f80c882c15b27b57f27e
51560 F20101205_AAALPB hetrick_s_Page_017.pro
cba8c5de8df442514516f3c890ba1c58
f7fa5adc6e525313662ac0bfd122f7a6aa8f1ba8
F20101205_AAALNY hetrick_s_Page_099.tif
7191c0d6d92977fabb983419ae7afea3
3203e266e060f2e11a726db77cff4da734df305f
53819 F20101205_AAALPC hetrick_s_Page_019.pro
236563448d235aa4a893c1472cd4aa76
02c84fbffb0458f2a4a4411e3ecd0a04732b440f
36157 F20101205_AAALON hetrick_s_Page_003.pro
77f7ca1a9f3d78f969247dbdd2b7050c
90d184d9dcc808ccc837a2a58306157334bff46e
F20101205_AAALNZ hetrick_s_Page_100.tif
b288bcc4321dddfe38950f00b7bfd9fc
77c299d42d192625b6930e707fceb70936a81d7f



PAGE 1

1 INVESTIGATION OF TUFTED TITMOUSE ( Baeolophus bicolor ) ANTI-PREDATOR VOCALIZATIONS By STACIA A. HETRICK A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2006

PAGE 2

2 Copyright 2006 by Stacia A. Hetrick

PAGE 3

3 ACKNOWLEDGMENTS I would like to thank my advisor, Dr. Kathr yn Sieving, for her endless guidance, patience, and friendship. I thank my comm ittee members, Dr. Michael Avery and Dr. Steven Phelps, for their advice in the planning stag es and throughout the duration of my study. I also thank Dr. Michael Avery for his generosity in allowi ng me to use the USDA/APHIS/WS/NWRC Florida Field Station to conduct my re search. I would like to ac knowledge the Ordway-Swisher Biological Station for allowing me the use of th eir property. I thank the generosity of Tina Brannon and Florida Wildlife Care, Inc. for allo wing me the use of several raptors during my study. My sincere thanks go to the numerous people who helped along the way: Dr. Thomas Contreras and my fellow colleagues in Dr. Sievin gs lab, Scarlett Howell, Jeremy Olson, Jennifer Teagarden, Rebeccah Scarborough, Travis Blunde n, Tracy and Clint Peters, Leander Lacy, Aletris Neils, Kandy Keacher, John Humphrey, Eric Tillman, Mike Milleson, Dr. Thomas Webber and all those who allowed me the use of th eir feeders for obtaining my study subjects. I also want to thank Dr. Melvin and Fiona Sunquist for their thoughtfulness and inspiration. Most of all, my tremendous gratitude goes to my pa rents, Denise and Gary Hetrick, for making my schooling a priority and for always giving me their constant support in every way.

PAGE 4

4 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................3 LIST OF TABLES................................................................................................................. ..........6 LIST OF FIGURES................................................................................................................ .........7 ABSTRACT....................................................................................................................... ..............9 CHAPTER 1 INTRODUCTION..................................................................................................................11 2 ANTI-PREDATOR VOCALIZATIONS OF THE TUFTED TITMOUSE ( Baeolophus bicolor ): DO THEY DENOTE PREDAT OR SPECIES OR CLASS?..................................13 Introduction................................................................................................................... ..........14 Predator-Specific and Risk-B ased Anti-Predator Calls...................................................14 Consequences of Specific Anti-Predator Calls................................................................15 Study System................................................................................................................... .......16 Research Design................................................................................................................ .....19 Hypothesis..................................................................................................................... ..19 Predictions.................................................................................................................... ...20 Methods........................................................................................................................ ..........20 Predator Presentations.....................................................................................................22 Spectrographic Analyses.................................................................................................22 Spectrum-Based Measures..............................................................................................23 Results........................................................................................................................ .............24 Discussion..................................................................................................................... ..........26 Encoding of Risk in Parid Anti-Predator Calls...............................................................28 Potential Biases...............................................................................................................30 Summary........................................................................................................................ ..31 3 INTERSPECIFIC RISK-BASED CALL SYSTEM OF TUFTED TITMICE ( Baeolophus bicolor ) IN RESPONSE TO PREDATORS.....................................................42 Introduction and Background.................................................................................................43 Anti-Predator Vocal Signaling........................................................................................43 Predator-Specific and Risk-Based Call Systems.............................................................44 Study System................................................................................................................... .......46 The Role of Tufted Titmice in Mixe d-Species Foraging and Mobbing Flocks..............46 Anti-Predator Calls of the Tufted Titmouse....................................................................47 Research Design................................................................................................................ .....48 Hypotheses..................................................................................................................... .48 Predictions.................................................................................................................... ...50

PAGE 5

5 Experiment 1: Titmice produce risk-based mobbing calls that are situationally specific..................................................................................................................50 Experiment 2: Chickadees exhibit in terspecific perception specificity to titmouse anti-pre dator calls...................................................................................51 Methods I: Situational Specificity Hypothesis.......................................................................51 Predator Presentations.....................................................................................................52 Spectrographic Analyses.................................................................................................53 Spectrum-Based Measures..............................................................................................55 Results I...................................................................................................................... ............56 Methods II: Interspecific Per ception Specificity Hypothesis.................................................58 Playback Presentations....................................................................................................58 Spectrographic Analyses.................................................................................................60 Results II..................................................................................................................... ............60 Discussion..................................................................................................................... ..........62 Situational Specificity of Titmouse Mobbing Calls with respect to Risk.......................62 Perception Specificity of Chickadees E xposed to Titmouse Anti-Predator Calls...........66 Titmice Give Interspecific Risk-Based Mobbing Calls in Response to Predators..........68 Potential Functions of Interspecific Risk-Based Mobbing Calls of Titmice...................70 Summary........................................................................................................................ ..72 4 CONCLUSION..................................................................................................................... ..89 APPENDIX A SUMMARY TABLES OF STATISTICAL TESTS FOR RESPONSES OF TUFTED TITMICE TO PREDATORS AND RESPONSES OF CAROLINA CHICKADEES TO PLAYBACKS...................................................................................................................... ...96 B SUMMARY TABLE OF MEAN RESPONS ES OF TUFTED TITMICE TO PREDATORS AND CAROLINA CH ICKADEES TO PLAYBACKS..............................102 LIST OF REFERENCES.............................................................................................................109 BIOGRAPHICAL SKETCH.......................................................................................................117

PAGE 6

6 LIST OF TABLES Table page 2-1 Mean number of each note type given by individual Tufted titmice in response to predator and control presentations.....................................................................................41 3-1 Acoustic (first two columns) and behavioral parameters (3rd column) used in analyzing the response of Tufted titmouse floc ks to high-risk and low-risk predator presentations and cont rol presentations.............................................................................87 3-2 Factor loadings of the 17 behavioral and general spectrographic parameters on the four principal components after varimax rotation. Eigenvalues and amount of variance explained by the resp ective components are given at the bottom of the table....88 A-1 Behavioral responses of Tufted titmouse fl ocks to highand low-risk predators and controls....................................................................................................................... ........97 A-2 General spectrographic measures and meas ures of acoustic stru cture of D notes of Tufted titmouse calls elicited in response to highand low-risk predators and controls....................................................................................................................... ........98 A-3 Behavioral responses of Carolina chicka dee pairs to playbacks of Tufted titmouse vocalizations in response to highand low-risk predators and controls and titmouse seet alarm calls............................................................................................................... ..100 A-4 General spectrographic measures of Caro lina chickadee calls e licited in response to playbacks of Tufted titmouse vocalizations in response to highand low-risk predators and controls a nd titmouse seet alarm calls.......................................................101 B-1 Mean responses, standard errors, and standard deviations of Tufted titmice to predator and control presentations...................................................................................102 B-2 Mean responses of Carolina chickadees to playbacks of Tufted titmouse vocalizations in response to pred ator and control presentations......................................106

PAGE 7

7 LIST OF FIGURES Figure page 2-1 Outdoor testing aviary..................................................................................................... ...33 2-2 Examples of the variation within the main vocalizations of the Tufted titmouse in response to the 4 predator treatments and control.............................................................34 2-3 Number of each type of note or call given in the firs t 5min following presentation by the fifteen individual Tufted titmice..................................................................................35 2-4 Number of chick and D notes per chick-a-dee complex call given by Tufted titmice in response to predator treatments and control..................................................................38 2-5 Mean number of overall song, seet, chick, and D notes that Tufted titmice gave in response to the predator treatments and control................................................................39 2-6 Mean entropy of the D notes of Tufted titm ice in response to predator treatments and control........................................................................................................................ ........40 3-1 Examples of the major anti-predat or vocalizations of the Paridae....................................74 3-2 Outdoor aviary at the USDA/APHI S/WS/NWRC Florida Field Station in Gainesville, Florida........................................................................................................... .75 3-3 Examples of the variation in the chick notes and the less variable D notes in the chick-a-dee call complex of the Tufted titmouse............................................................76 3-4 Closest approach distance of Tufted titmice to the stimuli during the predator and control treatments............................................................................................................. ..77 3-5 The proportion of Tufted titmice that appr oached within 1m and 3m of the stimuli during the predator a nd control treatments........................................................................78 3-6 Number of chick-a-dee complex calls given by Tufted titmice in response to the predator and control treatments.........................................................................................79 3-7 Number of chick and D notes per chick-a-dee complex call given by Tufted titmice in response to the predator and control treatments............................................................80 3-8 Mean number of D notes given by Tufted titmice in response to the predator and control treatments............................................................................................................. ..81 3-9 Plot of 17 behavioral and general spectrographic variables of Tufted titmouse calls in two-dimensional space defined by two principal components..........................................82

PAGE 8

8 3-10 Graph of the results of the Discriminant Function Analysis for Tufted titmice with 4 PCA factor score input variab les generated from 17 origin al behavioral and acoustic variables...................................................................................................................... .......83 3-11 Closest approach distance of Carolina ch ickadees to the speaker s during the playback treatments of Tufted titmouse vocalizations......................................................................84 3-12 The proportion of Carolina chickadees th at approached within 1m and 3m of the speakers during the playback treatments of Tufted titmouse vocalizations......................85 3-13 Number of chick and D notes per chick-a -dee call given by Carolina chickadees in response to the different playback treatm ents of Tufted titmouse vocalizations...............86 4-1 The flow diagram summarizes the vocal a nd behavioral responses of Tufted titmice to predators representing varying degrees of risk and heterospecific responses to playbacks of titmouse anti-predator vocalizations.............................................................95

PAGE 9

9 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science INVESTIGATION OF TUFTED TITMOUSE ( Baeolophus bicolor ) ANTI-PREDATOR VOCALIZATIONS By Stacia A. Hetrick December 2006 Chair: Kathryn E. Sieving Major Department: Wildlife Ecology and Conservation Tufted titmice ( Baeolophus bicolor ) are reported to produce differe nt types of anti-predator vocalizations in response to different predators and they are highly social with other species of birds. The goals of this study included investig ation of the anti-predato r vocalizations of the titmouse and determination of whether these calls c ontain information about the type of predator detected (predator-specific) and whether they contain information about the risk of the situation (risk-based). Additiona lly, I sought to determine whether sy mpatric bird species perceive and respond appropriately to the info rmation about predators encode d in titmouse anti-predator vocalizations. In the first experiment I tested the hypothesis that titmice give predator-specific vocalizations (unique vocalizations denoting the specific predator sp ecies or class) in response to different species, including av ian, mammalian, and reptilian cla sses of predators. Titmice produced a combination of vocalizations in resp onse to the predators and there was no evidence that any particular vocalization denoted a specific predator species or class. Titmice varied the note composition, note duration and note structur e of their chick-a-d ee mobbing calls with respect to predator type, which could indicate that they are pr oducing risk-based mobbing calls (in which the call structure varies as a function of risk). In th e second experiment, therefore, I designed a test to determine with certainty wh ether titmice produce risk-based mobbing calls.

PAGE 10

10 The hypothesis tested addressed whether titmouse mobbing calls are situationally specific (call structure varies according to the situation) in re sponse to predators that re present different levels of risk (i.e., highand low-risk). The results indicated that titm ice produce situationally specific risk-based mobbing calls in response to predat ors by varying their call rate, note composition, and frequency and temporal characte ristics, as a function of predator risk. In a final experiment, I tested whether Carolina chickadees ( Poecile carolinensis ) respond to titmouse risk-based mobbing calls and alarm calls with perception specifi city (calls alone elic it appropriate response in absence of original stimulus). Chickadees varied their behavior, note composition and temporal characteristics as a function of playb ack type in much the same way that titmice responded to the actual predators, indicating th at the chickadees responded with perception specificity. Interspecific risk-based call systems, like the one characterized here, likely play an important role in decreasing pr edation risk in animal social groups and, more generally, the larger community of animal species. Interspecific communication systems represent potential mechanisms underlying positive interactions, such as ecological facilitation, that help to structure and maintain aggregations within vertebrate communities.

PAGE 11

11 CHAPTER 1 INTRODUCTION Tufted titmice ( Baeolophus bicolor ) are nuclear species in mixe d species flocks of birds that form during the winter in eastern North Am erica. In these flocks predators are commonly encountered and titmice are thought to give diffe rent anti-predator vocal izations (in different situations) that are commonly iden tified as either seet calls (also known as the hawk or flying predator calls) in respons e to aerial raptors or chick-a-d ee mobbing calls in response to terrestrial predators. The goal of this study was to investigate the anti-predator vocal behavior of the titmouse and to ascertain whether these calls contain information about the species or class (avian, mammalian, or reptilian) of predator dete cted (predator-specific) or whether they contain information about the risk of th e situation (risk-based), or neit her. Additionally, I sought to determine whether sympatric bird species perceive and respond appropriate ly to the information about predators encoded in titm ouse anti-predato r vocalizations. First, I tested the hypothesis that titmice ha ve predator-specific calls in response to different types of predators. In chapter 2, I pres ent findings of predator presentations (of a hawk, an owl, a cat, and a snake) to individual, captiv e titmice, showing that they do not give unique calls for different predator speci es or predator classes (avia n, mammalian, reptilian), but vary their vocal response to the different predators by altering note composition, call structure, and temporal characteristics of their chick-a-dee calls. This type of response is characteristic of risk-based calls, where the calls vary in a grad ed manner according to the degree of risk. The degree of risk that each predator posed (or that titmice presented with each predator would perceive) in this first experiment was largel y unknown, so I was unable to determine positively whether titmice give risk-based calls in response to different predators. To determine

PAGE 12

12 experimentally whether titmice vary their ca lls according to ris k, I conducted a second experiment, which I present in the third chapter. The results from the first experiment led to the formation of tw o additional hypotheses, which I address in chapter 3. The situational sp ecificity hypothesis states that titmice will vary the structure of their calls according to distinct si tuations, such as encounters with predators that represent different degrees of risk. The interspe cific perception specificity hypothesis states that titmouse calls alone should elicit appropriate respons es from heterospecifics in the absence of the original stimulus. To address the first hypothesis, I presented captive titmous e flocks with highand low-risk predators and analy zed their behavioral and vocal re sponse. To address the second hypothesis, I presented cap tive Carolina chickadee ( Poecile carolinensis ) flocks with playbacks of the titmouse vocalizations in response to the hi ghand low-risk predators as well as titmouse seet calls and analyzed their be havioral and vocal response. In this chapter, I present my findings that titmice do give risk-based calls to highand low-risk predators by varying their call rate, note composition, call structure and temporal f eatures of their chick-a-dee mobbing calls. I also present findings that chickadees respond appr opriately to these calls and to titmouse seet calls by altering their behavioral and vocal response. These e xperiments have led to a better understanding of how titmice communicate about the predation risk environment and how chickadees, and potentially other sympatric sp ecies respond to titmouse vocal signals. These signals may play a critical role in decreasing the predation risk environment for chickadees and other sympatric species in this system.

PAGE 13

13 CHAPTER 2 ANTI-PREDATOR VOCALIZATIONS OF THE TUFTED TITMOUSE ( BAEOLOPHUS BICOLOR ): DO THEY DENOTE PREDAT OR SPECIES OR CLASS? Many species respond to predator encounters with specific vocalizations. Some species have different calls that denote pa rticular predator species or cl asses (predator-specific), while some vary one or more of their vocalizations accord ing to the degree of risk a predator represents (risk-based). In this study, I wanted to determine whether Tufted titmice produce predatorspecific vocalizations in respons e to different predator specie s or classes (avian, mammalian, reptilian). To reliably determine if titmice have calls that denote different predators, I presented captive, adult titmice with four pred atorsa hawk, an owl, a cat, a nd a snakeand a control, all in the same manner. I found that titmice most often produced a combination of different vocalizations, including chick -a-dee mobbing calls (composed of chick and D notes), seet alarm notes, contact (chip) notes, and song, none of which denoted a specific predator species or class. They did, however, vary their vocal respon ses to the different treatments in terms of note composition, note duration and note structure of their mobbing calls The cat elicited the least chick notes and the most D notes per call, foll owed by the hawk and owl, with the snake and control eliciting the most chick notes and fewest D notes per call. In a ddition, the bandwidth and entropy of the D notes elicited by th e hawk and cat were greater than those elicited by the owl, snake and control. These findings suggest th at titmice may be responding according to the degree of risk that the predators represent, rather than the predator species or class, indicating that they may be producing risk-based calls.

PAGE 14

14 Introduction Predator-Specific and Risk-B ased Anti-Predator Calls Many animals give specific vocalizations when they encounter a predator. Some species give predator-specific calls in wh ich different call types are used to denote particular predator species or classes. It has been well documented in the literature that many primate species give structurally-distinct alarm calls to different types of predators (Sey farth et al. 1980a, b; Macedonia 1990; Pereira and Macedonia 1991; Zuberbhler 2001; Kirchhof and Hammerschmidt 2006). These calls have also been documented in some carnivore species, namely the suricate ( Suricata suricatta ), a social mongoose, which gi ves distinct alarm calls to terrestrial predators, avian predators and snakes (Manser 2001). Domestic chickens ( Gallus domesticus ) also label predator classe s by giving qualitatively diffe rent vocalizations to aerial and terrestrial predators (Gyger et al. 1987; Evans et al. 1993). In contrast, some animals vary their vocal res ponse to predators according to the degree of risk (also called response urgenc y) that the predator poses. Some species vary the rate in which they call while others vary the quality of the calls they produce. Several species of marmot ( Marmota sp. ) vary the rate of their cal ls as a function of risk (Blumstein1995a; Blumstein and Arnold 1995; Blumstein and Armitage 1997a). In particular, yello w-bellied marmots ( Marmota flaviventris ) increase their call ra te and potentially give calls with a larger bandwidth in response to higher risk predator situations (Blumstein and Armitage1997a). When presented with different degrees of risk Mexican chickadees ( Poecile sclateri ) vary the pitch of a single kind of alarm call according to the degree of risk (Ficken 1989) and Black-capped chickadees ( Poecile atricapilla ) alter their call rate of chick-a-dee calls (B aker and Becker 2002).

PAGE 15

15 Consequences of Specific Anti-Predator Calls Call specificity with respect to predator type might be adaptive if an animal has predators that require different escape reactions. For example, in vervet monkeys, which have both terrestrial and aerial predators, they run up into trees when they hear an alarm call denoting a terrestrial predator and they look up, run into de nse bush, or both when they hear an alarm call denoting an aerial predator (Seyfa rth et al. 1980a). In social sp ecies with different types of predators, having predator-specific calls may allow group members to re spond appropriately to predator threats even if they themselves have not detected the predator. Contrastingly, riskbased calls may give group members an indication of the threat level, but they do not necessarily contain the predator-specific information that would allow for different specific escape responses. Many passerines are said to have different calls that are given to aerial and terrestrial predators (Marler 1957). If these calls are, if fact, predator-sp ecific calls that label different predator classes (i.e., aerial and terrestrial; avian, mammalian, and reptilian), this would lead to receivers being able to choose specific escape res ponses. But if these call s denote the degree of risk that is associated with the different t ypes of predators, receivers would not have the predator-specific information need ed to choose a specific escape re sponse. In animals that live in stable groups, as some mixe d-species flocks of birds do, evolution might favor calls that provide predator-specific inform ation to other flock members. Additionally, some anti-predator signals may af fect detected predator s in different ways, with some call types being potentially more e fficient at deterring or distracting different predators (Naguib et al.1999). On e possible function of anti-predat or calls is to signal to the predator that it has been detected and receptive predators might terminate the hunt rather than expend their energy pursuing prey that are awar e of its presence (reviewed in Smith 1986). Evidence that supports this idea comes from obs ervations by Morse (1973) of foraging accipiters

PAGE 16

16 that did not typically attack if tit flocks gave alarm calls but would attack before they called, although the specificity of the calls is unknown. Australian honeyeaters ( Phylidonyris novaehollandiae ) may be communicating with predators by producing loud aerial alarm calls that likely deter attacks from the predators by in forming them that they have been sighted and that the prey birds have alr eady gone into hiding (Wenzel 1997). It has been suggested that another important functi on of anti-predator calls specifically mobbing calls could be to warn predators that they are about to be harassed and therefore, sh ould retreat before they suffer potential injury from the mobbers (Frankenberg 1981). The use of different types of calls by prey species in response to predators may serve different functions and may play an important role in influencing the subsequent behavior of the detected predator. Therefore, it is important to understand these an ti-predator calls because they may be a key factor in decreasing the predation risk environment for signalers a nd receivers. Predator-specific calls, in particular, are likely to exist in syst ems where animals are in stable groups and have predators that require different escape techniques. In this st udy, I address whether the Tufted titmouse, a common passerine that participates in mixed-species flocks that share different types of predators in common (avian, mammalian, rept ilian), possesses predator-specific calls in response to these various predator types. Study System The subject of the present study is the Tufted titmouse ( Baeolophus bicolor ). The titmouse is a common songbird in deciduous forests in easte rn North America and is a regular visitor at bird feeders, especially during the fall and wi nter. Titmice are year-r ound residents in the study area of North-central Florida and participate in mixed-species flocks in winter (Gaddis 1979; Farley et al. in review). These flocks typica lly contain one or more Tufted titmice, Carolina chickadees ( Poecile carolinensis ) and usually include several atte ndant or satellite species.

PAGE 17

17 Most flocks contain a pair of titmice, their o ffspring, and/or other unrel ated juveniles (Pielou 1957; Brackbill 1970). Regular satellite spec ies include: Black-and-white warblers ( Mniotilta varia ), Downy woodpeckers ( Picoides pubescens ), Ruby-crowned Kinglets ( Regulus calendula ), Blue-headed and White-eyed vireos ( Vireo solitarius V. griseus ) and Blue-gray gnatcatchers ( Polioptila caerulea ; Farley et al. in review). Titmice and potentially Carolina chickadees play the role of the nuclear, or focal species, around which mixed-species foraging flocks form during the winter months and the other flock members play the role of satellite species (Gaddis 1983; Grubb and Pravosudov 1994; Greenberg 2000). Nuclear specie s in mixed-species bi rd flocks are generally characterized by behavioral traits that include dominance, soci ality, and a high level of vigilance (Munn and Terborgh 1979; Hutto 1994). In addition, parid nuc lear species act as sentinels by readily giving vocalizations in response to predators, ther eby potentially alerting flock members of danger (Gaddis 1983; Dolby and Grubb 1998). Typical predators of titmice include feral and house cats ( Felis domesticus ), hawks, owls, and snakes (Bent 1946). In my study area, the mo st common predators of forest passerines in winter include Sharp-shinned ( Accipitor striatus ) and Coopers hawks ( A. cooperii ); Eastern screech-owls ( Megascops asio ); Red-shouldered ( Buteo lineatus ) and Red-tailed hawks ( B. jamaicensis ); American kestrels ( Falco sparverius ); and feral and house cats (Sieving et al. 2004; S. A. Hetrick, pers. obs.). Snakes, most commonly rat snakes ( Elaphe sp. ), typically prey on the eggs, nestlings and sometimes adults of titmice and other small birds during the summer months (Jackson 1978; Halliday and Adler 1986; S. A. Hetrick, pers. obs.). During the winter months in eastern North America, Sharp-shi nned hawks are most likel y the most important

PAGE 18

18 predator of small woodland birds (Bent 1937; Morse 1970; Bildstein and Meyer 2000), including titmice and other flock associates (Ga ddis 1979, 1980; S. A. Hetrick, pers. obs.). Two main types of anti-predator vocalizations have been described for Tufted titmice: the seet alarm call (also known as the high whistle, see-see-see, hawk call, and flying predator alarm call) and the mobbing or scold call (known as seejert, chick-a-dee; Di xon 1955; Marler 1955; Gaddis 1979, 1980). Titmouse mobbing calls are va riants of the chick -a-dee call, which is a complex call composed of combinations of introductory chick note s, and subsequent D notes (dee notes, churr notes), with the number and presence of each not e type being variable (Latimer 1977; Hailman 1989). The broadband stru cture of the D notes in the mobbing calls causes them to be easily localizable, while the pure tone structure of seet calls causes them to be difficult to locate (Marler 1955). When mobbing calls are given in response to predators, many birds are attracted to the area a nd may harass the predator, whereas seet alarm calls result in the cessation of movement (freezing) by the caller and nearby birds or in rapid escape to cover (Gaddis 1980; Ficken 1989; Ba ker and Becker 2002; Howell 2006; S. A. Hetrick, pers. obs.). Although many studies have focused on aspects of the vocal repertoire of birds in the family Paridae, surprisingly little attention ha s been devoted to the voc alizations of Tufted titmice, especially in th e anti-predator context (D ixon 1955; Gaddis 1979, 1980, 1983; Hill 1986). It is not known whether titmice have specific vocaliza tions for different predators, although it is commonly stated that members of the family Paridae give seet or hawk calls to raptors flying overhead (Marler 1955; Latimer 1977; Harrap and Quinn 1995). Titmice could be giving seet calls to (1) denote the predator as a hawk or accipiter (predator-specific calls with respect to species); (2) denote the class of predator as avian (predator-specific calls with respect to class); (3) denote the class of predator as aeri al (predator-specific calls with respect to class);

PAGE 19

19 (4) signal the immediate degree of danger (risk-based calls); or (5) any combination of these. No study has yet to confirm whether titmice use the s eet call to reliably denot e accipiters, aerial or avian predators, or to denote a situation where th e titmouse is in immediate risk of attack. As Templeton et al. (2005) noted, prev ious studies have presented aeria l and terrestrial predators in different ways (Greene and Meagher 1998; Blumstein 1999b; Le Roux et al. 2001), which confounds the interpretation of re sponses given to predator type versus risk situation (i.e., predator proximity, location, or behavior). In or der to determine if the titmice were able to distinguish between predator species or betw een avian, mammalian, and reptilian predator classes, all predators were presented in the same manner in this experiment. Research Design The purpose of this study was to investigate th e anti-predator vocal be havior of the Tufted titmouse; specifically, whether titmice have specific vocalizations for different predators (predator-specific calls). This study dealt with the vocal responses of titmice to both stuffed predator mounts and live predat orsavian, reptilian and mamma lian. Wild-caught birds were tested in a captive situation w ith both the bird and the predator being in cages. I presented individual adult titmice w ith 4 different predators: an owl, a hawk, cat, snake and a control, all in the same manner. Hypothesis I tested the hypothesis that titmice have predat or-specific calls (call type uniquely covaries with predator species or class). In order to ha ve predator-specific vocalizations, the titmice must give different call types that are predictably associated with differe nt predator classes or species (Blumstein 1999a). Titmice could separate the predators into classes in various ways. For example, they may separate them into classes according to whether the pr edator is aerial or terrestrial; whether the predator is avian, mamma lian, or reptilian; or, they could be more

PAGE 20

20 specific and separate them into species such as ha wk, owl, cat, and snake. In this experiment, I tested the latter two of these possibilities. I conducted a true experiment under controlle d conditions by presenting four different predator species in a similar manner to individu al adult titmice to obtain vocal recordings of titmouse calls in response to the presentations. Subsequent analyses of the recordings were conducted to determine whether titmice give specif ic vocalizations in re sponse to the different predators. All predators were presented in a clear cage directly across from the cage containing the titmouse. If titmice give predator-specific vocal izations that label either predator species or predator class (avian, mammalian, or reptilian) then the manner in which the predators are presented should not matter. For example, if titmice have specific hawk calls that denote hawks, they should respond with hawk calls to aerial as well as perched hawks. On the other hand, if titmice respond according to the degree of risk, one or more of their calls would likely vary in response to an aerial versus a perched hawk (assuming that the two situations represent different degrees of risk). Predictions I predicted that the titmice w ould respond with different call t ypes to the different types of predator being presented. I pred icted that they would give pred ator-specific vocalizations that are associated either with predator class (avi an, mammalian, or reptilian) or with individual predator species. In the first case, the titmi ce would have different vocalizations for the hawk and owl (avian) than for the cat (mammalian) an d snake (reptilian) and in the second case, the titmice would have unique calls for one or more of the four different species. Methods I examined the responses of adult titmice to f our different predator species and a control. Fifteen adult titmice were caught between N ovember 6, 2004 and January 21, 2005 at various

PAGE 21

21 locations in Florida--the Ordway-Swisher Biological Station in Melrose, the USDA/APHIS/WS/NWRC Florida Fiel d Station (United States De partment of Agriculture, Animal & Plant Health Inspection Service, W ildlife Services, National Wildlife Research Center; USDA lab) in Gainesville, FL and various re sidences in the city of Gainesville. The birds were captured using mist-nets and transferred to the USDA lab. The birds were housed and tested in 0.5 x 0.5 x 0.5m cages containing several branches for perching. They were fed an adlib diet of mixed seed, mealworms, su et, and chopped fruit and vegetables. In order to be sure of acq uiring appropriate anti-predato r responses, I used only adult titmice in the experiment. Chickadees acquire information about predator identity through learning from older birds in th eir social groups (reviewed in Smith 1991). Since titmice have a similar social system to chickadees, I expected that young-of-the-year may not give informative responses in this experiment. To determine if th e birds were adults, I used the molt limit criteria in Pyle (1997) for aging. At least 24 hours before the first predator pr esentation, the cage containing the bird was brought into the testing e nvironment in order for the bird to acclimate. The testing environment was an outdoor aviary (9 x 3 x 2.3m) at the USDA lab that contained numerous branches and snags and was adjacent to forest, providing a semi-n atural environment. Within the large aviary, the cage containing the test subject was placed on a 0.5m platform with the predator presentation cage 0.75m away from the test cage on a 0.5m plat form (Fig. 2-1). The predator presentation cage was made of clear plexi-glass and a sheet cove red it at all times exce pt during the tests. A camouflage blind containing the researcher and a microphone were also within the aviary, 4.56m away from the test and predator cages.

PAGE 22

22 Predator Presentations Each bird underwent a series of 5 different treatments after the accl imation period. Each test consisted of one of 4 predator tr eatments a stuffed Sharp-shinned hawk ( Accipiter striatus ), a live Eastern screech-owl ( Megascops asio ), a live domestic house cat ( Felis domesticus ), and a live red rat snake ( Elaphe guttata ) or a control (an empty cag e) being presented in random order (hereafter the treatments will be referred to as hawk, owl, cat, snake and control). The tests began at 0800 and were conducted every 2 hours un til 1600. Each test consisted of one of the predators being placed into the cage 20 minutes before testing began. For the control, the researcher went through the steps as if they were placing a predator in th e cage. Care was taken to ensure that the bird in the test did not view the predator unt il the test began. The test began when the sheet covering the predator cage was re moved, allowing the bird to view the predator or empty cage. Recordings were made for 5min pre-stimulus and 7min post-stimulus; however, only the first 5min post-stimulus were incl uded in the analysis. A Sennheiser shotgun microphone (ME 66) was used to record the vocaliz ations directly onto a laptop computer using Raven Interactive Sound Analysis Software Vers ion 1.1 with a sampling rate of 44100 at 16-bit resolution. Spectrographic Analyses I analyzed vocal responses for 72 out of the 75 presentations for the ad ult titmice (n=14 for cat, n=15 for control, n=15 for owl, n=15 for snak e, and n=13 for hawk); the titmice did not vocally respond in 2 of the pres entations and one of the hawk recordings was lost due to equipment failure. Spectrographic analyses we re performed on the vocal recordings using Avisoft SASLabPro 4.39. To edit out noise, each sound file was FIR low-pass filtered at 12kHz and high-pass filtered at 1.4kHz. The spectr ogram parameters used were FFT=512, Frame Size=75%, Window=Hamming, Overlap= 87.5%. I classified and labe led all vocalizations given

PAGE 23

23 in response to the treatments, which included c hick-a-dee calls, songs, seet notes, and chip notes. The notes in each call, or chick-a-dee co mplex, were visually classified as introductory chick notes or subsequent D notes. In titmi ce, the various introductor y notes grade into each other and are not easily di stinguished into natura l categories; therefore, the introductory notes were classified together as chi ck notes (Fig. 2-2a). On the ot her hand, D notes can be reliably classified (Bloomfield et al. 2005) due to their harmonic-like structure and little frequency modulation. D notes also have a higher entropy and lower frequenc y than the introductory chick notes, and D notes always occur at the end of th e call, or are the only notes comprising a call, making them easily distinguished from introductory not es (Fig. 2-2a). Songs, seet notes and chip notes were also visually classi fied. Songs are unique and can be easily distinguished from other vocalizations (Fig. 2-2b). Seet notes were recognized by being high-pitched (around 8-10kHz for titmice) whistles with a narrow bandwidth that have no sharp onset or ending and cover only a narrow frequency range (Fig. 22c; Apel 1985; Marler 1955). Ch ip notes were recognized as being single-syllable notes that are typically chevron-shaped with a shorter duration than seet notes (Fig. 2-2d). I measured several aspects of the notes a nd calls, including some that were based on measures used in previous studies of parids (Baker and Becker 2002; Freeberg et al. 2003; Templeton et al. 2005). For each treatment, I av eraged the number of each type of vocalization (chick-a-dee, song, seet, chip). For the chick-a-dee calls, I aver aged the number of chick and D notes overall, the number and proportion of chick and D notes per call, and the duration of each chick and D note. Spectrum-Based Measures I also measured several fine-scaled acoustic parameters on the D notes in which I had highquality recordings (not overla pping with outside noise). I measured the spectrum-based

PAGE 24

24 parameters of the notes using a power spectr um with FFT length=512. For the D notes, the parameters were computed at the maximum spectr um of the entire D note (maxpeakhold) and are similar to those used by Nowicki (1989) and Templeton et al. (2005). The parameters were minimum and maximum frequency where the amplitude goes last below -30dB and where the amplitude goes last below -10dB (min and max frequency with the total option activated in Avisoft SASLabPro 4.39), bandwid th (calculated with min a nd max frequency described previously) and entropy. For each factor that was measured, I used uni variate analysis of variance (ANOVA) with the least significant difference (LSD) post-hoc test to conduct pairwise comparisons among the treatments. I transformed the data when appropria te to meet the assumptions of the analysis using arcsin(sqrt(n)) and log( n+1) transformations (Sokal an d Rohlf 1995). All statistical analyses were performed using SPSS 11.5 for window s and significance in al l statistical tests was set at the 0.05 alpha-level. Results Contrary to my predictions, the titmice did not give specific vocali zations to denote the different predator species or predator classe s (avian, mammalian, reptilian). The titmice typically gave a combination of vocalizations in response to the predator and control treatments, usually consisting of chip notes and possibly chick-a-dee calls, s eet notes, or both. Their vocal responses were highly variable (see Fig. 2-3 ae). Chip notes were by far the most common vocalization: 68 of the 74 five-min recordings contained chip notes (n=14 for control, n=15 for snake, n=13 for owl, n=13 for hawk and n=13 for cat ). About an eighth to a quarter of the birds responded with chick-a-dee calls to the control, snake and ca t treatments and about half the birds responded with chick-a-dee calls to the ow l and hawk treatments (n=2 for control, n=3 for snake, n=7 for owl, n=8 for hawk and n=4 for cat) More titmice gave seet calls in response to

PAGE 25

25 the hawk and cat than to the control, snake and owl treatments (n=4 for control, n=2 for snake, n=5 for owl, n=9 for hawk and n=7 for cat). Only a few birds responde d with singing when presented with the predators (n=1 for control, n=1 for snake, n=3 for owl, n=2 for hawk and n=1 for cat). The means of each type of vocalization given to the di fferent treatments are shown in Table 1. Six out of 11 of the general spectrographic va riables and all of the 7 fine-scale acoustic variables of the D notes varied significantly with treatment ( ANOVA p<0.05). In no case did the LSD pairwise comparisons indicate that titmice discriminated each predator from the other and the control. In most cases, the cat, hawk and owl were significantly different from the snake and control ((cat, hawk, owl) (snake, control)) or the cat and ha wk were significantly different from the owl, snake and control ((cat, hawk) (owl, snake, control)). Titmice did not give a significan tly different number of chick-a-dee mobbing calls to the different treatments (ANOVA F4, 74=0.41, p=0.798), although the numbe r of introductory chick notes per call and D notes per call di ffered significantly with treatment (F4, 133=44.34, p<0.001; F4, 133=21.96, p<0.001, respectively). Th e control and snake treatments elicited more chick notes per call and fewer D notes per call than the owl, cat and hawk. The cat treatment elicited the fewest chick notes and most D notes overall followed by the hawk and then the owl. The relationship can be clearly seen in Figure 2-4. The proportion of chic k notes per call also differed significantly with treatment (F4, 133=56.34, p<0.001), with the re lationships between the treatments the same as above. The number of seet, song, chip, chick a nd D notes did not differ with respect to treatment (p 0.157 in all cases, Fig. 2-5). The duration of the chick and D notes was significant with respect to treatment (F4, 134=3.69, p=0.007; F4, 299=20.29, p<0.001). The pairwise compar isons revealed that the chick

PAGE 26

26 notes in response to the cat were of greater duration than those given in response to the control and owl; the chick notes in response to the snake were of greater durati on than those given in response to the control and owl; a nd the chick notes in response to the hawk were of greater duration than those given in response to the ow l (with no other comparisons significant). The pairwise comparisons revealed th at the D notes in response to th e hawk were of greater duration than those given in response to all other treatment s and the D notes in resp onse to the cat were of greater duration than those given in response to the owl (with no other comparisons significant). Of the 7 fine-scaled acoustic parameters that were measured, all were significant. The parameters that had multiple comparisons with a clear relationship between the treatments were maximum frequency where amplitude goes last below -10dB, bandwidth at -10dB and entropy. The titmice gave D notes with a larger bandwid th and higher maximum frequency at -10dB to the cat and hawk treatments than to the owl, control and snake treatments. Additionally, the D notes that the titmice produced in response to the hawk had a highe r entropy than in response to the cat; both hawk and cat treatments elicited D not es with a higher entropy than the owl, control and snake treatments (Fig. 2-6). Discussion Contrary to my predictions, th e results indicate that the ti tmice did not use different call types to label different predat or species or predator classe s (avian, mammalian, reptilian); therefore, they do not po ssess predator-specific calls in these contexts. They gave a variety of notes and call types during the predator presentati ons (chick-a-dee, seet, song, and chip), none of which were reliably associated w ith a particular predator. All of the predator presentations and occasionally the control presentation elicited seet notes from some of the titmice. Dixon (1955) and Marler (1955) have called the titmouse seet call the hawk call or flying predator call. But in this study titmice gave seet notes in response to other predators besides hawks and to predators

PAGE 27

27 that were perched (not flying). The results from this experiment show that titmice do not vocally discriminate (by giving different call types) between predator species when they are presented in similar ways. In addition, th e results indicate that seet al arm calls are given in other circumstances besides just in response to aerial predators. Although the titmice didnt give predator-specifi c vocalizations, they did vary the call and note structure of one specific voc alization, the chick -a-dee mobbing call, in response to the different predators, which is characteristic of ri sk-based calls (Blumstein and Armitage 1997a). One way in which they varied their calls was by altering the note composition of the chick-adee mobbing call. The cat treatment elicited the fewest chick notes and the most D notes, followed by the hawk and owl treatments. The snake and control treatments, which did not differ significantly from each othe r, elicited the most chick notes and the fewest D notes. Gompertz (1961) and Latimer (1977) noted that so me parids decrease or drop the introductory notes of the mobbing call wh ile extending the D note sec tion as the level of risk increases. If this were true for Tufted titmice, then it would lead us to believe that the cat treatment represents the greatest risk, followed by the hawk and owl treatm ents, with the snake treatment representing the same amount of risk as the control. Evans et al. (1993) suggested that, under some conditions, terrestrial predators pose a greater threat than avia n predators. Given that all birds used in this study were captured in or near suburban ne ighborhoods where outdoor cats are common around bird feeders frequented by titmice and other species (S. A. Hetrick, per. obs.), all test subjects would likely have experience with this species. Unlike a perched raptor (species dependent upon aerial attack to capture prey), a crouching cat in close proximity to a titmouse poses a very high risk to the adult bird. Therefore it is quite likely that the titmice in this study could have viewed the cat treatment as repr esenting the greatest risk.

PAGE 28

28 Encoding of Risk in Parid Anti-Predator Calls Note composition may be particularly importa nt in encoding information about predators and degree of risk. Studies of other birds in the family Paridae, namely chickadees, suggest that different variants of the chick-a-dee call mi ght encode information about the presence of different environmental stimuli (including predat ors) or the motivational state of the caller (Smith 1972; Gaddis 1985; Ficken et al. 1994). Hailman et al. (1985, 1987) suggested that note composition variation may encode information re lated to many factor s, including potential predators, and the results from Ficken, Ha ilman, and Hailmans (1994) study with Mexican chickadees supports that view. Black-capped chickad ees altered the number of introductory A and B syllables in their chick-a-dee calls in re sponse to a predator mount presented at different distances, which represented different risk levels (Baker and Becker 2002). They also altered the number of D notes in the chick -a-dee call in response to raptor s of different sizes, with the smallest raptors eliciting the most D notes and th e largest raptors eliciting the least (Templeton et al. 2005). The small raptors represented higher-risk situations to the ch ickadees and the large raptors represented lower risk situ ations, thus it is likely that the chickadees responded according to the degree of risk that they encountered. Besides varying the note composition of the chick-a-dee calls, I found that titmice also altered the structure of the D notes within th e calls. Titmice gave D notes with a larger bandwidth and higher maximum frequency in respons e to the hawk and cat, while the owl, snake and control elicited D notes with a smaller ba ndwidth and lower maximum frequency. Other animals change their call structure in different predator situations. Ye llow-bellied marmots vary several frequency characteristics of their calls, including bandwidth, as a function of distance to certain predators (Blumstein and Armita ge 1997a). White-browed scrubwrens ( Sericornis frontalis ) vary the call struct ure of their aerial trill call by increasing the minimum frequency

PAGE 29

29 (pitch) according to the distan ce from a suddenly appearing pr edator (Leavesley and Magrath 2005). In both of the aforementioned studies, the distance from the pred ator to the subject represents the amount of risk, w ith the closer distances represen ting higher risks; thus, we can conclude that both the marmot and the scrubwren vary the structure of their calls according to the degree of risk that they are pres ented with. In the present stud y, the hawk and the cat treatments elicited D notes from the titmice that had higher entropy than the D notes elicited from the owl, snake and control treatments. Entropy is a meas ure of the amount of randomness or noise a note contains, with pure tones having no entr opy and white noise having the most entropy; subsequently, entropy can be used as an indicati on of note harshness. Latimer (1977) noted that many birds in the family Paridae give calls that are harsher (higher en tropy) as the level of aggression rises. Additionally, Morton (1977) documented that the motivational state of the caller influences signal structure and that ha rsh (high entropy), broadba nd (large bandwidth), low-frequency sounds are associated with aggr essive behavior, whereas more tonal, highfrequency sounds are associated with non-aggressive or fearful behavior. Because the hawk and cat treatments elicited D notes with larger bandwidth and high er entropy (harshness), we can speculate that the titmice perceived a higher level of risk and had a higher level of aggressiveness in response to the hawk and cat than to the owl, snake and control treatments. Thus, I conclude that the level of risk according to titmice may be associated, at least in part, with the type of predator and not solely the proximity or location of a predator with respect to the test subject. The results of both the note composition and D-note structure suggest that the titmice perceived the cat and hawk as having the mo st risk, followed by the owl and the snake and control had the least risk. In addition, more individual titmice responded with seet notes to the high-risk species; and the mean number of seet no tes was highest in response to the cat and hawk

PAGE 30

30 presentations. This evidence further supports the idea that the titmice viewed these two predators as posing the most risk. Apel (1985) concluded that Black-capped chickadees recognized the difference in the degree of dange r between Sharp-shinned hawks and American kestrels by responding with more se et notes (high sees) to the higher-risk Sharp-shinned hawk. It is interesting to note that the titmice did note respond differe ntly to the snake and control in terms of note composition and D-note struct ure, thus suggesting th at the titmice did not perceive the snake as more of a risk than the co ntrol. There are several explanations as to why this occurred. The time of year that the study was conducted was wint er, which could have affected the titmouses response to the snake, as red rat snakes prey on birds and their eggs typically during the summer when most birds are nesting (Jackson 1978; Halliday and Adler 1986; S. A. Hetrick, pers. obs.). Another reason th at the titmice may not have reacted differently to the snake and control could be due to the activity level and movements of the snake. The snake was stationary for the most part during the presentations and was not actively foraging, probably due to the cold outside temperature. On the other hand, the hawk in this study did not move because it was a stuffed mount and the titm ice still responded as if it was a high-risk situation, suggesting that movement was not an im portant factor in determining their response. Potential Biases The responses of the individual titmice to th e 4 predator treatments and control treatment were highly variable (as shown in Fig. 2-3). Some of this varia tion may be due to the existence of social dominance hierarchies within wild ti tmouse flocks (Brawn and Samson 1983). In these flocks, the alpha (most dominant) male possibly c ontributes the majority of the vocalizations in response to predators. This idea is suppor ted by evidence in Pale-winged trumpeters ( Psophia leucoptera ), where the dominant male in the group gives the majority of the anti-predator vocalizations (Seddon et al. 2002). Older, and most likely more dominant, male Willow tits

PAGE 31

31 ( Parus montanus ) give alarm calls more frequently th an females or young males (Alatalo and Helle 1990). In the present experiment, all of th e individuals were adults, but the sex could not be determined in most of the cases. The mean number of chick-a -dee mobbing calls given by t itmice in response to the predators was lower than expected based on observations of natural encounters between titmice and their predators (S. A. Hetrick, pers. obs.). A f actor that could have co ntributed to this could be that the titmice were housed alone and aw ay from other titmice during the predator presentations. A few of the hypotheses as to why animals give anti-predator calls include alerting others and transmitting cultural informa tion about predator characteristics (Klump and Shalter 1984). For either of thes e to occur, the caller must have an audience. Evans et al. (1993) found that domestic chickens ( Gallus domesticus ) rarely give aerial alar m calls unless they have an appropriate audience. My re sults show that not all of the titmice gave anti-predator vocalizations (seet or chick-a-d ee mobbing calls) in response to th e predator presentations. The most common and numerous vocalization given was th e single syllable chip note, which is not typical of natural predator encounters (S. A. He trick, pers. obs.). Moreover, some individuals sang during trials, but primarily in response to th e cat, hawk and owl, which likely represent the three highest-risk predators (Fig. 23). In the present experiment, the birds were held in a cage in an outdoor aviary. There were sometimes other bi rds such as cardinals and towhees in the area outside of the aviary that coul d have provided an audience for th e titmice and wild titmice could occasionally be heard in the far distance. Even so, the titmice may have felt a reduced motivation to vocalize because they were not around their natural flock members. Summary In summary, my results clearly indicate that titmice do not produce predator-specific calls in response to different predator species or clas ses of predators (avian, reptilian, and mammalian)

PAGE 32

32 when the predators are presented in the same manne r. However, results do indicate that titmice may give risk-based calls in response to predat ors due to the fact that they varied their call structure and note structure in resp onse to the different predators. Given that this experiment controlled for the situation under wh ich predators were encountered, it is possible that the type of predator affected the amount of risk encoded in the calls of ti tmice; house cat and hawk elicited call structures typically associated with greater risk than either an owl or a snake. In chapter 3, I investigate further the nature of the risk-based communication sy stem of titmice in response to predators.

PAGE 33

33 Figure 2-1. Outdoor testing aviary showing the test cage containing the test bird on the left and the predator presentation cage on the right (with the sheet cover removed).

PAGE 34

34 A B C D Figure 2-2. Examples of the variation within th e main vocalizations of the Tufted titmouse in response to the 4 predator treatments and control. A) Chick -a-dee call complex showing the variation in introductory ch ick notes and subseq uent D notes. B) Variations of song notes. C) Variations of seet notes. D) Variations of chip notes.

PAGE 35

35 A () Individual titmouse15 14 13 12 11 10 9 8 7 6 5 4 3 2 1Mean # of notes or calls300 200 100 0 song notes seet notes chick-a-dee calls chip notes B () Individual titmouse15 14 13 12 11 10 9 8 7 6 5 4 3 2 1Mean # of notes or calls300 200 100 0 song notes seet notes chick-a-dee calls chip notes Figure 2-3. Number of each type of note or call given in the fi rst 5min following presentation by the fifteen individual Tufted titmice in response A) to control, B) to snake, C) to owl, D) to hawk, E) to cat treatments.

PAGE 36

36 C Individual titmouse15 14 13 12 11 10 9 8 7 6 5 4 3 2 1Mean # of notes or calls300 200 100 0 song notes seet notes chick-a-dee calls chip notes D. Individual titmouse15 14 13 11 10 9 8 7 6 5 4 3 2 1Mean # of notes or calls300 200 100 0 song notes seet notes chick-a-dee calls chip notes Figure 2-3. (cont.)

PAGE 37

37 E Individual titmouse15 14 13 12 11 10 9 8 7 6 5 4 3 2 1Mean # of notes or calls300 200 100 0 song notes seet notes chick-a-dee calls chip notes Figure 2-3. (cont.)

PAGE 38

38 cat hawk owl snake control Treatment type 4 3 2 1 0 Mean # of notes Mean # of D notes per call Mean # of chick notes per call Figure 2-4. Number of chick and D notes per c hick-a-dee complex call given by Tufted titmice in response to predator treatments and cont rol in the 5min following presentation. Both variables were signif icant with respect to treat ment (ANOVA, p<0.001). All pairwise comparisons were significant (LSD, p<0.05) except for snake and control (p=0.647, p=0.442 for chick and D notes per call, respectively). Error bars: +/1 SE.

PAGE 39

39 Treatment typecat hawk owl snake controlMean # of notes20 10 0 song notes seet notes chick notes D notes Figure 2-5. Mean number of overall song, seet, ch ick, and D notes that Tufted titmice gave in response to the predator treatments an d control in the first 5min following presentation. None of the differences ar e significant with respect to treatment (ANOVA, p>0.05).

PAGE 40

40 141 88 34 21 15 N =Treatment typecat hawk owl snake controlMean entropy.60 .58 .56 .54 .52 .50 .48 .46 Figure 2-6. Mean entropy of the D notes of Tufted titmice in response to predator treatments and control in the first 5min following presen tation. All pairwise comparisons were significant (LSD, p<0.05) ex cept between control and owl (p=0.868), control and snake (p=0.166), and snake and owl (p=0.061). Error bars: +/1 SE.

PAGE 41

41 Table 2-1. Mean number of each note type gi ven by individual Tufted titmice in response to predator and control presentations in the first 5min following presentation. Note type Predator N Mean SD SE Chick Control 152.937.531.94 Snake 382.538.492.19 Owl 241.604.871.26 Hawk 211.503.781.01 Cat 70.470.990.26 D Control 70.732.150.56 Snake 211.404.171.08 Owl 342.273.410.88 Hawk 216.2910.062.69 Cat 1419.4021.445.54 Seet Control 432.875.781.49 Snake 90.601.840.48 Owl 271.803.430.88 Hawk 433.797.602.03 Cat 543.606.441.66 Song Control 40.271.030.27 Snake 30.200.780.20 Owl 412.738.492.19 Hawk 362.579.062.42 Cat 503.3312.913.33 Chip Control 94260.8052.4313.54 Snake 113870.4080.4420.77 Owl 71345.6045.7111.80 Hawk 80951.3658.3215.59 Cat 44825.6735.729.22

PAGE 42

42 CHAPTER 3 INTERSPECIFIC RISK-BASED CALL SYSTEM OF TUFTED TITMICE ( BAEOLOPHUS BICOLOR ) IN RESPONSE TO PREDATORS Many animals give alarm vocalizations in re sponse to predators but little work has focused on characterizing responses to such si gnals by sympatric heterospecifics. Tufted titmice ( Baeolophus bicolor ) and other members of family Paridae respond to predators that do not pose an immediat e risk (e.g., perched predator s) with complex mobbing calls that have been described as chick-a -dee calls. Black-capped chickadee ( Poecile atricapilla ) mobbing calls vary with th e degree of risk represente d by different species of perched predators, and conspecifics, isolated from predator stimuli, will respond to these calls with risk-appropriate behaviors. M obbing calls of the Tufted titmouse attract many bird species and generate vi gorous interspecific mobbing floc ks that harass and scold predators. I wanted to determine if titmice al so vary their mobbing calls according to the degree of risk that predators pose and if ot her species respond to mobbing and other antipredator calls of Tufted titmice with risk-appropriate behaviors. I presented captive flocks of titmice with live highand low-risk predators and controls under semi-natural conditions to acquire vocal recordings. I then played these recordi ngs and recordings of titmouse seet alarm calls (given when the bird is startled or in a state of fear) to captive pairs of Carolina chickadees ( Poecile carolinensis ) without a predator stimulus. To the high-risk predator presentations titmice approached the pred ator more closely and gave significantly more mobbing cal ls with different note composition and shorter note intervals and note bandwidths compared to vocalizations given to low-risk predator presentations and controls. In response to the playback of titmouse mobbing vocalizations to the high-risk predator, chickadees approached the speakers more closely, gave significantly more mobbing calls with different note composition and longer note

PAGE 43

43 duration than calls given to the low-risk pred ator and control vocalizations of titmice and they froze and became silent in response to titmouse seet calls. Thus, titmice did vary their mobbing calls according to the degree of risk they experience, and chickadees responded appropriately to the various titmouse mobbing calls and alarm calls. Possession of an interspecific risk-based cal l system provides one explanation for the socially dominant role that parid species pl ay in interspecific a ssociations (e.g., winter foraging flocks and predator-mobbing aggregations) involving multiple bird species. Interspecific risk-based calls like those characterized here may underlie ecological facilitation in vertebrate communities more generally Introduction and Background Anti-Predator Vocal Signaling Many vertebrate species respond to predator encounters by giving anti-predator vocal signals. Passerine birds typically have two main types of anti-p redator vocalizations: alarm calls and mobbing calls. Typical passerine alarm calls are difficult to locate and are usually given when the birds are in a state of fear, such as when a predator poses an immediate threat of attack (Marler 1957; Ficken and Witk in 1977; Morton 1977; Apel 1985). Responses to alarm calls usually involve either the cessati on of movement or abrupt flight to cover by the caller and other birds nearby (Marler 1955, 1957; Gompertz 1961 ; Ficken and Witkin 1977; Latimer 1977; Ficken 1989; Evans et al. 1993). In contrast, mo bbing vocalizations are easily localizable and are usually given to perched predators posing litt le immediate risk, and th ese calls attract other species that often harass the predator, sometime s encouraging it to leave the area (Klump and Shalter 1984; Ficken and Popp 1996; Naguib et al. 1999; Baker and Becker 2002). The production and description of antipredator vocalizations has comm only been presented as part of broader analyses of species modes of communi cation and overall vocal complexity. Specific

PAGE 44

44 focus on the production and context of individual species anti-predat or calls has shown that such calls can communicate information about predator s, including the level or type of risk (Struhsaker 1967; Seyfarth et al. 1980b; Maced onia 1990; Dasilva et al. 1994; Blumstein and Armitage 1997a; Greene and Meagher 1998; Zuber bhler 2001). However, understanding when and how conspecifics and heterosp ecifics receive and respond to ri sks conveyed in anti-predator signals is just now coming to light (Naguib et al. 1999; Baker and Becker 2002; Templeton et al. 2005). Predator-Specific and Risk-Based Call Systems Different species vocalizations in response to predators vary in complexity. Some vocalizations contain detailed information about th e type of predator (pre dator-specific calls), while some contain information about the immediacy of threat that the caller faces (risk-based calls; Macedonia and Evans 1993; Greene and Meagher 1998). These two calling systems differ with respect to their production specificity. In production, both pr edator-specific and risk-based calls are referred to as situationa lly specific because the call stru cture in some way varies with distinct situations. If the vocal response unique ly (or categorically) covaries with the stimulus type, as in predator-specific calls where a differe nt type of call is associated with different predator species or classes, then there is a hi gh degree of production spec ificity (Blumstein and Armitage 1997a). On the other hand, if the vocal response varies continuously, i.e., the same call type is produced with graded frequency or intens ity according to the degree of risk, then there is not a high degree of production spec ificity. However, such a system is still considered to be situationally specific because highe r and lower risk situations can be distinguished (Blumstein and Armitage 1997a,b; Blumstein 1999a). Risk-bas ed (also called urgency-based) call systems have been found in several taxa including ground squirrels, marmots, scrubwrens, babblers and

PAGE 45

45 chickadees (Robinson 1980, 1981; Sherman 1985; Fi cken 1989; Blumstein 1995a,b; Blumstein and Arnold 1995; Naguib et al. 1999; Baker a nd Becker 2002; Leavesley and Magrath 2005). Perception specificity refers to the nature of the signal-receivers r eaction to the immediacy of threat conveyed in anti-predator vocalizat ions. If a vocal signal produced under different situations elicits contextually ap propriate responses from conspecifi cs and/or heterospecifics, in the absence of other cues, the vocal signals are said to generate perception specificity (Evans et al. 1993; Macedonia and Evans 1993; Blumstein 1999a). Intraspecific perception specificity is common in vertebrates but work ad dressing interspecific perception sp ecificity of signals is rare relative to the number of systems with sympatric heterospecifics that associate with one another (Fichtel and Kappeler 2002). Species in the family Paridae are known to produce risk-based calls. Ficken (1989) demonstrated that Mexican chickadees ( Poecile gambeli ) vary their alarm or high zee calls according to the degree of risk and Baker a nd Becker (2002) showed that Black-capped chickadees ( Poecile atricapilla ) vary their mobbing calls accordi ng to the immediacy of threat. Neither study conducted playbacks to address intra or interspecific communication to see if others were able to recognize th e variation in the calls and res pond appropriately. This is an important distinction because birds may produce situ ationally specific calls, but unless others are able to recognize them and unders tand their meanings, successful co mmunication does not occur. Recently, Templeton et al. (2005) clearly de monstrated intraspecific communication among Black-capped chickadees. They found that the ch ickadees gave risk-based calls to different predators and that these calls el icited appropriate re sponses from conspecifics. In the present study, I tested for the presence of interspecifi c communication of preda tion risk between two species of parids that co-occur in the southeastern United States.

PAGE 46

46 Study System The Role of Tufted Titmice in MixedSpecies Foraging and Mobbing Flocks Tufted titmice ( Baeolophus bicolor ) and (potentially) Ca rolina chickadees ( Poecile carolinensis ) play the role of the nuclear, or focal species, around which mixed-species foraging flocks form during the winter months. In these flocks, chickadees are socially subordinate to titmice (Waite and Grubb 1988) and are frequently found outside of foraging flocks (Contreras unpubl. data; Farl ey et al. in review). Other species (more than 12-15) that associate regularly with titmouse fl ocks play the role of satel lite species (Gaddis 1983; Grubb and Pravosudov 1994; Greenberg 2000). Regular satellite species in North-central Florida include a diverse set of species : Black-and-white warblers ( Mniotilta varia ), Downy woodpeckers ( Picoides pubescens ), Ruby-crowned kinglets ( Regulus calendula ), Blue-headed and White-eyed vireos ( Vireo solitarius V. griseus ) and Blue-gray gnatcatchers ( Polioptila caerulea ; Farley et al. in review). Contreras (unpubl. data) has s hown that heterospecific flock members follow titmice around the flock territory, providing direct evidence that titmice are likely to play an active leadership role as the do minant nuclear species in this system. Nuclear species of mixed-species bird fl ocks are generally characterized by behavioral traits that lend themselves to interspecific communication; incl uding interspecific dominance, a high level of vigilance, and intraspecific sociality (Munn a nd Terborgh 1979; Hutto 1994). While not often tested effectively, nuclear species in bird floc ks are thought to facilita te flock formation and sometimes food-finding, initiate and guide floc k movements, and reduce predation risk for satellite species (Mnkknen et al. 1996; Dol by and Grubb 2000). In addition, parid nuclear species act as sentinels by givi ng anti-predator vocali zations in response to predators, thereby alerting other flock members to dang er (Gaddis 1983; Dolby and Grubb 1998).

PAGE 47

47 Tufted titmice serve as nuclear species in m obbing aggregations of birds as well as in foraging flocks. Mobbing aggregations, where on e or more bird species gather around and harass a predator, are relatively common in North-cen tral Florida. These aggregations, that can include more than 20 or 30 species of forest birds (Sieving et al. 2004) are formed when a predator is spotted that is not an immediate mortality threat (e .g., a perched predator). Titmice are the most vigilant and aggressive species in mobbing aggregations and these aggregations appear to form around them (Greenberg 2000; S. A. Hetrick, pers. obs.) In North-central Florida, up to half of the fore st bird community responds to tit mouse mobbing calls (more than to other common local species mobbing or alar m calls in this system) by approaching the sounds and engaging in mobbing behavi or (Sieving et al. 2004). Thus, it is likely that a complex interspecific communication system exists between the Tufted titmouse and sympatric heterospecifics involving both th e production of situa tionally specific anti -predator calls and contextually appropriate responses to thes e calls by others (Morse 1973; Sullivan 1984; Zimmerman and Curio 1988). Anti-Predator Calls of the Tufted Titmouse The Paridae have two main anti-predator vocaliz ations in their repert oire: the seet alarm call (Fig. 3-1a; also known as th e high zee, high see, aerial pr edator call) and the mobbing or scold call (known vari ously as, churring, seejert, chick-adee; Smith 1972; Ficken and Witkin 1977; Gaddis 1979). Mobbing calls are variants of the chick-a-dee call, which is a complex call composed of combinations of introductory chick notes and subsequent D notes (dee notes, churr notes), with the number and presence of each note type being variable (Fig. 3-1b, c; Latimer 1977; Hailman 1989). The chick-a-dee cal l complex (or portions of it) is produced in many non-predator situations in addition to be ing the dominant mobbing vocalization (Latimer 1977; Hailman 1989; Grubb and Pravosudov 1994). Which call is given depends on the

PAGE 48

48 situation; for example, an alarm call (seet) will typi cally be given to a raptor in flight that poses an immediate threat of attac k, while mobbing calls are typical ly given to perched raptors representing much less risk of a ttack (Ficken 1989, S. A. Hetrick, pers. obs.). Mobbing calls are easily localizable and are given when other birds are attracted to harass the perched predator, whereas alarm calls are difficult to locate and re sult in the cessation of all vocalizations and movement (freezing) by the calle r and nearby birds or in rapi d escape to cover (Ficken 1989; Baker and Becker 2002; Howell 2006). While it seems clear that interspecific risk-based signaling is likely to be quite common in vertebrate communities, based on the diverse studies showing perception specificity among heterospecific receivers of anti-predator calls (Nuechterlei n 1981; Sullivan 1984; Seyfarth and Cheney 1990; Hurd 1996; Shriner 1998; Windfelde r 2001), the degree to which most of these signaling systems are risk-based an d/or predator-specific is unknown. Research Design I conducted two experiments to address whet her Tufted titmice possess an interspecific risk-based call system in response to predator s. Experiment 1 involved presenting highand low-risk predators and controls to titmouse flocks to address the situational specificity of the titmouses vocal responses. Experiment 2 was a playback study that addressed the perception specificity of Carolina chickadee pairs to titmouse anti -predator vocalizations in response to the highand low-risk predators (obtained in Exp. 1) and titmouse seet vocalizations. Hypotheses I propose that Tufted titmice have an interspe cific risk-based call system in response to predators. This type of call system is composed of 2 parts, which represent 2 distinct hypotheses and I conducted a separate experiment for each hypothesis. The first experiment tested the situational specificity hypothesis that titmice will vary the st ructure of their mobbing calls

PAGE 49

49 according to the situation (Blumstein 1999a; Blum stein and Armitage 1997a,b). One way to do this is to vary the structure of the calls according to the degree of risk that they are presented with. Titmice can vary their mobbing calls in many ways. They can vary the number of mobbing calls given, and within a mobbing call they can vary the type of notes gi ven (introductory chick notes or D notes), how many times a note is given, the temporal parameters of the notes and calls (e.g., note duration, interval between notes and cal ls) and the acoustic structure of notes (e.g., bandwidth, entropy). Titmice may use all or some of these ways to vary their mobbing calls with respect to threat level. Several researchers have found that pa rids respond to hi gher threat levels by increasing their call rate and increasing the number of D notes with in the calls (Apel 1985; Baker and Becker 2002; Templeton et al. 2005). Some parids decr ease or drop the introductory notes of the mobbing call while extending the D note section as the level of risk increases (Gompertz 1961; Latimer 1977). In addition, Temp leton et al. (2005) found that Black-capped chickadees ( Poecile atricapilla ) vary certain temporal measures and the acoustic structure of certain notes when presented with predators of di fferent risk levels. I tested for all of these possible variations across di fferent risk situations. I also conducted an analysis of the behavioral responses of th e titmice to the predators and controls as a standard for determining whether the chickadee responses in the next experiment were appropriate. Parids are known to m ob perched predators, which do not pose an immediate predation risk (Langham et al. in press, S. A. Hetric k, pers. obs.). To characterize a mobbing response in Black-capped ch ickadees, Templeton et al. ( 2005) observed the number of chickadees that came within certain distances of the stimulus and the closest distance that any bird approached the stimulus. Since one of th e accepted hypotheses for mobbing is to drive the predator out of the territory (Shedd 1982; Kl ump and Shalter 1984), it would make sense that

PAGE 50

50 birds would have a greater motivation for mobbi ng higher-risk predators more intensely. One way to mob more intensely would be for more bi rds to approach the pred ator and to approach closer. In the second experiment, I tested the interspecific perception specificity hypothesis that titmouse anti-predator calls should elicit appropriate responses from heterospecific flock members who hear the calls in the absence of the original stimulus (M acedonia and Evans 1993). For chickadees to give appropria te responses, I would expect them to respond to the playback of titmouse vocalizations to predator stimuli in th e same way they would if the stimuli were present in this case, indicated by the responses of titmice to the actual predator situations in Exp. 1. Predictions Experiment 1: Titmice produce risk-based mo bbing calls that are situationally specific. I predict that Tufted titmice will produce si tuationally specific mobbing calls that vary according to the degree of risk to which they are exposed. Specifi cally, I predict th at titmice will vary their mobbing calls according to one or more of the parameters listed in Table 1. To the Eastern screech-owl presentation, which represents a high-risk predator situation, I predict that the titmice will re spond with greater mobbing intensity than to the Great horned owl presentation, which represents a low-risk pred ator situation. To re spond with greater mobbing intensity to the Eastern screech-owl, I predict that the titmice will in crease their mobbing call rate, approach the owl closer, and a greater propor tion of titmice will come within 1m and 3m of the owl. I also predict that titmice will chan ge their note composition by decreasing the number of chick notes and increasing the number of D no tes as the risk level increases (Latimer 1977). In addition to call rate and note composition, vari ation in mobbing calls according to degree of

PAGE 51

51 risk will be identified using the parameters listed in Table 1 (Apel 1985; Baker and Becker 2002). Experiment 2: Chickadees exhibit interspeci fic perception specificity to titmouse antipredator calls. In addition, I predict that Ca rolina chickadees will respond to the titmouse anti-predator calls with some degree of per ception specificity. I predict th at the behavioral and vocal responses of chickadees will vary in response to playbacks of titmouse seet calls and calls for highand low-risk predator situations. More sp ecifically, in response to the playback of titmouse vocalizations given to a high-risk predator situation, I predict that the ch ickadees will respond with more intense mobbing by approaching the speaker more closely and giving a relatively larger number of chick-a-dee cal ls (higher call rate) compared to the response to the playback of titmouse vocalizations elicited from a low-risk predator situation. A similar response was elicited from Black-capped chickadees when th ey were played conspecific mobbing calls in response to different predators (T empleton et al. 2005). Mobbing cal ls, in general, attract birds to the area of the caller to participate in mobbing (Hurd 1996; Baker and Becker 2002; S. A. Hetrick, pers. obs.); therefore, playbacks of titmouse mobbing calls would likely attract chickadees to the area of the speaker. I also predict that chickadees will alter their note composition as described for the titmice in Exp.1, as well as possibly varying other parameters listed in Table 1 in response to the different pl ayback treatments. In response to titmouse seet calls, I predict that the chickadees will not ge nerate mobbing, but will dive to cover and freeze while remaining silent without approaching th e area of the speaker (Gaddis 1980; Ficken 1989). Methods I: Situational Specificity Hypothesis Five flocks consisting of 3 Tufted titmice were captured in Gainesvill e, Florida between 13 October 2005 and 5 January 2006. All 3 birds in each flock were captured at the same time from

PAGE 52

52 the same location to ensure that the three bird s knew each other. The birds were captured around suburban seed feeders using mist-nets and/or bait ed walk-in potter traps a nd then all birds were banded with uniquely colored leg bands. Immediately following capture the birds were transferred to a 12 x 8 x 4m outdoor aviary cont aining numerous live trees and snags, providing a semi-natural habitat at the USDA/APHIS/WS/N WRC Florida Field Station (United States Department of Agriculture, Animal & Plant Health Inspection Service, Wildlife Services, National Wildlife Research Center; USDA lab) in Gainesville, Florida (Fig. 3-2). The aviary was constructed of inch plastic mesh attached to 4 x 4in. post s. After the 24 hr habituation period, during which they were monitored for norma l feeding activity and ge neral health, flocks were tested for each of 4 morn ings in a row while being fed ad-libitum from a feeder in the aviary. Birds were held for up to 7 days before be ing released back at thei r original capture site. Predator Presentations Each titmouse flock was presented with 4 trea tmentsa live Eastern screech-owl (high-risk predator, Megascops asio ), a live Great horned owl (low-risk predator, Bubo virginianus ) and 2 controlsa procedural control wi th a live Northern bobwhite quail ( Colinus virginianus ) and an experimental control with no stimulus (an em pty perch). The owls were non-releasable, rehabilitated owls that were borrowed from Flor ida Wildlife Care, Inc. The 4 presentations were made in randomized order for each flock and spaced approximately 24 hours apart. Most hunting by both owls is nocturnal, but both ow ls occasionally hunt during the day (Packard 1954; Spendlow 1979; Gehlbach 1994). Diet studies have shown that both the Eastern screech-owl (screech owl) and the Great horned owl (great horned) prey on birds, but small songbirds comprise a much greater proportion of the diet of the small, maneuverable screech owl than the larger, less maneuverable great horned (reviewed in Gehlbach 1995 and Houston et al. 1998;

PAGE 53

53 Gehlbach 1994; Turner and Dimmick 1981), thus making the screech owl a higher risk predator to the titmice (reviewed in Templeton et al. 2005). I placed one of the stimuli in the aviary on a 1.2 m perch or pl atform (for the quail) under a removable cover and for the experimental contro l, I placed an empty perch under the cover, approximately 10min before the trial began a nd then retreated to a camouflaged blind just outside the aviary. Care was taken to ensure th at the titmice did not vi ew the stimulus until it was uncovered. The flock was given 5min or long er after the observer ex ited the aviary to resume normal behavior. Audio and video record ings and behavioral observations were then made for 5 minutes pre-stimulus and 7 minutes post-stimulus; however, only the first 2 minutes post-stimulus were included in the analysis. A Sennheiser omni-directional microphone (ME 62) was used to record vocal responses of the titmi ce directly onto a laptop computer using Raven Interactive Sound Analysis Software Versi on 1.1 with a sampling ra te of 44100 at 16-bit resolution. Behavioral responses within the first 2min post-stimulus were characterized using the following behavioral variables adapted from Templeton et al. (2005): a) the closest distance any bird approached the stimulus (in m); b) th e proportion of birds that came within 3m of the stimulus; c) the proportion of bird s that came within 1m of the stimulus; and d) whether the birds were frozen in place during the en tire treatment (Table 1). For the behavioral variables (a-c), I used Kruskal-Wallis non-parametric test with one -tailed Mann-Whitney U (MWU) post-hoc tests to conduct pairwise compar isons among the treatments. Spectrographic Analyses I analyzed vocal responses in 13 of the 20 pr esentations (n=5 for the screech, n=4 for the great horned, n=4 for the controls) because t itmice did not vocally respond in 7 of the presentations. Due to the low instance of vocal response to th e procedural control (n=2) and experimental control (n=2), the controls were lu mped together in the analyses, resulting in n=4

PAGE 54

54 control samples. Spectrographic analyses were performed on the vocal recordings using Avisoft SASLabPro 4.39. To edit out noise, each sound f ile was FIR low-pass filtered at 12kHz and high-pass filtered at 1.8kHz. The spectrogr am parameters used were FFT=512, Frame Size=75%, Window=Hammi ng, and Overlap=87.5%. The notes in each call, or chick-a-dee call complex, were visually classified as introductory chick notes or subsequent D not es. The various introductory notes grade into each other and are not reliably di stinguished into natural categories; therefore, they were classified together as chick notes (Fig. 33). On the other hand, D notes can be reliably classified (Bloomfield et al. 2005) due to their harmonic-like structure and little frequency modulation. D notes also have a higher entropy and lower frequenc y than the introductory chick notes, and D notes always occur at the end of th e call, or are the only notes comprising a call, making them easily distinguished from introducto ry notes (Fig. 3-3). Seet calls given by responding titmice were omitted from analysis b ecause the recording equipment could not pick them up due to their extremely low amplitude. Single A or chip notes were omitted in the analysis due to their preval ence and predominantly low amplitude and peter songs were omitted in order to include only vocalizations us ed in the anti-predator context (Latimer 1977; Gaddis 1979). For the predator treatments, there was only one titmouse contributing to the majority of the vocal mobbing in 4 out of the 5 flocks. This may be explained by the existence of social dominance hierarchies within the flock, with the alpha male possibl y contributing th e majority of the anti-predator vocalizations (Brawn and Samson 1983), as o ccurs in the closely related Willow tit ( Parus montanus ; Hogstad 1993) and in Pa le-winged trumpeters ( Psophia leucoptera ; Seddon et al. 2002). In 3 of the 5 flocks, it was a known adult male that responded; in 1 flock, it

PAGE 55

55 was a hatch year male; and in 1 of the flocks it was an adult of unknown sex (likely a male). Age and sex were determined according to the molt li mit and wing chord criteria established by Pyle (1997). If more than one titmouse responded vocally to the treatment, the calls of the dominant titmouse responding were isolated and measured. I measured several aspects of the notes and calls in the 2min post-stimulus recordings, including some that were based on measures used in previous studies of parids (Baker and Becker 2002; Freeberg et al. 2003; Templeton et al. 2005). For each treatment, I averaged the number of calls, the number of chick and D not es overall, the number of notes per call, the number and proportion of chick a nd D notes per call, the duration of each chick and D note, the call duration, the duration of the 1st D note per call, the interval between notes, the interval between the chick and D section, and th e interval between calls (Table 1). Spectrum-Based Measures I also measured several fine-scaled acoustic pa rameters on a sub-sample of 10 D notes that were randomly chosen from each 2min post-stimulus recording. If there were fewer than 10 D notes in the 2min post-stimulus recording, as was frequently the case for the controls, I chose as many D notes as possible from the recording. In all cases, the D notes were chosen from highquality recordings. I measured the spectrum-base d parameters of the D notes in the sub-sample using a power spectrum with FFT length=512. Th e parameters were computed at the maximum spectrum of the entire D note (maxpeakhold) a nd are similar to those used by Nowicki (1989) and Templeton et al. (2005). The parameters were minimum and maximum frequency where the amplitude goes last below -30dB and where the amplitude goes last below -10dB (min. and max. frequency with the total option activated in Avisoft SASLabPro 4.39), bandwidth at -30dB and -10dB (calculated with min and max frequency described previously), entropy, and the number of peaks above -10dB.

PAGE 56

56 For each acoustic factor that was measured, I used univariate analysis of variance (ANOVA) with the least signi ficant difference (LSD) post-hoc test to conduct pairwise comparisons among the treatments. I transformed the data when appropriate to meet the assumptions of the analysis using sqrt(n), arcs in(sqrt(n)), and log(n+1) transformations (Sokal and Rohlf 1995). All statistic al analyses were conducte d using SPSS 11.5 for windows. Because many of the variables were correlated with each other, I performed a Principal Components Analysis (PCA) on all 14 general spectrographic measures and 3 behavioral measures (excluding the measure of whether the birds froze in pl ace). I used the uncorrelated composite variables generated from the PCA in a Discriminant Function Analysis (DFA) to determine if the discriminant functions could corr ectly classify the flock responses to one of the three treatments. Significan ce in all statistical tests wa s set at the 0.05 alpha-level. Results I The results of the univariate ANOVAs showed th at all of the behavi oral variables and 13 out of 14 of the general spectro graphic variables that were m easured varied with treatment (p<0.05). I determined which treatments diffe red from the others with the LSD pairwise comparisons test. In some cases, the Eastern sc reech-owl treatment was significantly different from the Great horned owl, which were both signif icantly different from the lumped (procedural and experimental) controls (hereaf ter referred to as control). A ll transformations and the results for all pairwise comparisons are shown in Appendi x A (Table A-1, A-2). Means, SD and SE for all measures taken are presented in Appendix B (Table B-1). Titmice exhibited more intense mobbing behavi or when presented with the screech owl than with the great horned or c ontrols. Titmice approached the screech owl more closely and more of the titmice came within 1m and 3m of the screech owl than of the great horned or control (Kruskal-Wallis 2 2=7.057, p=0.029; 2 2=8.670, p=0.013; 2 2=7.355, p=0.025,

PAGE 57

57 respectively, Fig. 3-4, 5), alt hough the pairwise comparisons be tween the great horned and control were not significant. None of the titmouse flocks froze in place during any of the treatments. The titmice gave different numbers of mobbi ng calls to the different treatments (ANOVA F2, 20=12.2, p=0.001; Fig. 3-6), with the high-risk screec h owl eliciting a highe r call rate than the low-risk great horned or control treatments (LSD p=0.006, p<0.001, respectively) and the great horned eliciting a higher call rate than the cont rol, although the differen ce was not significant (LSD p=0.200). In particular, the high-risk scre ech owl elicited fewer chick notes per call (F2, 266=378.4, p<0.001; Fig. 3-7), mo re D notes per call (F2, 266=837.8, p<0.001; Fig. 3-7), and a greater number of D notes overall (F2, 20=16.2, p<0.001; Fig. 3-8) th an the great horned or control treatments. The overa ll number of notes per call (F2, 266=359.9, p<0.001) and the duration of the entire call (F2, 266=280.0, p<0.001) were greater and the interval between notes (F2, 1313=7.7, p<0.001) was shorter for the screech owl tr eatment than the grea t horned or control treatment. Of the fine-scaled acoustic parameters that were measured, bandwidth at -30dB was different between the D notes in each treatment, with the screech owl treatment eliciting titmice to give D notes with a lo wer bandwidth than the gr eat horned and control (F2, 105=10.7, p<0.001; all pairwise comparisons were significant). The screech owl treatment also elicited D notes with a higher minimum frequency (where amplitude goe s last below -30dB) than the great horned or control (F2, 105=13.1, p<0.001; all pairwise comparisons were significant except between the screech and great horned treatment, p=0.065). The results of the PCA on the 17 measures are summarized in Table 2. Four components with eigenvalues >1 were extracted from the data set. The first principal component (PC1) was

PAGE 58

58 determined mostly by behavioral variables, call rate, and note compos ition variables. The second, third, and fourth principal components (P C2-4) were determined mostly by temporal features of the notes and calls. After varimax rotations, the first four principal components explained 85.3% of the variance, with PC1 accounting for 48.6% and PC2, 3 and 4 accounting for 15.7%, 10.8%, and 10.2% of the variance, respec tively. The 17 variables are depicted in a bivariate plot that shows their respective values for PC1 and PC2 (Fig. 3-9). The DFA using the PCA factor scores led to 92.3% corr ect classification of the variable s with the eliciting stimuli. The first of the two discriminant functions (D F1) accounted for 91.9% of the variation, and the second (DF2) accounted for 8.1 % of the variati on. The resulting graph (Fig. 3-10) shows the separation of flock responses into three basic groups with clearly se parated group centroids between the three treatments. The only misclass ification was one of the flocks response to the great horned treatment classified as a response to the sc reech owl treatment. Methods II: Interspecific Perception Specificity Hypothesis Ten pairs of Carolina chickadees were capture d in Gainesville, Flor ida between 10 January 2006 and 5 March 2006. The birds were capture d, banded, and housed in the same manner as the titmice in Exp. 1. In the study region, observa tions of normal group sizes at feeders indicated that while titmice normally travel in groups of 3 or more, chickadees are nearly always in pairs. Thus the number of birds per flock was different for the two species but was determined to be the most natural combinations of individuals likely to be related to, or at least familiar with, each other (flock members were always captured within a few minute s of each other at the same feeder in the same mist net or in adjacent or same potter traps). Playback Presentations Playback recordings (2min duration) were constructed from recordings of titmouse mobbing calls, alarm (seet) calls a nd control vocalizations to make 5 independent replicates of

PAGE 59

59 each of 4 playback treatment types. Two of the playback treatment types were of titmouse vocalizations acquired in Exp. 1 in response to th e screech owl and the great horned. The control playback was constructed from pre-stimulus r ecordings of titmouse vo calizations (containing mostly contact vocalizations) acquired in Exp. 1. For each playback treatment type, there were 5 unique exemplars (5 different titmouse flock responses recorded in Exp. 1). The fourth playback treatment type was of titmouse seet calls acquire d from other sources. Five unique seet call recordings (2min each, 123 to 217 seet calls per min) were made us ing seet calls recorded from one flock of free-living Tufted titmice responding to unknown stimuli (by Lang Elliot; http://www.naturesound.com/) and from a previous experiment with captive titmice responding to presentations of predators in close proximity (hawk, owl or cat; n=4 different birds, by S. A. Hetrick; see Chapter 2). Each of the 20 playback recordings (5 variants of each of 4 treatment types) was used twice for a total of 40 playback s (n=40). Each of ten chickadee pairs received 4 of the playbacks (one each of the 4 treatments). The 4 treatm ents were presented in random order for each pair and spaced approximately 24 hours apart. At the beginning of each playback treat ment, SAH placed a pair of camouflaged RadioShack speakers (Model 40-143 1) in the aviary on a 1.2m platform approximately 10min before the trial began and then retreated to a cam ouflaged blind just outsid e the aviary. For each treatment, the speakers were randomly placed in one of three locations in the aviary in order to reduce habituation to a particular direction. The pair was given 5min or longer after the observer entered the aviary to resume normal behavior. Recordings and behavioral observations were then made for 5min pre-playback, during the 2min playback of the titmouse vocalizations and for 5 additional min post-playback. The methodol ogy from Exp. 1 was followed to record and

PAGE 60

60 video-tape the chickadees behavioral and vocal responses. This methodology was also used to analyze the behavioral va riables associated with the chickadees response. Spectrographic Analyses For each treatment, I analyzed 3min of beha vioral responses and conducted spectrographic analyses of 3min of recordings. Some chicka dee pairs gave strong re sponses during the 2min playback while some waited until th e playback had finished to be gin responding, so analysis of 3min total (2min during the playback and 1min pos t-playback) was deemed the most appropriate. Chickadee behavioral responses a nd acoustic parameters were char acterized as in Exp.1. I did not analyze fine-scale acoustic measures on the chickadee vocalizations because most of the chickadee vocalizations overlapped with the titmouse vocalizations in recordings of the playback trial. The acoustic data were analyzed usi ng univariate ANOVA with the LSD post-hoc tests to conduct pairwise comparisons among the treatments. The seet playback was only considered in the analysis of the behavior al responses and excluded from the spectrographic analyses of vocalizations due to the low instance of calling du ring the playback treatment. Significance in all statistical tests was se t at the 0.05 alpha level. Results II The results of the univariate ANOVAs showed th at all of the behavi oral variables and 9 out of 10 of the general spectro graphic variables that were m easured varied with playback treatment (p<0.05). The LSD pairwise comparisons showed that in some cases, the chickadees response to the screech mobbing playback (playback of the titmouse vocalizations elicited from a screech owl) differed from the response to the great horned mobbing pl ayback (playback of titmouse vocalizations elicited from a great horned), which both differed from the response to the control playback (playback of pre-stimulus titmouse vocalizations). In other cases, one or more of the playback treatments did not differ from the other playback treatments. The details of the

PAGE 61

61 transformations and the results for all pairwise comparisons are shown in Appendix A (Table A3, A-4). Means, SD and SE of all measures ta ken are presented in Appendix B (Table B-2). Chickadees exhibited more intense mobbing behavior when they heard the screech mobbing playback than when they heard the great horned mobbing and control playbacks. Chickadees approached the speakers that broa dcast the vocalizations more closely and both members of the pair were more likely to come w ithin 1m and 3m of the camouflaged speaker in response to the screech mobbing playback than to the great horned mobbing, control, or seet playbacks (Kruskal-Wallis 2 3=23.666, p<0.001; 2 3=24.118, p<0.001; 2 3=16.687, p=0.001, respectively; Fig. 3-11, 3-12), although some pair wise comparisons were not significant (see Appendix A, Table A-3). The chickadee pair s did not freeze in place during any of the playbacks except the seet playback in which they each froze for 100% of the 3min experimental periods (and for long periods afterwards). The screech mobbing playback elicited an overall greater numbe r of notes per call (ANOVA F2, 338=9.1, p<0.001) with a fewer numbe r of chick notes per call (F2, 338=32.3, p<0.001; Fig. 3-13) and a greater number of D notes per call (F2, 338=27.6, p<0.001; Fig. 3-13) than the great horned mobbing or control playback s. All of the pairwise comparisons were significant for these three variables except for th e number of chick notes per call for the great horned mobbing playback and control playba ck (LSD p=0.108). The call duration and the average duration of each D note were greater fo r the screech mobbing playback than the great horned mobbing or control playbacks (F2,335=27.8, p<0.001; F2, 775=119.9, p<0.001, respectively).

PAGE 62

62 Discussion Situational Specificity of Titmouse Mobbing Calls with respect to Risk The results of the predator presentation experiment confirmed that Tufted titmice have situationally specific mobbing calls in that their calls varied acco rding to the situation. These calls are risk-based because the calls varied according to risk. More specifically, the results confirmed my prediction that titmice would e xhibit a stronger mobbing response to the higher risk predator. Behavioral responses of titmice to the predator presentations showed that titmice clearly distinguished high and low degrees of risk by exhibiting a more intense mobbing response to the high-risk Eastern screech-owl. They approached it more closely (within 1m) and gave a greater number of chick-a-dee mobbing calls than to the low-risk Great horned owl (Fig. 3-4, 3-6). These results are cons istent with the results of othe r researchers who found that other species in the family Paridae increase their m obbing call rate as the le vel of risk increases (Latimer 1977; Apel 1985; Baker and Becker 2002; Templeton et al. 2005). Black-capped chickadees ( Poecile atricapilla ) altered their rate of calling in response to a stuffed falcon at near and far distances (Baker and Becker 2002) and in response to different predator species (Templeton et al. 2005), by calling at higher rates in the higher risk situations. Many species of rodent also increase their rate of calling as risk increases (Nikols kii and Pereladova 1994; Blumstein and Armitage 1997a; Randall and Rogovin 2002). Titmice also varied their note composition in response to the different predators by decreasing the amount of introductory chick notes per call and increasing the amount of D notes per call as risk increased (Fig. 3-7). These findings agree with the observations of past authors who noted that as the level of f ear or risk increases, titmice drop the prefix (chick) notes and increase the churr (D) notes (Odum 1942; Gompertz 1961; Latim er 1977). Variation in note composition was previously thought to encode information related to many factors, including

PAGE 63

63 information about predators (Hailman et al. 1985, 1987; Ficken et al. 1994). Black-capped chickadees alter the number of in troductory A and B notes in th eir chick-a-dee mobbing calls in response to predators presente d at different distances (Baker and Becker 2002) and vary the number of D notes in the calls in response to raptors of different sizes, with the smaller, higher risk raptors eliciting the most D notes (Templeton et al. 2005). According to my findings and the findings of the authors listed a bove, it appears that call rate an d note composition, in particular, are important for titmice and chicka dees in communicating about risk. I also found that titmice varied the note in terval and the bandwidth of the D notes according to the level of risk, which the chickadees could be cueing in on to help them interpret the content of the titmouse calls. Other birds and mammals have also been documented to change the structure of their call notes in differe nt predator situations. White-browed scrubwrens ( Sericornis frontalis ) vary the structure of their aerial trill call by increasing the minimum frequency (pitch) of their calls according to th e distance from a suddenly appearing predator (Leavesley and Magrath 2005). Yellow-bellied marmots also vary several frequency characteristics of their calls, in cluding bandwidth, as a function of distance to certain predators (Blumstein and Armitage 1997a). In both of thes e studies, the distance fro m the predator to the subject represents the amount of risk, with the closer distances representing higher risk of being caught or attacked by a predator; thus, we can conclude that both the scrubwren and the marmot vary the structure of their calls according to the degree of risk that they perceive. As Templeton et al. (2005) noted, variation in th e structure of calls in response to different predat ors, reflecting the degree of risk perceived, likely occurs in ma ny species, but few researchers have tested for this.

PAGE 64

64 The results of the DFA indicate that the highand low-risk predator treatments and control treatment can be reliably (in 92.3% of the cases), and potentially uniquel y, distinguished by the combined behavioral and general spectrographic variables that were measured in this experiment (Fig. 3-10). The combination of variables meas ured clearly discriminate between the titmouse responses to the treatments, rais ing the possibility that the pr oduction of anti-p redator signals could be predator-specific. Davis (1991) made a similar argument that yellow-bellied marmots had predator-specific calls based on results from a multivariate DFA. If the titmouses response were predator-specific, this would mean that wh ile individual acoustic and behavioral parameters may vary in a graded fashion across predator situations (e.g., more and fewer D notes), the combination of multiple acoustic characteristics encoded in titmouse anti-predator signals could (together) uniquely identify distinct predator species or classes. It is impo ssible to conclude this, however, without further study of the nature of call production in Tufted titmice. Some variation was evident in the vocal res ponses between the different titmouse flocks, and this could be attributed to the predators mo vement and behavior at the time of presentation. Many authors have observed that the behavior of a predator affects birds reactions to it. Increased call rates and more intense mobbing as a result of predator movement were reported in several species of birds including Carolina wrens ( Thryothorus ludovicianus ; Morton and Shalter 1977), Eurasian blackbirds ( Turdus merula ; Frankenberg 1981), and Pied flycatchers ( Ficedula hypoleuca ; Shalter 1978; reviewed in Apel 1985). It has also been suggested that some passerines can even detect differences in posture and behavior that ar e associated with how hungry a predator is, and mob more frequently if the predator is hungry (Hamerstrom 1957). In the present study, we did not detect that the two owls exhibited differences in posture or that their movement differed at different times, but we did not examine this, since they were tethered

PAGE 65

65 to the perch and were generally calm during expe riments. I did observe that the owls were sometimes looking in the direction of the titmice and sometimes looking in the opposite direction. On one occasion, the great horned ow l jumped down from the perch to the ground and this seemed to evoke a momentar y increase in titmouse call rate. In general, it is well accepted that parids give alarm calls in response to predators that pose an immediate threat, such as an aerial predator in a low attack flight, and that they give mobbing calls to predators that do not pose an immediate threat, such as a perched predator (Gaddis 1980; Ficken 1989). Having different calls for different classes of predators suggests that titmice may be giving predator-specific calls, although this argument is weakened because there are many exceptions as to when titmice give these calls (see below). Here, I show that mobbing calls given by titmice to highand low-risk (but all pe rched) predators and no-r isk controls are also clearly situationally specific. Whether they are categorically different (predator-specific) or graded into one anothe r (risk-based) may depend on how the calls are pe rceived by receivers. For example, if a receiver is listening to only the number of D notes gi ven per minute (Fig. 3-8) then the level of risk being communicated will be graded ac ross the range of numbers of notes given and these can vary among individuals presen ted with the same stimulus. In these cases, the communication would be risk-based and not predator-specific by standard definitions (Macedonia and Evans 1993; Blumstein and Armitage 1997a; Blumstein 1999a). However, if a receiver is basing its assessment of risk on more than one parameter ch aracterizing the titmouse mobbing calls, then 3 distinct (non-overlapping) classes of risk could be represented in the titmouse vocal responses to the three types of stimuli (2 owls vs. control). In either case, titmice are giving vocal signals that dist inguish between highand low-risk predators (both perched and not of immediate threat) and no-ri sk controls (quail or empty perc h). More work is needed to

PAGE 66

66 determine whether single or multiple call charac teristics generate the most appropriate responses in receivers. Some bird species possess predator-induced calls that are functionally referentialthat is the calls are predator-specific w ith respect to predator species or class (e.g., high degree of production specificity) and they pr oduce appropriate responses in the individuals hearing them (e.g., high degree of perception specificity; Klump and Shalte r 1984; Evans et al. 1993; Blumstein 1999a; Seddon et al. 2002). Within this framework, titmouse anti-predator calls appear to be functionally referentia l in that they usually give seet alarm calls to aerial predators and chick-a-dee mobbing calls to perched and te rrestrial predators (Ga ddis 1980; Ficken 1989). In purely experimental situations used here (Chapters 2 and 3), I have found that this is not always the case, as perched predators and terres trial predators presented to titmice in captivity sometimes elicited seet calls. Th is would imply a lower degree of production specificity for the seet call than is generally supposed. Seet call us e may be more related to the level of fear or surprise that an individual experiences (Marle r 1957; Apel 1985) and in captivity, fear levels could be generally higher than in the wild. Chick-a-dee calls al so appear to have a low degree of production specificity because they are given in a variety of non-predator situations as well as being the main mobbing vocalization (S. A. Hetr ick, pers. obs.). Because one of the main requirements for functional reference is high pro duction specificity, there is weak evidence that titmouse anti-predator calls are functionally refere ntial because both calls appear to have low production specificity. Perception Specificity of Chickadees Exposed to Titmouse Anti-Predator Calls The results of the playback experiment conf irmed my prediction that chickadees would respond with greater mobbing intensity to titmou se calls given in response to the high-risk screech owl and with freezing and silence to th e titmouse seet playback. In other words,

PAGE 67

67 chickadees that heard the titmouse calls were ab le to interpret them and respond appropriately. When the chickadees heard the screech mobbing pl ayback, they responded (as the titmice did in response to the screech owl) by a pproaching the stimulus more cl osely (within 1m) than to the great horned mobbing playback (Fig. 3-11). Bl ack-capped chickadees respond in a similar way to conspecific vocalizations elic ited by highand low-risk pred ators by altering their approach distance to the stimulus (Templeton et al. 2005). In the present study, chickadees responded to the screech mobbing playbacks by altering the note composition of their chick-a-dee mobbing calls. They decreased the amount of introductory chick notes and increased the amount of D notes per call (Fig. 3-13). I was unable to examine more subtle variati ons in call and note stru cture of the chickadee calls in this study (see methods) but it is likely that these varied in other ways similar to those of Black-capped chickadees and titmice (e.g., bandwidth, frequenc y; Templeton et al. 2005; Exp. 1). But by documenting that the relative frequency of chick ve rsus D notes per call va ried by treatment, my conclusions about risk-appropria te responses by chickadees ar e supported by the findings of others (Apel 1985; Baker and Becker 2002; Temp leton et al. 2005). And responses of the chickadees clearly paralleled those of the t itmice (Exp.1), indicating th at chickadees responded to the titmouse calls, in the absence of other cu es, in much the same way that the titmice responded to the actual predators. When the chickadees heard playbacks of titmouse seet alarm calls, they froze in place every time and were almost totally silent for 5min or longer. This is similar to the observations of Ficken and Witkin (1977) who found that Black-capped chickadees immediately became motionless or moved to cover if in the ope n and froze upon hearing the alarm (e.g., high zee) calls of another chickadee. Several other aut hors have observed that re cipients stop moving and

PAGE 68

68 become silent upon hearing these calls (Ficke n et al. 1978; Gaddis 1980; Waite and Grubb 1987; Ficken 1989). These responses make sense becaus e seet calls are thought to be given when birds are fearful and/or perceive im minent attack, as with the s udden emergence of any potential predator including (especially) fl ying or low-cruising raptors (La timer 1977; Ficken et al. 1978; Apel 1985; Ficken 1989). Smith (1972) suggested that seet calls probably function to alert recipients to danger. If the message that titm ice are giving with their seet calls is extreme danger, it seems appropriate that the chic kadees would respond by freezing in place and becoming silent. These behaviors would allow an i ndividual to remain inc onspicuous in the face of potential immediate danger, for example, to lessen the risk of being detected by a cruising predator that had not been located by the prey. In sum, responses to the situationally specific anti-predator calls of the titmouse suggest a high degree of perception specificity in chickadees. Titmice Give Interspecific Risk-Based Mo bbing Calls in Response to Predators Although many researchers have demonstrated that anti-predator calls can communicate specific predator information to conspecifics (i ntraspecific communication) few have shown that these calls can also have meaning to heterosp ecifics (interspecific communication). Mixed species groups of lemurs, ( Eulemur fulvus rufus and Propithecus verreauxi verreauxi ) have evolved an interspecific functi onally referential alarm system for diurnal raptors where both species respond to the calls of conspecifics a nd heterospecifics in the group (Fichtel and Kappeler 2002). Western grebes ( Aechmophorus occidentalis ) that nest in association with Forsters terns ( Sterna forsteri ) respond to the alarm calls of th e terns by leaving their nests and swimming to open water (Nuechterlein 1981). Ot her primate, rodent, and avian groups may have interspecific anti-predator calling systems as well (Marler 1957; Francis et al. 1989; Shriner 1998; Windfelder 2001; Langham et al. in press) Compared to th e large proportion of vertebrates that join mixed species groups (r eviewed in Greenberg 2000), and the pervasive

PAGE 69

69 evidence that most animal groups form, at least in part, to gain anti-pre dator advantages (Caro 2005), a lack of interspecific communication re lated to predator avoidance may be more surprising than its presence but more work is needed to survey for the prevalence of these communication systems. Most researchers that have f ound risk-based (or urgency-based) call systems in response to predators have focused on typical alarm calls as opposed to mobbi ng calls (Robinson 1980; Blumstein 1995a; Blumstein and Armitage 1997a; Leavesley and Magrath 2005). This distinction is important because alarm calls are a ssociated with a flight or freeze response while mobbing calls are associated w ith approach and harassment of the predator. Many species mobbing calls, as well as their alarm calls, may al so be able to communicate differences in the risk environment. It is likely that perception specificity of titmouse risk-based mobbing calls occurs among many other species, besides the closely related chic kadees, that associate with titmice in foraging and mobbing flocks and that simply share hab itats with titmice. Downy woodpeckers, for example, respond with risk-averse behaviors wh en they hear titmouse alarm calls (Sullivan 1984) and this makes sense because they spend a great deal of time with titmice in winter foraging flocks (Farley et al. in review). More telling are the results of an experiment by Howell (2006) in which the same playback recordings of titmouse calls that I used in Exp. 2 (for chickadees) were presented to free-living Northern cardinals ( Cardinalis cardinalis ) that were feeding at platform feeders in the open. The cardinals responded with risk-appropriate behaviors (e.g., freezing to seet calls, and diving for cover more often in response to mobbing calls to highrisk predators than to mobbing calls to low-risk predators). Northern car dinals will join mobbing flocks but not titmouse-led foraging flocks, so are not classified as a mixed-flock joiner.

PAGE 70

70 However, they are sympatric with titmice in many habitats throughout their common range and so are exposed to titmouse anti-predator calls very frequently. Most surprising yet, is the evidence found by Schmidt (unpubl. data) that squirrel s feeding in trays in the open also respond appropriately to titmouse alarm (seet) and mobbing cal ls. These results suggest that a wide range of species that are and are not associated wit h, or taxonomically related to, titmice may perceive titmouse anti-predator calls with a high degree of perception specificity. Finally, though I conclude based on findings presented here that both production and perception of titmouse anti-predator calls are ri sk-based (graded, or continuous), the nearcomplete separation in multivariate space achie ved by the DFA between the titmouse vocal and behavioral responses to control, great horned an d screech owl presentations (Fig. 3-10) raises a second possibility. If receiving species are able to discern multiple acoustic characteristics of titmouse anti-predator signals (i.e., hear the calls in a multivariate fashion), they could be discriminating among predator-risk situations acco rding to class, as in a predator-specific (categorical) call system rather than a risk-b ased (continuous) system of perception (Davis 1991). If so, then the anti-predator signaling syst em of titmice may actually be interspecifically functionally referential (Evans et al. 1993; Blumstein 1999a). Howe ver, it is impossible to draw this conclusion without further study of the natu re of call perception in species that exhibit appropriate responses to titmouse anti-predator calls. Potential Functions of Interspecific Risk-Based Mobbing Calls of Titmice My study revealed characteristics of risk-based mobbing calls of titmice that could be used by chickadees and other species to assess situation-specific risks, and clearly shows what riskappropriate behaviors are when mobbing and alarm calls are given. But why should titmice possess a complex communication system with inte rspecific risk-based mobbing calls? Alarm calls (for extreme danger) are produced in many sp ecies that are intrasp ecifically social, and

PAGE 71

71 have been reasonably explained on the basi s of kin selection (Hamilton 1963, 1964; Maynard Smith 1965; Sherman 1977;Woolfendon and Fitzpatr ick 1984). Researchers have formulated many hypotheses as to why birds give mobbing cal ls to perched predators (Curio 1978, 1980; Smith 1991). One hypothesis that seems to make se nse for our particular system is the moveon hypothesis where individuals se ek to drive the predator from the area (Curio 1978). Titmouse flocks hold stable winter territories and therefore would benefit by moving a predator out of their territory to decrea se the risk of future predation (Brawn and Samson 1983). By giving interspecific risk-based m obbing calls, titmice can alert others to the specific situation and induce them to join in the mobbing of the predator. Risk-based calls in response to higher risk predators would likely generate a more intense mobbing response because titmice and the receivers would have a greater motivation for driv ing these predators out of the area. A more intense response would likely resu lt in a greater chance of moving the predator out of the area, which would benefit all flock members. Most of the birds that partic ipate in foraging and mobbing flocks with titmice share the same predators. Shriner (1998) noted that when species have predators in common, they might be able to obtain important information about pr edation risk from the an ti-predator calls of the other species, and so natural selection would drive the evolut ion of interspecific perception specificity. It has been shown that more bird s are attracted to titmouse mobbing calls than other local forest species mobbing calls (Sieving et al. 2004) and my work shows that this could happen, in part, because the titmice are provi ding detailed information about the risk environment in their calls that may help reduce the predation risk for other species that evolve to use that information. The possession of such calls by titmice suggests a mechanism underlying their socially dominant role in mixed-species foraging flocks. Tufted titmice give information-

PAGE 72

72 laden vocalizations according to the risk enviro nment and Carolina chickadees and potentially other flock members are able to interpret and exploit this inform ation, which may be an incentive for them to join these mixed-species flocks (Gaddis 1980; Sullivan 1984; Howell 2006). Other flock members may also benefit from the titmouse high rates of vigilance, which allows them to reduce their own vigilance and put more of their energy into foraging (Cimprich and Grubb 1994). Therefore, there are many reasons to associ ate with titmice, and the specific information about predation risk that titmice provide, in ad dition to their high vigilance and aggressiveness towards predators, may each play a role in decr easing the predation risk for other species that associate, or merely live, with titmice. Summary In summary, this study clearly demonstrates that titmice possess an interspecific risk-based call system with respect to predators. I have described many call characte ristics that could be used to communicate situational specificity a nd I have shown that chickadees respond to titmouse anti-predator calls with a high degree of perception specificity. Other work is showing that the risk-based anti-predator calls of titm ice are perceived by, and generate situationally specific responses in, a wide range of unrelated species. If mo st of the species that share predators with titmice participate in an interspe cific risk-based communication system, then this suggests a mechanism that could underlie intersp ecific facilitation via predation-risk reduction among diverse sympatric vertebrate species. The majo rity of species in the family Paridae, that are distributed throughout the Holarctic, exhibi t highly conserved (similar) mobbing calls and a high proportion of passerine specie s that live with titmice respond to titmouse mobbing calls by exhibiting typical predator-mobbing behaviors (L angham et al. in press). This study (and others) suggests that heterospec ifics respond appropriately to parid anti-preda tor calls with a high degree of perception specificity. Therefore, pari d anti-predator vocalizati ons support a system of

PAGE 73

73 interspecific communication about risk in whic h many species that share their predators may participate. The existence of such a br oad-based communication system suggests that interspecific facilitation within bird communities of the Holarctic (where parids are distributed) may be as common as other, more widely-studie d, ecological interactions in organizing bird communities (e.g., predation and competition).

PAGE 74

74 A B C Figure 3-1. Examples of the major anti-predator vocalizations of the Paridae. A) Seet call of the Tufted titmouse. B) Chick-a-dee m obbing call of the Tufted titmouse with introductory chick notes and subsequent D notes. C) Chick-a-dee mobbing call of the Carolina chickadee w ith introductory chick notes and subsequent D notes.

PAGE 75

75 Figure 3-2. Outdoor aviary at the USDA/APHIS/WS/NWRC Florida Field Station in Gainesville, Florida. Top picture: Platform with removable cover is in the center of the picture with camouflaged blind in bac kground. Bottom picture: Another view of aviary with cover removed, revealin g the Great horned owl on the perch.

PAGE 76

76 Figure 3-3. Examples of the variation in the ch ick notes and the less variable D notes in the chick-a-dee call complex of the Tufted titmouse. The chick notes grade into each other and are not reliabl y distinguished into natural sub-categories.

PAGE 77

77 screech great horned control Treatment Type 4.0 3.0 2.0 1.0 0.0 Closest approach distance (m) Figure 3-4. Closest approach distance of Tufted titmice to the stimuli during the predator and control treatments in the first 2min followi ng presentation. All pairwise comparisons were significant (MWU, p<0.05) except between the great horned and control treatments (p=0.594). Error bars: +/1 SE.

PAGE 78

78 screech great horned control Treatment type 1 0.8 0.6 0.4 0.2 0 Proportion Prop. of titmice within 3m of stimulus Prop. of titmice within 1m of stimulus Figure 3-5. The proportion of Tuft ed titmice that approached with in 1m and 3m of the stimuli during the predator and cont rol treatments in the first 2min following presentation. For the proportion of titmice within 1m a nd 3m of the stimuli, respectively, all pairwise comparisons were significant (M WU, p<0.05) except between great horned and control (p=0.953, p=0.953). Error bars: +/1 SE.

PAGE 79

79 screech great horned control Treatment type 50 40 30 20 10 0 Number of calls Figure 3-6. Number of chick-a-dee complex ca lls given by Tufted titmice in response to the predator and control treatments in the firs t 2min following presentation. All pairwise comparisons were significant (LSD, p<0.05) except between the great horned and control treatment (p=0.200). Error bars: +/1 SE.

PAGE 80

80 screech great horned control Treatment type 7 6 5 4 3 2 1 0 Mean # of notes Mean # of D notes per call Mean # of chick notes per call Figure 3-7. Number of chick and D notes per c hick-a-dee complex call given by Tufted titmice in response to the predator and contro l treatments in the first 2min following presentation. All pairwise comparisons we re significant (LSD, p<0.05). Error bars: +/1 SE.

PAGE 81

81 screech great horned control Treatment type 300 250 200 150 100 50 0 Mean # of D notes Figure 3-8. Mean number of D notes given by Tu fted titmice in response to the predator and control treatments in the first 2min followi ng presentation. All pairwise comparisons were significant (LSD, p<0.05). Error bars: +/1 SE.

PAGE 82

82 1.0 0.5 0.0 -0.5 -1.0 PC1 1.0 0.5 0.0 -0.5 -1.0 PC2 11 2 1 7 5 4 3 6 8 9 10 12 13 14 15 16 17 Figure 3-9. Plot of 17 behavioral and general spectrographic variables of Tufted titmouse calls in two-dimensional space defined by two pr incipal components. PC1 is determined mostly by behavioral variable s, call rate, and note compos ition variables, and PC2 is mostly by temporal features of notes and calls. The variables corresponding to the numbers in the plot are listed in Table 3-2.

PAGE 83

83 DF14 3 2 1 0 -1 -2 -3 -4DF22 1 0 -1 -2 -3 TreatmentGroup Centroids 2screech 1great horned 0control 2 1 0 Figure 3-10. Graph of the results of the Discrimi nant Function Analysis for Tufted titmice with 4 PCA factor score input vari ables generated from 17 origin al behavioral and acoustic variables listed in Table 3-2.

PAGE 84

84 seet screech great horned control Playback treatment 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 Closest approach distance (m) Figure 3-11. Closest approach distance of Caro lina chickadees to the speakers during the playback treatments of Tufted titmouse vocaliz ations in the first 3min after the start of each playback. All pairwise comparis ons were significant (MWU, p<0.05) except between great horned and control (p=0.538). Error bars: +/1 SE.

PAGE 85

85 seet screech great horned control Playback treatment 0.8 0.6 0.4 0.2 0 Proportion Prop. of chickadees within 3m of speaker Prop.of chickadees within 1m of speaker Figure 3-12. The proportion of Carolina chickadees that approached within 1m and 3m of the speakers during the playback treatments of Tufted titmouse vocalizations in the first 3min after the start of each playback. No t all pairwise comparis ons were significant (see Appendix A, Table A-4). Error bars: +/1 SE.

PAGE 86

86 screech great horned control Playback treatment 3 2 1 0 Mean # of notes Mean # of D notes per call Mean # of chick notes per call Figure 3-13. Number of chick and D notes per chick-a-dee call given by Carolina chickadees in response to the different playback trea tments of Tufted titm ouse vocalizations in the first 3min after the start of each pl ayback. All pairwise comparisons were significant (LSD, p<0.05) except for the num ber of chick notes per call between the great horned mobbing and c ontrol playback (p=0.108). Error bars: +/1 SE.

PAGE 87

87 Table 3-1. Acoustic (first two colu mns) and behavioral parameters (3rd column) used in analyzing the response of Tufted titmouse floc ks to high-risk and low-risk predator presentations and c ontrol presentations. General spectrographic measures Measures of acoustic structure of D notes Behavioral measures *call ratenumber of calls minimum frequency where amplitude goes last below -10dB *closest distance any bird approached the stimulus *number of chick notes overall maximum frequency where amplitude goes last below -10dB *proportion of birds that came within 3m of stimulus *number of D notes overall minimum frequency where amplitude goes last below -30dB *proportion of birds that came within 1m of stimulus *number of notes per call maximum frequency where amplitude goes last below -30dB *whether the birds were frozen in place during the entire treatment *number of chick notes per call bandwidth at -10dB *number of D notes per call bandwidth at -30dB *proportion of chick notes per call entropy *duration of each chick note number of peaks above -10dB *duration of each D note *call duration duration of 1st D note of each call interval between notes interval between chick and D sections in each call interval between calls parameters also used in Experiment 2 with Carolina chickadee pairs

PAGE 88

88 Table 3-2. Factor loadings of the 17 behavioral and general spectrographic parameters on the four principal components after varimax rotation. Eigenvalues and amount of variance explained by the resp ective components are given at the bottom of the table. Parameter PC1 PC2 PC3 PC4 1 close approach -0.8840.295-0.142 0.230 2 prop. in 1m 0.777-0.4570.150 -0.021 3 prop. in 3m 0.700-0.253-0.046 0.071 4 call rate 0.833-0.1780.177 0.151 5 number of chick notes overall -0.407-0.0550.274 0.616 6 number of D notes overall 0.9570.023-0.151 -0.047 7 number of notes per call 0.760-0.152-0.557 -0.014 8 number of chick notes per call -0.944-0.1450.000 0.186 9 number of D notes per call 0.880-0.065-0.395 -0.190 10 prop. of chick notes per call -0.932-0.0510.168 0.250 11 duration of each chick note 0.0380.1480.015 -0.861 12 duration of each D note 0.0220.9360.017 -0.227 13 call duration 0.8500.184-0.429 -0.099 14 duration of 1st D note of each call -0.2030.9350.157 -0.060 15 interval between notes 0.0280.6090.308 0.515 16 interval between chick and D sections in each call -0.0750.2150.889 0.145 17 interval between calls -0.854-0.0420.275 0.193 Eigenvalue 8.262.671.83 1.73 % variance explained 48.615.710.8 10.2

PAGE 89

89 CHAPTER 4 CONCLUSION The Tufted titmouse is a vocally complex species that possesses a sophisticated anti-predator call system (Gaddis 1979, 1980, this st udy). I conducted several experiments to investigate characteristics of this system and summarize the results in Figure 4-1. In the figure, I present titmouse vocal and behavioral re sponses to highand low-risk predators and heterospecific responses to playbacks of the titmouse vocalizations (in the absence of the original predator stimulus ). In response to highand low-risk perched predators, titmice exhibite d mobbing behavior (Fig. 4-1a) and produced chick-a-dee mobbing calls with the high-ri sk Eastern screech-owl (screech owl) eliciting more intense mobbing (closer appro ach to the predator and more mobbing calls) than the low-risk Great horned owl (great horned) or control (see Fig. 3-4, 3-6). When Carolina chickadees (a heterospecific associat e of titmice in the wild) heard playbacks of these calls in the absence of the predator s timuli, they responded in a similar manner as the titmice by exhibiting mobbing behavior and producing chick-a-dee mobbing calls. Playbacks of titmouse mobbing calls produced in response to the high-risk screech owl elicited more mobbing calls and closer appro aches by the chickadees than playbacks of titmouse calls given in response to the great horned (Fig. 4-1a, second box; see Fig. 3-11; Chapter 3). In response to aerial predators (Gaddis 1980) or when startled by the sudden emergence of a potential predator (Chapter 2), which represent extremely high-risk situations, titmice sought cover or froze in place and produced seet alarm calls (Fig. 41b). When chickadees heard playbacks of titmouse seet calls in the absence of the original stimuli, they also responded by s eeking cover or freezing in place and becoming silent (Fig. 4-1b, second box; Chapter 3).

PAGE 90

90 It appears that titmice gi ve more mobbing calls and e xhibit more intense mobbing behavior as risk increases up to a point. But if the risk is too great, as is the case in an aerial predator encounter, mobbing is no longe r appropriate and titmice exhibit fearful behavior (giving alarm calls, freezing in pl ace and becoming inconspicuous). The fact that titmice, in general, give mobbing calls and exhibit mobbing behavior in response to highand low-risk perched pred ators that do not pose an imme diate threat and give alarm calls and become still and inconspicuous in response to aerial predators that do pose an immediate threat, can be explained by Morton s (1977) motivationstructural rules. The rules state that low-frequency, broadband s ounds (like titmouse mobbing calls) will be produced when the caller is in an aggressive state and is likely to attack; whereas highfrequency, pure tones (like titmouse alarm calls ) will be produced when the caller is nonaggressive or fearful. In my first experiment, I found that titmice do not produce predator-specific vocalizations that denote predator type or predator class (Chapter 2). They gave a combination of different vocalizations in res ponse to the control a nd predator treatments, which included avian, mammalian and reptilian predators (see Fig. 2-3) Seet alarm calls were occasionally given upon the removal of the cover from the presentation cage, and were most often given when the cage containe d a hawk or a cat, which were likely the highest risk predators presented (explained in Chapter 2; see Fig. 2-5). The sudden emergence of these predators in close proximity to individual titmice in this experiment represents an extremely high-risk situation. It makes sense that titmice would produce seet alarm calls to all the predators presente d in the manner used, but more of them in response to the predator species that represent the most risk. In this experiment the cat

PAGE 91

91 and hawk likely represented the highest risk (see Chapter 2) and these were also the predator treatments that elicited the highest mean number of seet calls. Additionally, chick-a-dee calls were given by titmice in response to all the treatments, but were most often given in response to the cat and haw k. According to Mortons (1977) motivationstructural rules, these combined results likely indicate that titmice were most fearful (more seet calls were produced) in response to the cat and hawk treatment at first, as the seet calls were mostly elicited at the very be ginning of the presentati on. After their initial response, the titmice were most aggressi ve (more mobbing calls were produced) in response to the hawk and cat treatments. Over all, the results of this experiment clearly show that titmice do not produce predator-speci fic vocalizations in re sponse to predator species or predator class (a vian, mammalian, reptilian), but instead may be producing risk-based anti-predator calls. Therefore, in the second experiment, I tested for and found that titmice produce situationally specific risk-based mobbing calls that vary acco rding to the degree of risk that a predator represents (Chapter 3). In addition to produci ng more chick-a-dee mobbing calls (see Fig. 3-6) and approaching the stimulus closer in response to the highrisk screech owl (see Fig. 3-4), titmice also varied the note com position of their mobbing calls (see Fig. 3-7) and varied several tem poral and frequency characteristics of their vocal response with respect to risk. Chickadees that heard playbacks of titmouse mobbing calls in response to hi ghand low-risk predators responded in much the same way that titmice responded to the actual pred ators (Fig. 4-1a). In addition to producing more mobbing calls and approaching closer (see Fig. 3-11) when they heard titmouse calls in response to the high-risk screech owl, chickadees also varied the note

PAGE 92

92 composition of the mobbing calls (see Fig. 313) and varied a few temporal note and call characteristics as a function of playback t ype. These combined results indicate that titmice possess an interspecific risk-based call system in response to predators (Fig. 41a). The chick-a-dee mobbing call and the seet alarm call appear to be functionally referential in terms of the type of predator encounter, as noted by Templeton et al. 2005, because, in general, titmice give mobbing calls in response to perched and terrestrial predators and seet calls in response to aeria l predators (Gaddis 1980; Evans et al. 1993; Blumstein 1999a). For Tufted titmice, the evidence for this claim is weak because neither of these call types ha s a high degree of production spec ificity (a requirement for functional reference) because both calls are gi ven in a variety of situations, including some non-predator situations. Fo r example, the seet call is of ten given in any situation in which the bird experiences alarm and can be evoked by the sudden emergence of any potential predator, eith er aerial or terrestrial (Ficken and Witkin 1977; Latimer 1977), as I found in Chapter 2. In this study, titmice so metimes gave seet calls during the control treatment in response to the removal of the c over from an empty cage (Chapter 2). Even though no predator was present, the movement of the cover likely startled the titmouse and resulted in the titmouse producing seet call s. Additionally, the chick-a-dee call of titmice is multifunctional and is given as a contact call, in coordinating group movements, and is also the primary predat or mobbing call (Gaddis 1979; S. A. Hetrick, pers. obs.). Therefore, both call types have low production specificity and there is weak evidence for functional reference.

PAGE 93

93 In summary, I present my four central c onclusions. First, I documented that titmouse anti-predator calls are not predator-specific with respect to predator species or predator class (avian, mammalian, reptilian; Ch apter 2), and this dispels some confusion in the literature about the produc tion of seet alarm calls a nd chick-a-dee mobbing calls. Seet calls have been described as hawk or aerial predator calls, and while they usually are given in response to a flying hawk, they are also given in other situations. Therefore, they are not reliably associ ated only with flying predator s. Second, I documented that titmouse chick-a-dee mobbing cal ls should be classified as risk-based in their production because the number and quality of the this call varied in a graded fashion as a function of risk. The third conclusion is that the nature of risk encoded in titmouse calls can be used by other species, as eviden ced by appropriate beha vioral responses by heterospecifics upon hearing the calls. This st udy was the first to document interspecific communication about such fine-scale differences in predation risk; i.e., that chickadees discern differences between calls given in re sponse to two different species of perched owls. Finally, the last two conclusions ta ken together indicate that titmice possess an interspecific risk-based call systema more sophisticated interspecific communication system with respect to predation risk than was previously known. Previous work showed that many species respond to titmous e mobbing calls, but we have improved understanding of these interactions by doc umenting that both the production and the perception of the calls by other species are ri sk-based. Titmice give specific information about different risk situations and heterosp ecific responses indicate that other species tailor their responses accordi ng to the specific le vel of risk being communicated. Many species that are sympatric with titmice res pond to their anti-predator calls and may be

PAGE 94

94 benefiting from the specific information that th ey contain. The posit ive benefits gained by receivers of predation risk informati on suggest that titmice may be playing an important facilitative role in animal communities.

PAGE 95

95A B Figure 4-1. The flow diagram summ arizes the vocal and behavioral responses of Tu fted titmice to predators representing varying degrees of risk and heterospecific respons es to playbacks of titmouse anti-predator vocalizations. The two shaded rows demonstrate the production of and responses to A) titmouse c hick-a-dee mobbing calls, and B) titmouse seet alarm calls.

PAGE 96

APPENDIX A SUMMARY TABLES OF STATISTICAL TESTS FOR RESPONSES OF TUFTED TITMICE TO PREDATORS AND RESPONSES OF CA ROLINA CHICKADEES TO PLAYBACKS

PAGE 97

97Table A-1. Behavioral responses of Tufted titmouse flocks to highand low-risk predators and controls in the 2min following presentation. One-tailed Mann-Whitney U tests were used to generate pairwise comparisons. scr = Eastern screech-owl, gh = Great horned owl, cont = control Measure KruskalWallis 2 df Asymp. Sig.* p-value screech versus great horned p-value screech versus control p-value great horned versus control p-value Multiple Comparisons at p<0.05 closest approach (m) 7.057 2 0. 0290.0530.019 0.594scr<(gh=cont) prop. in 1m 8.670 2 0.0130.0160.013 0.953scr>(gh=cont) prop. in 3m 7.355 2 0.0250.0530.008 0.953scr>(gh=cont) *Asymp. Sig.= Asymptotic significance

PAGE 98

98Table A-2. General spectrographic measures and measures of acoustic structure of D notes of Tufted titmouse calls elicited in response to highand low-risk predators and controls in the 2min following presen tation. scr = Eastern screech-owl, gh = Great horned owl, cont = control Measure TransformationANOVA p-value *Fdf Adjusted r2 screech versus great horned p-value screech versus control p-value great horned versus control p-value Multiple Comparisons at p<0.05 call rate (#/2min) sqrt 0.00112.22, 200.540 0.006<0.0010.200scr>(gh=cont) chick notes overall (#/2min) log(n+1) 0.3841.02, 200.001 0.7060.1980.381scr=gh=cont D notes overall (#/2min) log(n+1) <0.00116.22, 200.616 0.045<0.0010.007scr>gh>cont notes per call (#/2min) <0.001359.92, 2660.726 <0.001<0.001<0.001scr>gh>cont chick notes per call (#/2min) sqrt <0.001378.42, 2660.740 <0.001<0.001<0.001scrgh>cont prop. of chick notes per call arcsinsqrt 0.00116.32, 140.702 0.260<0.0010.002scrscr>cont duration of each D note (s) sqrt <0.00163.72, 14920.078 <0.0010.4990.001gh>(cont=scr) call duration (s) <0.001280.02, 2660.679 <0.001<0.001<0.001scr>gh>cont duration of 1st D note in each call (s) <0.00152.02, 2980.256 <0.001<0.0010.199scr<(gh=cont) interval between notes (s) sqrt <0.0017.72, 13130.010 0.0110.0010.049scrcont max. freq. where amplitude goes last below -10dB (Hz) 0.0214.03, 1050.054 0.5520.0190.007(scr=gh)>cont min. freq. where amplitude goes last below -30dB (Hz) <0.00113.12, 1050.189 0.065<0.0010.001(scr=gh)>cont max. freq. where amplitude goes last below -30dB (Hz) 0.7340.32, 105-0.013 0.5300.8210.480scr=gh=cont bandwidth at -10dB (Hz) 0.2791.32, 1050.006 0.6680.1950.119scr=gh=cont bandwidth at -30dB (Hz) <0.00110.72, 1050.157 0.047<0.0010.006scr
PAGE 99

99Table A-2 (cont) Measure TransformationANOVA p-value *Fdf Adjusted r2 screech versus great horned p-value screech versus control p-value great horned versus control p-value Multiple Comparisons at p<0.05 entropy 0.6070.52, 105-0.010 0.5240.6190.329scr=gh=cont peaks above -10dB (#/D note) 0.9880.02, 1050.019 0.9150.9440.879scr=gh=cont *Fdf = Fdf treatment, df total

PAGE 100

100Table A-3. Behavioral responses of Carolin a chickadee pairs to playbacks of Tufted titmouse vocalizations in response to highand low-risk predators and controls and titmous e seet alarm calls in the 3min following the start of each pla yback. One-tailed Mann-Whitney U tests were used to generate pairwise compar isons. scr = Eastern screech-o wl, gh = Great horned owl, cont = control Measure KruskalWallis 2 df Asymp. Sig.* p-value screech versus great horned p-value screech versus control p-value great horned versus control p-value screech versus seet p-value great horned versus seet p-value control versus seet p-value Multiple Comparisons at p<0.05 closest approach (m) 23.666 3 <0.0010.001<0. 0010.579 <0.0010.0430.004scr<(gh=cont)(gh=seet=cont) prop. in 3m 16.687 3 0.0010.0110.1050.436 <0.0010.2470.052scr>gh scr=cont gh=cont scr>seet gh=seet cont>seet *Asymp. Sig.= Asymptotic significance

PAGE 101

101Table A-4. General spectrographic measures of Carolina chickadee calls elicited in response to pl aybacks of Tufted titmouse vocalizations in response to highand low-risk predators and controls and titmouse seet alarm calls in the 3min following the start of each playback. scr = Eastern scre ech-owl, gh = Great horne d owl, cont = control Measure Transformation ANOVA p-value *Fdf Adjusted r2 screech versus great horned p-value screech versus control p-value great horned versus control p-value Multiple Comparisons at p<0.05 call rate (#/3min) sqrt 0.0274.12, 300.178 0.6090.0120.037(scr=gh)>cont chick notes overall (#/3min) sqrt 0.0752.92,300.113 0.6930.0770.034scr=gh scr=cont gh>cont D notes overall (#/3min) sqrt 0.0443.52, 300.148 0.3810.0150.097scr=gh scr>cont gh=cont notes per call (#/3min) <0.0019.12, 3380.046 0.009<0.0010.051scr>gh>cont chick notes per call (#/3min) <0.00132.32, 3380.157 <0.001<0.0010.108scr>(gh=cont) D notes per call (#/3min) sqrt <0.00127.62, 3380.137 <0.001<0.0010.010scr>gh>cont prop. of chick notes per call arcsinsqrt 0.0533.32, 270.152 0.3740.0180.096scr=gh scr(gh=cont) duration of each D note (s) <0.001119.92, 7750.235 <0.001<0.001<0.001scr>gh>cont call duration (s) sqrt <0.00127.82,3350.138 <0.001<0.0010.008scr>gh>cont *Fdf treatment, df total

PAGE 102

102 APPENDIX B SUMMARY TABLE OF MEAN RESPONSES OF TUFTED TITMICE TO PREDATORS AND CAROLINA CHICKADEES TO PLAYBACKS Table B-1. Mean responses, standard errors, and standard deviations of Tufted titmice to predator and control presentations in the 2min following presentation. Measure Predator N Mean SE SD closest approach (m) control 10 2.90 0.62 1.969 great horned 5 2.10 0.51 1.140 screech 5 0.76 0.15 0.336 prop. in 1m control 10 0.20 0.10 0.322 great horned 5 0.13 0.08 0.181 screech 5 0.80 0.13 0.299 prop. in 3m control 10 0.40 0.11 0.345 great horned 5 0.40 0.16 0.366 screech 5 0.93 0.07 0.148 call rate (#/2min) control 10 4.40 2.07 6.552 great horned 5 10.20 5.08 11.367 screech 5 35.00 5.01 11.203 chick notes overall (#/2min) control 10 6.50 3.09 9.767 great horned 5 6.60 2.09 4.669 screech 5 6.80 1.16 2.588 D notes overall (#/2min) control 10 4.20 2.32 7.345 great horned 5 79.60 36.40 81.402 screech 5 210.60 41.80 93.466

PAGE 103

103 Table B-1 (cont.) Measure Predator N Mean SE SD notes per call (#/2min) control 43 2.19 0.1195 0.8283 great horned 51 4.69 0.1357 0.9784 screech 172 6.21 0.0722 0.9469 chick notes per call (#/2min) control 43 1.49 0.0934 0.6126 great horned 51 0.49 0.4683 0.3344 screech 172 0.19 0.0100 0.1314 D notes per call (#/2min) control 43 0.95 0.0527 0.3459 great horned 51 4.29 0.9446 0.6746 screech 172 6.02 0.0775 1.0158 prop. of chick notes per call control 5 0.68 0.1053 0.2355 great horned 4 0.16 0.0850 0.1699 screech 5 0.04 0.0138 0.0307 duration of each chick note (s) control 64 0.0221 0.000599 0.004792 great horned 33 0.0278 0.000969 0.005564 screech 34 0.0241 0.000303 0.001768 duration of each D note (s) control 42 0.1860 0.00511 0.03311 great horned 398 0.2048 0.00200 0.03993 screech 1052 0.1821 0.00839 0.02722 call duration (s) control 43 0.3340 0.00618 0.04050 great horned 51 1.0617 0.01710 0.12215 screech 172 1.5200 0.02796 0.36556

PAGE 104

104 Table B-1 (cont.) Measure Predator N Mean SE SD duration of 1st D note of each call (s) control 34 0.1830 0.00660 0.3847 great horned 88 0.1927 0.00565 0.5303 screech 176 0.1454 0.00197 0.2607 interval between notes (s) control 62 0.0858 0.00210 0.01828 great horned 339 0.0812 0.00398 0.01179 screech 912 0.0791 0.001397 0.01480 interval between chick and D sections in each call (s) control 27 0.0641 0.56963 0.010897 great horned 26 0.0510 0.06375 0.020275 screech 34 0.0472 0.03606 0.008144 interval between calls (s) control 38 8.23 0.56963 3.5114 great horned 85 1.74 0.06374 0.5876 screech 170 1.37 0.03606 0.4701 min. freq. where amplitude goes last below -10dB (Hz) control 21 1864.00 85.05 380.227 great horned 36 2161.11 84.78 508.695 screech 50 2133.40 60.44 427.382 max. freq. where amplitude goes last below -10dB (Hz) control 21 6140.00 296.30 1357.501 great horned 36 6854.72 113.64 681.823 screech 50 6730.60 129.66 916.850

PAGE 105

105 Table B-1 (cont.) Measure Predator N Mean SE SD min. freq. where amplitude goes last below -30dB (Hz) control 21 668.95 29.54 128.750 great horned 36 841.94 34.90 209.391 screech 50 914.40 23.79 168.210 max. freq. where amplitude goes last below -30dB (Hz) control 21 7407.00 22.86 102.243 great horned 36 7426.11 17.50 104.970 screech 50 7412.80 12.44 87.971 bandwidth at -10dB (Hz) control 21 4242.00 282.70 1264.277 great horned 36 4688.89 144.65 867.877 screech 50 4593.20 142.88 1010.305 bandwidth at -30dB (Hz) control 21 6736.84 32.24 140.517 great horned 36 6580.00 38.93 233.556 screech 50 6494.00 26.01 183.937 entropy control 21 0.4347 0.00557 0.0255 great horned 36 0.4278 0.00504 0.0298 screech 50 0.4314 0.00310 0.0219 peaks above -10dB (#/D note) control 21 2.19 0.18 0.814 great horned 36 2.17 0.07 0.447 screech 50 2.18 0.07 0.523

PAGE 106

106 Table B-2. Mean responses of Carolina chic kadees to playbacks of Tufted titmouse vocalizations in response to predator and control presentations in the 3min following the start of each playback. Measure Playback N Mean SE SD closest approach (m) control 10 3.05 0.47 1.499 great horned mobbing 10 3.45 0.68 2.140 screech mobbing 10 0.70 0.13 0.420 seet 10 5.60 0.59 1.852 prop. in 1m control 10 0.10 0.07 0.211 great horned mobbing 10 0.20 0.11 0.350 screech mobbing 10 0.70 0.08 0.258 seet 10 0.00 0.00 0.000 prop. in 3m control 10 0.40 0.12 0.394 great horned mobbing 10 0.25 0.11 0.354 screech mobbing 10 0.70 0.08 0.258 seet 10 0.05 0.05 0.158 call rate (#/3min) control 10 4.90 1.62 5.109 great horned mobbing 10 12.10 2.47 7.795 screech mobbing 10 16.50 3.97 12.563 chick notes overall (#/3min) control 10 13.70 4.53 14.330 great horned mobbing 10 30.00 5.50 17.404 screech mobbing 10 28.60 6.96 22.006

PAGE 107

107 Table B-2 (cont.) Measure Playback N Mean SE SD D notes overall (#/3min) control 10 5.70 3.42 10.822 great horned mobbing 10 22.20 6.74 21.327 screech mobbing 10 49.80 24.43 77.270 notes per call (#/3min) control 51 3.80 0.17 1.222 great horned mobbing 121 4.27 0.09 1.040 screech mobbing 166 4.72 0.13 1.710 chick notes per call (#/3min) control 51 2.69 0.110 0.784 great horned mobbing 121 2.44 0.062 0.689 screech mobbing 166 1.72 0.085 1.092 D notes per call (#/3min) control 51 1.137 0.125 0.899 great horned mobbing 121 1.900 0.103 1.124 screech mobbing 166 3.000 0.174 2.249 prop. of chick notes per call control 8 0.86 0.061 0.172 great horned mobbing 10 0.66 0.072 0.229 screech mobbing 9 0.58 0.108 0.323 duration of each chick note (s) control 137 0.0377 0.000524 0.00613 great horned mobbing 300 0.0376 0.000386 0.00669 screech mobbing 286 0.0398 0.000339 0.00573

PAGE 108

108 Table B-2 (cont.) Measure Playback N Mean SE SD duration of each D note (s) control 55 0.0849 0.000416 0.00308 great horned mobbing 222 0.0907 0.000369 0.00549 screech mobbing 498 0.0994 0.000463 0.01034 call duration (s) control 49 0.3510 0.018360 0.12852 great horned mobbing 121 0.4279 0.010797 0.11877 screech mobbing 165 0.5782 0.021590 0.07700

PAGE 109

109 LIST OF REFERENCES Alatalo RV, Helle P (1990) Alarm calling by individual willow tits Parus montanus Anim Behav 40:437-442. Apel KM (1985) Antipredator behavior in the Black-capped Chickadee ( Parus atricapillus ). PhD diss (Univ of Wisconsin, Milwaukee). Baker MC, Becker AM (2002) Mobbing calls of black-capped chickadees: effects of urgency on call production. Wilson Bull 114:510. Bent AC (1937) Life histories of Nort h American birds of prey (part 1). US Natl Mus Bull 167:95-111. Bent AC (1946) Life histories of Nort h American jays, crows, and titmice. US Natl Mus Bull 191. Bildstein KL, Meyer K (2000) Sharp-shinned Hawk ( Accipiter striatus ). In The Birds of North America eds Poole A, Gill F (The Birds of North America, Inc., Philadelphia), no 482. Bloomfield LL, Phillmore LS, Weisman RG, Stur dy CB (2005) Note types and coding in parid vocalizations. III: The chick-a-d ee call of the Carolina chickadee ( Poecile carolinensis ). Can J Zool 83:820-823. Blumstein DT (1995a) Golden marmot alarm calls I. The production of situationally specific vocalizations. Ethology 100:113-125. Blumstein DT (1995b) Golden-marmot alarm calls : II. Asymmetrical production and perception of situationally specific vocalizations? Ethology 101:25-32. Blumstein DT (1999a) The evolution of functiona lly referential alarm communication: multiple adaptations; multiple constraints. Evol Commun 3:135-147. Blumstein DT (1999b) Alarm calling in three species of marmots. Behaviour 136:731-757. Blumstein DT, Armitage KB (1997a) Alarm-calling in yellow-bellied marmots. I. The meaning of situationally variable alarm calls. Anim Behav 53:143-171. Blumstein DT, Armitage KB (1997b) Alarm ca lling in yellow-bellied marmots: II. The importance of direct fitness. Anim Behav 53:173-184. Blumstein DT, Arnold W (1995) Situational-speci ficity in alpine-mar mot alarm communication. Ethology 100:1-13. Brackbill H (1970) Tufted Titmouse breeding behavior. Auk 87:522-536. Brawn JD, Samson FB (1983) Winter behavior of tufted titmice. Wilson Bull 95:222-232. Caro T (2005) Antipredator Defenses in Birds and Mammals (Univ of Chicago Press, Chicago).

PAGE 110

110 Cimprich DA, Grubb TC (1994) Consequences for Carolina Chickadees of foraging with Tufted Titmice in winter. Ethology 75:1615-1625. Curio E (1978) The adaptive significance of avian mobbing. I. Teleonomic hypotheses and predictions. Z Tierpsychol 48:175-183. Curio E (1980) An unknown determin ant of a sex-specific altruism. Z Tierpsychol 53:139-152. Dasilva KB, Kramer DL, Weary DM (1994) Cont ext-specific alarm ca lls of the eastern chipmunk, Tamias striatus Can J Zool 72:1087-1092. Davis RO (1991) Semantical communication in anti-pr edator alarm calls. In Natural History of Eastern California and High-altitude Research eds Hall CA, Jr, Doyle-Jones V, Widawski B (Univ of California, White Mountain Research Station, Los Angeles), pp 271-312. Dixon KL (1955) An ecological analysis of the interbreeding of crested titmice. Univ Calif Publ Zool 54:125-205. Dolby AS, Grubb TC, Jr (1998) Benefits to sate llite members in mixed species foraging groups: an experimental analysis. Anim Behav 56:501-509. Dolby AS, Grubb TC, Jr (2000) Soci al context affects risk taking by satellite species in a mixedspecies foraging group. Behav Ecol 11:110-114. Evans CS, Evans L, Marler P (1993) On the mean ing of alarm calls: functional reference in an avian vocal system. Anim Behav 46:23-38. Farley EA, Sieving KE, Contreras TA (in revi ew) Characterizing complex facultative mixedspecies bird flocks using an objective me thod for determining species participation. J Ornithol Fichtel C, Kappeler PM (2002) Anti-predator be havior of group-livi ng Malagasy primates: mixed evidence for a referential alarm call system. Behav Ecol Sociobiol 51:262-275. Ficken MS (1989) Acoustic characteristics of al arm calls associated with predation risk in Chickadees. Anim Behav 39:400-401. Ficken MS, Ficken RW, Witkin SR (1978) Vocal repertoire of the Black-capped Chickadee. Auk 95:34-48. Ficken MS, Hailman ED, Hailman JP (1994) Th e chick-a-dee call system of the Mexican chickadee. Condor 96:70-82. Ficken MS, Popp J (1996) A comparative analysis of passerine mobbing calls. Auk 113:370-380. Ficken MS, Witkin SR (1977) Responses of Bl ack-capped Chickadee flocks to predators. Auk 94:156-157.

PAGE 111

111 Freeberg TM, Lucas JR, Clucas B (2003) Variat ion in chick-a-dee ca lls of a population of Carolina chickadees, Poecile carolinensis : identity and redundanc y within note types. J Acoust Soc Am 113:2127-2136. Francis AM, Hailman JP, Woolfenden GE (1989) Mobbing of Florida scrub joys: behavior, sexual asymmetry, role of helpers and ontogeny. Anim Behav 38: 795-816. Frankenberg E (1981) The adap tive significance of avian mo bbing. 4. Alerting others and perception advertisement in blackbirds facing an owl. Z Tierpsychol 55:97-118. Gaddis P (1979) A comparative analysis of th e vocal communication sy stems of the Carolina Chickadee and Tufted Titmouse. PhD di ss (Univ of Florida, Gainesville). Gaddis PK (1980) Mixed flocks, accip iters, and antipredator behavior. Condor 82:348-349. Gaddis PK (1983) Composition and be havior or mixed-species flocks of forest birds in Northcentral Florida. Fla Field Nat 11:25-44. Gaddis PK (1985) Structure and va riability in the vocal repertoi re of the mountain chickadee. Wilson Bull 97:30-46. Gehlbach FR (1994) The Eastern Screech-Owl: li fe history, ecology, and behavior in suburbia and the countryside. (Texas A&M Univ Press, College Station). Gehlbach FR (1995) Eastern Screech-Owl ( Otus asio ). In The Birds of North America eds Poole A, Gill F (Academy of Natural Sciences, Ph iladelphia, and American Ornithologists Union, Washington, DC), no 165. Gompertz T (1961) The voca bulary of the great tit. Brit Birds 54:369-418. Greene E, Meagher T (1998) Red squirrels, Tamiasciurus hudsonicus produce predator-class specific alarm calls. Anim Behav 55:511-518. Greenberg R (2000) Birds of many feathers: the formation and stru cture of mixed-species flocks of forest birds. In On the Move: How and Why Animals Travel in Groups eds Boinski S, Garber PA (Univ of Chica go Press, Chicago), pp 521-558. Grubb TC, Jr, Pravosudov VV (1994) Tufted titmouse ( Parus bicolor ). In The Birds of North America eds Poole A, Gill F (Academy of Natural Sciences, Philadelphia, and American Ornithologists Union, Washington, DC), no 86. Gyger M, Marler P, Pickert R (1987) Semantics of an avian alarm call system: the male domestic fowl, Gallus domesticus Behaviour 102:15-39. Hailman JP (1989) The organization of major vocalizations in the Paridae. Wilson Bull 101:305343.

PAGE 112

112 Hailman JP, Ficken FS, Ficken RW (1985) The chick-a-dee calls of Parus atricapillus : a recombinant system of animal communi cation compared to written English. Semiotica 56:191-224. Hailman JP, Ficken FS, Ficken RW (1987) Constr aints on the structure of combinatorial chicka-dee calls. Ethology 75:62-80. Halliday TR, Adler K (eds) (1986) The Encyclopedia of Reptiles and Amphibians (Facts on File Inc., New York). Hamerstrom F (1957) The influence of a hawks appetite on mobbing. Condor 59:192-194. Hamilton WD (1963) The evoluti on of altruistic behavior. Am Nat 97:354-356. Hamilton WD (1964) The genetical theo ry of social behavior. I, II. J Theor Biol 7:1-52. Harrap S, Quinn D (1995) Chickadees, Tits, Nuthatches and Treecreepers (Princeton Univ Press, Princeton). Hill GE (1986) The function of distress calls given by Tufted Titmice ( Parus bicolor ): an experimental approach. Anim Behav 34:590-598. Hogstad O (1993) Alarm calling by willow tits, Parus montanus as mate investment. Anim Behav 49:221-225. Houston CS, Smith DG, Rohber C (1998) Great Horned Owl ( Bubo virginianus ). In The Birds of North America eds Poole A, Gill F (The Birds of North America, Inc., Philadelphia), no 372. Howell SE (2006) When titmice give the alarm, cardi nals listen: an experi mental case of interspecific functionally referential communication in wild birds. BS thesis (Univ of Florida, Gainesville). Hurd CR (1996) Interspecific a ttraction to the mobbi ng calls of Black-capped Chickadees ( Parus atricapillus ). Behav Ecol Sociobiol (38) 287-292. Hutto RL (1994) The composition and social organi zation of mixed-species flocks in a tropical deciduous forest in western Mexico. Condor 96:105-117. Jackson JA (1978) Predation by a gray rat sn ake on red-cockaded woodpecker nestlings. BirdBanding 49:187-188. Kirchhof J, Hammerschmidt K (2006) Functiona lly referential alarm calls in tamarins ( Saguinus fuscicollis and Saguinus mystax ) evidence from playback experiments. Ethology 112:346-354.

PAGE 113

113 Klump GM, Shalter MD (1984) Acoustic behavi our of birds and mammals in the predator context. I. Factors affecting the structure of alarm signals. II. The functional significance and evolution of alarm signals. Z Tierpsychol 66:189-226. Langham GM, Contreras TA, Sieving KE (in pr ess) Why pishing works: titmouse (Paridae) scolds elicit a generalized re sponse in bird communities. Ecoscience Latimer W (1977) A comparative study of the songs and alarm calls of some Parus species. Z Tierpsychol 45:414-433. Leavesley AJ, Magrath RD (2005) Communicati ng about danger: urgency alarm calling in a bird. Anim Behav 70:365-373. Le Roux A, Jackson TP, Cherry MI (2001) Does Brants whistling rat ( Parotomys brantsii ) use an urgency-based alarm system in reacti on to aerial and terr estrial predators? Behaviour 138:757-773. Macedonia JM (1990) What is communicated in th e antipredator calls of lemurs: evidence from playback experiments with ringtailed and ruffed lemurs. Ethology 86:177-190. Macedonia JM, Evans CS (1993) Variation am ong mammalian alarm call systems and the problem of meaning in animal signals. Ethology 93: 177-197. Manser MB (2001) The acoustic stru cture of suricates alarm calls va ries with predator type and level of response urgency. Proc R Soc London, Ser B 268:2315-2324. Marler P (1955) Characteris tics of some animal calls. Nature 176:6-8. Marler P (1957) Specific distinctiveness in the communication signals of birds. Behaviour 11:1339. Maynard Smith J (1965) The evolution of alarm calls. Am Nat 99:59-63. Mnkknen M, Forsman JT, Helle P (1996) Mi xed-species foraging aggregations and heterospecific attraction in boreal bird communities. Oikos 77:127-136. Morse DH (1970) Ecological aspe cts of some mixed species foraging flocks of birds. Ecol Monogr 40:119-168. Morse DH (1973) Interactions betwee n tit flocks and sparrowhawks Accipiter nisus Ibis 115:591-593. Morton ES (1977) On the occurrence and significan ce of motivation-structural rules in some bird and mammal sounds. Am Nat 111:855-869. Morton ES, Shalter MD (1977) Vo cal response to predators in pair-bonded Carolina Wrens. Condor 79:222-227.

PAGE 114

114 Munn CA, Terborgh JW (1979) Multi -species territoriality in neotropical foraging flocks. Condor 81:338-347. Naguib M, Mundry R, Ostreiher R, Hultsch H, Schrader L, Todt D (1999) Cooperatively breeding Arabian babblers call differently when mobbing in differe nt predator-induced situations. Behav Ecol 10:636-640. Nikolskii AA, Pereladova OB (1994) An alarm call of Menzbiers marmot ( Marmota menzbieri Kaschk., 1925). In Actual Problems of Marmots Investigation ed Rumiantsev VY (ABF Publishing House, Moscow) pp149. Nowicki S (1989) Vocal plasticity in captive bl ack-capped chickadees: the acoustic basis and rate of call convergence. Anim Behav 37:64-73. Nuechterlein GL (1981) Information parasitism in mixed colonies of western grebes and Forsters terns. Anim Behav 29:985-989. Odum EP (1942) Annual cycle of the black-capped chickadee-3. Auk 59:499-531. Packard RL (1954) Great Horned Owl attacking squirrel nests. Wilson Bull 66:272. Pereira ME, Macedonia JM (1991) Ringtailed lemur an tipredator calls denote predator class, not response urgency. Anim Behav 41:543-544. Pielou WP (1957) A life-history study of the Tufted Titmouse, Parus bicolor Linnaeus. PhD diss (Michigan State Univ, East Lansing). Pyle P (1997) Identification Guide to North American Birds (Slate Creek Press, Bolinas). Randall JA, Rogovin KA (2002) Variati on in and meaning of alarm calls in a social desert rodent Rhombomys opimus Ethology 108:513-527. Robinson SR (1980) Antipredator behaviour a nd predator recognition in Beldings ground squirrels. Anim Behav 28:840-852. Robinson SR (1981) Alarm communicati on in Beldings ground squirrels. Z Tierpsychol 56:150168. Seddon N, Tobias JA, Alvarez A (2002) Vocal communication in the pale-winged trumpeter ( Psophia leucoptera ): repertoire, context and functional reference. Behaviour 139:13311359. Seyfarth RM, Cheney DL, Marler P (1980a ) Vervet monkey alarm calls: semantic communication in a free-ranging primate. Anim Behav 28:1070-1094. Seyfarth RM, Cheney DL, Marler P (1980b) Monke y responses to three different alarm calls: evidence of predator classifica tion and semantic communication. Science 210:801-803.

PAGE 115

115 Seyfarth R, Cheney D (1990) The assessment by vervet monkeys of their own and another species alarm calls. Anim Behav 40:754-764. Shalter MD (1978) Mobbing in the pied flycat cher. Effect of expe riencing a live owl on responses to a stuffed facsimile. Z Tierpsychol 47:173-179. Shedd DH (1982) Seasonal variation and func tion of mobbing and related antipredator behaviours of the American robin ( Turdus migratorius ). Auk 99:342-346. Sherman PW (1977) Nepotism and the evolution of alarm calls. Science 197:1246-1253. Sherman PW (1985) Alarm calls of Beldings ground squirrels to aerial predators: nepotism or self-preservation? Behav Ecol Sociobiol 17:313-323. Shriner WM (1998) Yellow-bellied marmot a nd golden-mantled ground squirrel responses to heterospecific alarm calls. Anim Behav 55:529-536. Sieving KE, Contreras TA, Maute KL (2004) He terospecific facilitation of forest-boundary crossing by mobbing understory bird s in North-central Florida. Auk 121:738751. Smith RJF (1986) Evolution of animal signals: ro le of benefits of retaining group members or territorial neighbors. Am Nat 128:604-610. Smith ST (1972) Communication and Other Social Behavior in Parus carolinensis (Nuttall Ornithological Club, Cambridge). Sm ith SM (1991) The Black-capped Chickadee: Behavio ral Ecology and Natural History (Cornell Univ Press, Ithaca). Sokal RR, Rohlf FJ (1995). Biometry: The Principles and Prac tice of Statistics in Biological Research 3rd edition (WH Freeman and Co, New York). Spendlow JA (1979) Screech owl hunting during day light. N Am Bird Bander 1979:111. Struhsaker TT (1967) Auditory co mmunication among vervet monkeys ( Cerpithecus aethiops ). In Social Communication among Primates ed Altmann SA (Univ of Chicago Press, Chicago), pp 281-384. Sullivan K (1984) The advantages of social foraging in Downy Woodpeckers. Anim Behav 32:16-22. Templeton CN, Greene E, Davis K (2005) Allome try of alarm calls: Black-capped chickadees encode information about predator size. Science 308:1934-1937. Turner LJ, Dimmick RW (1981) Seasonal prey capture by the screech owl in Tennessee. J Tenn Acad Sci 56:56-59. Waite TA, Grubb TC, Jr (1987) Dominance, foraging and predation risk in the Tufted Titmouse. Condor 89:936-940.

PAGE 116

116 Waite TA, Grubb TC, Jr (1988) Copying of foragi ng locations in mixed-species flocks of temperate-deciduous woodland bi rds: an experimental study. Condor 90:132-140. Wenzel S (1997) Alarm calls and their function in avian predato r-prey interactions. BS thesis (Flinders Univ of South Australia). Windfelder TL (2001) Interspeci fic communication in mixed-sp ecies groups of tamarins: evidence from playback experiments. Anim Behav 61:1193-1201. Woolfendon GE, Fitzpatrick JW (1984) The Florida Scrub Jay: Demography of a CooperativeBreeding Bird (Princeton Univ Press, Princeton). Zimmerman U, Curio E (1988) Two conflicting n eeds affecting predator mobbing by great tits, Parus major Anim Behav 36:926-932. Zuberbhler K (2001) Predator-specific alarm calls in Campbells monkeys Cercopithecus campbelli Behav Ecol Sociobiol 50:414-422.

PAGE 117

117 BIOGRAPHICAL SKETCH Stacia Ann Hetrick was born and raised in sout hwest Florida. She is the youngest of three children and has two older brothe rs. Stacia was a curious youngste r and loved the outdoors. Her early pursuits in wildlife i nvolved chasing lizards and frogs, raising gerbils, cats, dogs, guinea pigs and, as might be predicted, a parakeet she cal led Peety. She took a keen interest when her father, who enjoyed watching a nd photographing long-legged wa ders, would stop at a pond or canal alongside the road to snap pictures of w ood storks and herons. Family boating trips also afforded her frequent opportunities for observing os preys, eagles, and coloni es of ibis as well as the occasional manatee, gator or water moccasin. Her interest in animals from an early age evolved to her pursuit of a major in wildlife ecology and conservati on. It was during her years as an undergraduate at the University of Florida th at she consciously bega n to realiz e her lifes passion would be ornithology. Stacia graduated from UF with a BS degree in the spring of 2002 after which she felt fortunate to be able to take on a masters project studying birds. In addition to her studies, discretionary time was spent volunteering, working bird-related field jobs, and making frequent scouting trips to the woods just to enjoy the simple wonders of nature. During her time as a masters student, she worked as a Teaching Assist ant and helped teach several wildlife classes to undergraduates. This led her to discover a sec ond passion in teaching. Stacia intends to pursue a career in environmental educati on and hopes to never stop teachi ng others about the wonderful world of birds.


Permanent Link: http://ufdc.ufl.edu/UFE0017822/00001

Material Information

Title: Investigation of Tufted Titmouse (Baeolophus bicolor) Anti-Predator Vocalizations
Physical Description: Mixed Material

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0017822:00001

Permanent Link: http://ufdc.ufl.edu/UFE0017822/00001

Material Information

Title: Investigation of Tufted Titmouse (Baeolophus bicolor) Anti-Predator Vocalizations
Physical Description: Mixed Material

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0017822:00001


This item has the following downloads:


Full Text





TITMOUSE (Baeolophus bicolor) ANTI-PREDATOR
VOCALIZATIONS


INVESTIGATION OF TUFTED


By

STACIA A. HETRICK


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2006

































Copyright 2006

by

Stacia A. Hetrick









ACKNOWLEDGMENTS

I would like to thank my advisor, Dr. Kathryn Sieving, for her endless guidance, patience,

and friendship. I thank my committee members, Dr. Michael Avery and Dr. Steven Phelps, for

their advice in the planning stages and throughout the duration of my study. I also thank Dr.

Michael Avery for his generosity in allowing me to use the USDA/APHIS/WS/NWRC Florida

Field Station to conduct my research. I would like to acknowledge the Ordway-Swisher

Biological Station for allowing me the use of their property. I thank the generosity of Tina

Brannon and Florida Wildlife Care, Inc. for allowing me the use of several raptors during my

study. My sincere thanks go to the numerous people who helped along the way: Dr. Thomas

Contreras and my fellow colleagues in Dr. Sieving's lab, Scarlett Howell, Jeremy Olson, Jennifer

Teagarden, Rebeccah Scarborough, Travis Blunden, Tracy and Clint Peters, Leander Lacy,

Aletris Neils, Kandy Keacher, John Humphrey, Eric Tillman, Mike Milleson, Dr. Thomas

Webber and all those who allowed me the use of their feeders for obtaining my study subjects. I

also want to thank Dr. Melvin and Fiona Sunquist for their thoughtfulness and inspiration. Most

of all, my tremendous gratitude goes to my parents, Denise and Gary Hetrick, for making my

schooling a priority and for always giving me their constant support in every way.












TABLE OF CONTENTS


page

ACKNOWLEDGMENT S .............. ...............3.....


LI ST OF T ABLE S ................. ...............6................


LI ST OF FIGURE S .............. ...............7.....


AB S TRAC T ......_ ................. ............_........9


CHAPTER


1 INTRODUCTION ................. ...............11.......... ......


2 ANTI-PREDATOR VOCALIZATIONS OF THE TUFTED TITMOUSE (Baeolophus
bicolor): DO THEY DENOTE PREDATOR SPECIES OR CLASS? ................. ...............13

Introducti on ............... .. ..........._ .. ....... ..... ...... .... .........1
Predator-Specific and Risk-Based Anti-Predator Calls............... ...............14.
Consequences of Specific Anti-Predator Calls............... ...............15.
Study System .............. ...............16....
Research Design .............. ...............19....
Hypothesis ............ _...... ._ ...............19....
Predictions ............ _...... ._ ...............20....
Methods .............. ..... ...............20.
Predator Presentations .............. ...............22....

Spectrographic Analyses .............. ...............22....
Spectrum-Based Measures .............. ...............23....
Re sults ................ ...............24.................
Discussion ................... .... .... ........ .... ...... .... ... ...........2
Encoding of Risk in Parid Anti-Predator Calls ................. ...............28........... ..
Potential Biases .............. ...............30....
Sum m ary ................. ...............3.. 1..............

3 INTERSPECIFIC RISK-BASED CALL SYSTEM OF TUFTED TITMICE

(Baeolophus bicolor) IN RESPONSE TO PREDATORS .............. ...............42....

Introduction and Background .............. ...............43....
Anti-Predator Vocal Signaling .................... ...............43.
Predator-Specific and Risk-Based Call Systems ......___ ..... ... ._ ..........__.....44
Study System ............... ......... .... .. .. .. ....... .............4
The Role of Tufted Titmice in Mixed-Species Foraging and Mobbing Flocks ..............46
Anti-Predator Calls of the Tufted Titmouse ........ ................. ............. ......4
Research Design .............. ...............48....
Hypotheses .............. ...............48....
Predictions ......... ................ ...............50 .....











Experiment 1: Titmice produce risk-based mobbing calls that are situationally
specific. ................... .......... .. ........... ..... ............5
Experiment 2: Chickadees exhibit interspecific perception specifieity to
titmouse anti-predator call s............... ...............5 1.
Methods I: Situational Specifieity Hypothesis .............. ...............51....
Predator Presentations .............. ...............52....
Spectrographic Analyses .............. ...............53....
Spectrum-Based Measures .............. ...............55....
R results I .............. ........ . ..... ................5
Methods II: Interspecific Perception Specifieity Hypothesis ................. .................5
Playback Presentations .............. ...............58....
Spectrographic Analyses .............. ...............60....
Results II............... ...............60...
D discussion ............... ... .... ...._ ..... ............ .. ....... .......... ........6
Situational Specificity of Titmouse Mobbing Calls with respect to Risk .......................62
Perception Specificity of Chickadees Exposed to Titmouse Anti-Predator Calls...........66
Titmice Give Interspecific Risk-Based Mobbing Calls in Response to Predators......... .68
Potential Functions of Interspecific Risk-Based Mobbing Calls of Titmice .................. .70
Summary ................. ...............72.................

4 CONCLU SION............... ...............8

APPENDIX

A SUMMARY TABLES OF STATISTICAL TESTS FOR RESPONSES OF TUFTED
TITMICE TO PREDATORS AND RESPONSES OF CAROLINA CHICKADEES TO
PLAY BACK S ................ ...............96........... ....

B SUMMARY TABLE OF MEAN RESPONSES OF TUFTED TITMICE TO
PREDATORS AND CAROLINA CHICKADEES TO PLAYBACKS .............. ................102

LIST OF REFERENCES ................. ...............109................

BIOGRAPHICAL SKETCH ................. ...............117......... ......










LIST OF TABLES


Table page

2-1 Mean number of each note type given by individual Tufted titmice in response to
predator and control presentations .......................__ ...............41. ...

3-1 Acoustic (first two columns) and behavioral parameters (3rd COlumn) used in
analyzing the response of Tufted titmouse flocks to high-risk and low-risk predator
presentations and control presentations. ............. ...............87.....

3-2 Factor loadings of the 17 behavioral and general spectrographic parameters on the
four principal components after varimax rotation. Eigenvalues and amount of
variance explained by the respective components are given at the bottom of the table. ...88

A-1 Behavioral responses of Tufted titmouse flocks to high- and low-risk predators and
controls ................. ...............97.................

A-2 General spectrographic measures and measures of acoustic structure of D notes of
Tufted titmouse calls elicited in response to high- and low-risk predators and
controls ................. ...............98._._._.......

A-3 Behavioral responses of Carolina chickadee pairs to playbacks of Tufted titmouse
vocalizations in response to high- and low-risk predators and controls and titmouse
seet alarm calls ........... __..... ._ ...............100...

A-4 General spectrographic measures of Carolina chickadee calls elicited in response to
playbacks of Tufted titmouse vocalizations in response to high- and low-risk
predators and controls and titmouse seet alarm calls............... ...............101.

B-1 Mean responses, standard errors, and standard deviations of Tufted titmice to
predator and control presentations. ........._._. ...._... ...............102...

B-2 Mean responses of Carolina chickadees to playbacks of Tufted titmouse
vocalizations in response to predator and control presentations............... .............10










LIST OF FIGURES


Figure page

2-1 Outdoor testing aviary............... ...............33.

2-2 Examples of the variation within the main vocalizations of the Tufted titmouse in
response to the 4 predator treatments and control. ............. ...............34.....

2-3 Number of each type of note or call given in the first 5min following presentation by
the fifteen individual Tufted titmice .............. ...............35....

2-4 Number of chick and D notes per 'chick-a-dee' complex call given by Tufted titmice
in response to predator treatments and control. ............. ...............38.....

2-5 Mean number of overall song, seet, chick, and D notes that Tufted titmice gave in
response to the predator treatments and control .............. ...............39....

2-6 Mean entropy of the D notes of Tufted titmice in response to predator treatments and
control .............. ...............40....

3-1 Examples of the maj or anti-predator vocalizations of the Paridae. ............._ ..............74

3-2 Outdoor aviary at the USDA/APHIS/WS/NWRC Florida Field Station in
Gainesville, Florida............... ...............75

3-3 Examples of the variation in the chick notes and the less variable D notes in the
'chick-a-dee' call complex of the Tufted titmouse ................ ............... ......... ...76

3-4 Closest approach distance of Tufted titmice to the stimuli during the predator and
control treatments............... ...............7

3-5 The proportion of Tufted titmice that approached within Im and 3m of the stimuli
during the predator and control treatments .............. ...............78....

3-6 Number of 'chick-a-dee' complex calls given by Tufted titmice in response to the
predator and control treatments .............. ...............79....

3-7 Number of chick and D notes per 'chick-a-dee' complex call given by Tufted titmice
in response to the predator and control treatments .............. ...............80....

3-8 Mean number of D notes given by Tufted titmice in response to the predator and
control treatments............... ...............8

3-9 Plot of 17 behavioral and general spectrographic variables of Tufted titmouse calls in
two-dimensional space defined by two principal components .............. ....................8










3-10 Graph of the results of the Discriminant Function Analysis for Tufted titmice with 4
PCA factor score input variables generated from 17 original behavioral and acoustic
variables ................. ...............83......... .....

3-11 Closest approach distance of Carolina chickadees to the speakers during the playback
treatments of Tufted titmouse vocalizations ....__ ......_____ .......___ ...........8

3-12 The proportion of Carolina chickadees that approached within Im and 3m of the
speakers during the playback treatments of Tufted titmouse vocalizations ......................85

3-13 Number of chick and D notes per 'chick-a-dee' call given by Carolina chickadees in
response to the different playback treatments of Tufted titmouse vocalizations............_...86

4-1 The flow diagram summarizes the vocal and behavioral responses of Tufted titmice
to predators representing varying degrees of risk and heterospecific responses to
playbacks oftitmouse anti-predator vocalizations............... .............9









Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

INVESTIGATION OF TUFTED TITMOUSE (Baeolophus bicolor) ANTI-PREDATOR
VOCALIZATIONS

By

Stacia A. Hetrick

December 2006

Chair: Kathryn E. Sieving
Major Department: Wildlife Ecology and Conservation

Tufted titmice (Baeolophus bicolor) are reported to produce different types of anti-predator

vocalizations in response to different predators and they are highly social with other species of

birds. The goals of this study included investigation of the anti-predator vocalizations of the

titmouse and determination of whether these calls contain information about the type of predator

detected (predator-specific) and whether they contain information about the risk of the situation

(risk-based). Additionally, I sought to determine whether sympatric bird species perceive and

respond appropriately to the information about predators encoded in titmouse anti-predator

vocalizations. In the first experiment I tested the hypothesis that titmice give predator-specific

vocalizations (unique vocalizations denoting the specific predator species or class) in response to

different species, including avian, mammalian, and reptilian classes of predators. Titmice

produced a combination of vocalizations in response to the predators and there was no evidence

that any particular vocalization denoted a specific predator species or class. Titmice varied the

note composition, note duration and note structure of their 'chick-a-dee' mobbing calls with

respect to predator type, which could indicate that they are producing risk-based mobbing calls

(in which the call structure varies as a function of risk). In the second experiment, therefore, I

designed a test to determine with certainty whether titmice produce risk-based mobbing calls.









The hypothesis tested addressed whether titmouse mobbing calls are situationally specific (call

structure varies according to the situation) in response to predators that represent different levels

of risk (i.e., high- and low-risk). The results indicated that titmice produce situationally specific

risk-based mobbing calls in response to predators by varying their call rate, note composition,

and frequency and temporal characteristics, as a function of predator risk. In a Einal experiment,

I tested whether Carolina chickadees (Poecile carolinensis) respond to titmouse risk-based

mobbing calls and alarm calls with perception specifieity (calls alone elicit appropriate response

in absence of original stimulus). Chickadees varied their behavior, note composition and

temporal characteristics as a function of playback type in much the same way that titmice

responded to the actual predators, indicating that the chickadees responded with perception

specifieity. Interspecific risk-based call systems, like the one characterized here, likely play an

important role in decreasing predation risk in animal social groups and, more generally, the

larger community of animal species. Interspecific communication systems represent potential

mechanisms underlying positive interactions, such as ecological facilitation, that help to structure

and maintain aggregations within vertebrate communities.









CHAPTER 1
INTTRODUCTION

Tufted titmice (Baeolophus bicolor) are nuclear species in mixed species flocks of birds

that form during the winter in eastern North America. In these flocks, predators are commonly

encountered and titmice are thought to give different anti-predator vocalizations (in different

situations) that are commonly identified as either 'seet' calls (also known as the 'hawk' or

'flying predator' calls) in response to aerial raptors or 'chick-a-dee' mobbing calls in response to

terrestrial predators. The goal of this study was to investigate the anti-predator vocal behavior of

the titmouse and to ascertain whether these calls contain information about the species or class

(avian, mammalian, or reptilian) of predator detected (predator-specific) or whether they contain

information about the risk of the situation (risk-based), or neither. Additionally, I sought to

determine whether sympatric bird species perceive and respond appropriately to the information

about predators encoded in titmouse anti-predator vocalizations.

First, I tested the hypothesis that titmice have predator-specific calls in response to

different types of predators. In chapter 2, I present Eindings of predator presentations (of a hawk,

an owl, a cat, and a snake) to individual, captive titmice, showing that they do not give unique

calls for different predator species or predator classes (avian, mammalian, reptilian), but vary

their vocal response to the different predators by altering note composition, call structure, and

temporal characteristics of their 'chick-a-dee' calls. This type of response is characteristic of

risk-based calls, where the calls vary in a graded manner according to the degree of risk. The

degree of risk that each predator posed (or that titmice presented with each predator would

perceive) in this first experiment was largely unknown, so I was unable to determine positively

whether titmice give risk-based calls in response to different predators. To determine










experimentally whether titmice vary their calls according to risk, I conducted a second

experiment, which I present in the third chapter.

The results from the first experiment led to the formation of two additional hypotheses,

which I address in chapter 3. The situational specificity hypothesis states that titmice will vary

the structure of their calls according to distinct situations, such as encounters with predators that

represent different degrees of risk. The interspecific perception specificity hypothesis states that

titmouse calls alone should elicit appropriate responses from heterospecifics in the absence of the

original stimulus. To address the first hypothesis, I presented captive titmouse flocks with high-

and low-risk predators and analyzed their behavioral and vocal response. To address the second

hypothesis, I presented captive Carolina chickadee (Poecile carolinensis) flocks with playbacks

of the titmouse vocalizations in response to the high- and low-risk predators as well as titmouse

seet calls and analyzed their behavioral and vocal response. In this chapter, I present my

findings that titmice do give risk-based calls to high- and low-risk predators by varying their call

rate, note composition, call structure and temporal features of their 'chick-a-dee' mobbing calls.

I also present findings that chickadees respond appropriately to these calls and to titmouse seet

calls by altering their behavioral and vocal response. These experiments have led to a better

understanding of how titmice communicate about the predation risk environment and how

chickadees, and potentially other sympatric species respond to titmouse vocal signals. These

signals may play a critical role in decreasing the predation risk environment for chickadees and

other sympatric species in this system.









CHAPTER 2
ANTI-PREDATOR VOCALIZATIONS OF THE TUFTED TITMOUSE (BAEOLOPHUS
BICOLOR): DO THEY DENOTE PREDATOR SPECIES OR CLASS?

Many species respond to predator encounters with specific vocalizations. Some species

have different calls that denote particular predator species or classes (predator-specific), while

some vary one or more of their vocalizations according to the degree of risk a predator represents

(risk-based). In this study, I wanted to determine whether Tufted titmice produce predator-

specific vocalizations in response to different predator species or classes (avian, mammalian,

reptilian). To reliably determine if titmice have calls that denote different predators, I presented

captive, adult titmice with four predators- a hawk, an owl, a cat, and a snake- and a control, all in

the same manner. I found that titmice most often produced a combination of different

vocalizations, including 'chick-a-dee' mobbing calls (composed of chick and D notes), 'seet'

alarm notes, contact (chip) notes, and song, none of which denoted a specific predator species or

class. They did, however, vary their vocal responses to the different treatments in terms of note

composition, note duration and note structure of their mobbing calls. The cat elicited the least

chick notes and the most D notes per call, followed by the hawk and owl, with the snake and

control eliciting the most chick notes and fewest D notes per call. In addition, the bandwidth and

entropy of the D notes elicited by the hawk and cat were greater than those elicited by the owl,

snake and control. These Eindings suggest that titmice may be responding according to the

degree of risk that the predators represent, rather than the predator species or class, indicating

that they may be producing risk-based calls.









Introduction

Predator-Specific and Risk-Based Anti-Predator Calls

Many animals give specific vocalizations when they encounter a predator. Some species

give predator-specific calls in which different call types are used to denote particular predator

species or classes. It has been well documented in the literature that many primate species give

structurally-distinct alarm calls to different types of predators (Seyfarth et al. 1980a, b;

Macedonia 1990; Pereira and Macedonia 1991; Zuberbiihler 2001; Kirchhof and

Hammerschmidt 2006). These calls have also been documented in some carnivore species,

namely the suricate (Suricata suricatta), a social mongoose, which gives distinct alarm calls to

terrestrial predators, avian predators and snakes (Manser 2001). Domestic chickens (Gallus

domesticus) also label predator classes by giving qualitatively different vocalizations to aerial

and terrestrial predators (Gyger et al. 1987; Evans et al. 1993).

In contrast, some animals vary their vocal response to predators according to the degree of

risk (also called response urgency) that the predator poses. Some species vary the rate in which

they call while others vary the quality of the calls they produce. Several species of marmot

(Marmota sp. ) vary the rate of their calls as a function of risk (Blumsteinl1995a; Blumstein and

Arnold 1995; Blumstein and Armitage 1997a). In particular, yellow-bellied marmots (Marmota

flaviventris) increase their call rate and potentially give calls with a larger bandwidth in response

to higher risk predator situations (Blumstein and Armitagel997a). When presented with

different degrees of risk, Mexican chickadees (Poecile sclateri) vary the pitch of a single kind of

alarm call according to the degree of risk (Ficken 1989) and Black-capped chickadees (Poecile

atricapilla) alter their call rate of 'chick-a-dee' calls (Baker and Becker 2002).









Consequences of Specific Anti-Predator Calls

Call specificity with respect to predator type might be adaptive if an animal has predators

that require different escape reactions. For example, in vervet monkeys, which have both

terrestrial and aerial predators, they run up into trees when they hear an alarm call denoting a

terrestrial predator and they look up, run into dense bush, or both when they hear an alarm call

denoting an aerial predator (Seyfarth et al. 1980a). In social species with different types of

predators, having predator-specific calls may allow group members to respond appropriately to

predator threats even if they themselves have not detected the predator. Contrastingly, risk-

based calls may give group members an indication of the threat level, but they do not necessarily

contain the predator-specific information that would allow for different specific escape

responses. Many passerines are said to have different calls that are given to aerial and terrestrial

predators (Marler 1957). If these calls are, if fact, predator-specific calls that label different

predator classes (i.e., aerial and terrestrial; avian, mammalian, and reptilian), this would lead to

receivers being able to choose specific escape responses. But if these calls denote the degree of

risk that is associated with the different types of predators, receivers would not have the

predator-specific information needed to choose a specific escape response. In animals that live

in stable groups, as some mixed-species flocks of birds do, evolution might favor calls that

provide predator-specific information to other flock members.

Additionally, some anti-predator signals may affect detected predators in different ways,

with some call types being potentially more efficient at deterring or distracting different

predators (Naguib et al. 1999). One possible function of anti-predator calls is to signal to the

predator that it has been detected and receptive predators might terminate the hunt rather than

expend their energy pursuing prey that are aware of it' s presence (reviewed in Smith 1986).

Evidence that supports this idea comes from observations by Morse (1973) of foraging accipiters









that did not typically attack if tit flocks gave alarm calls but would attack before they called,

although the specificity of the calls is unknown. Australian honeyeaters (Phylidonyris

novaehollandiae)1~~11~~~11~ may be communicating with predators by producing loud aerial alarm calls

that likely deter attacks from the predators by informing them that they have been sighted and

that the prey birds have already gone into hiding (Wenzel 1997). It has been suggested that

another important function of anti-predator calls, specifically mobbing calls, could be to warn

predators that they are about to be harassed and therefore, should retreat before they suffer

potential injury from the mobbers (Frankenberg 1981). The use of different types of calls by

prey species in response to predators may serve different functions and may play an important

role in influencing the subsequent behavior of the detected predator.

Therefore, it is important to understand these anti-predator calls because they may be a key

factor in decreasing the predation risk environment for signalers and receivers. Predator-specific

calls, in particular, are likely to exist in systems where animals are in stable groups and have

predators that require different escape techniques. In this study, I address whether the Tufted

titmouse, a common passerine that participates in mixed-species flocks that share different types

of predators in common (avian, mammalian, reptilian), possesses predator-specific calls in

response to these various predator types.

Study System

The subj ect of the present study is the Tufted titmouse (Baeolophus bicolor). The titmouse

is a common songbird in deciduous forests in eastern North America and is a regular visitor at

bird feeders, especially during the fall and winter. Titmice are year-round residents in the study

area of North-central Florida and participate in mixed-species flocks in winter (Gaddis 1979;

Farley et al. in review). These flocks typically contain one or more Tufted titmice, Carolina

chickadees (Poecile carolinensis) and usually include several attendant or 'satellite' species.









Most flocks contain a pair of titmice, their offspring, and/or other unrelated juveniles (Pielou

1957; Brackbill 1970). Regular satellite species include: Black-and-white warblers (M~niotilta

varia), Downy woodpeckers (Picoides pubescens), Ruby-crowned Kinglets (Regulus calendula),

Blue-headed and White-eyed vireos (Vireo solitarius, y. griseus) and Blue-gray gnatcatchers

(Polioptila caerulea; Farley et al. in review).

Titmice and potentially Carolina chickadees play the role of the 'nuclear', or focal species,

around which mixed-species foraging flocks form during the winter months and the other flock

members play the role of 'satellite' species (Gaddis 1983; Grubb and Pravosudov 1994;

Greenberg 2000). Nuclear species in mixed-species bird flocks are generally characterized by

behavioral traits that include dominance, sociality, and a high level of vigilance (Munn and

Terborgh 1979; Hutto 1994). In addition, parid nuclear species act as sentinels by readily giving

vocalizations in response to predators, thereby potentially alerting flock members of danger

(Gaddis 1983; Dolby and Grubb 1998).

Typical predators of titmice include feral and house cats (Felis domesticus), hawks, owls,

and snakes (Bent 1946). In my study area, the most common predators of forest passerines in

winter include Sharp-shinned (Accipitor striatus) and Cooper's hawks (A. cooperii); Eastern

screech-owls (M~egascops a~sio); Red-shouldered (Buteo lineatus) and Red-tailed hawks (B.

jamnaicensis); American kestrels (Falco sparverius); and feral and house cats (Sieving et al.

2004; S. A. Hetrick, pers. obs.). Snakes, most commonly rat snakes (Elaphe sp.), typically prey

on the eggs, nestlings and sometimes adults of titmice and other small birds during the summer

months (Jackson 1978; Halliday and Adler 1986; S. A. Hetrick, pers. obs.). During the winter

months in eastern North America, Sharp-shinned hawks are most likely the most important









predator of small woodland birds (Bent 1937; Morse 1970; Bildstein and Meyer 2000), including

titmice and other flock associates (Gaddis 1979, 1980; S. A. Hetrick, pers. obs.).

Two main types of anti-predator vocalizations have been described for Tufted titmice: the

'seet' alarm call (also known as the high whistle, see-see-see, 'hawk' call, and 'flying predator'

alarm call) and the mobbing or scold call (known as seejert, chick-a-dee; Dixon 1955; Marler

1955; Gaddis 1979, 1980). Titmouse mobbing calls are variants of the 'chick-a-dee' call, which

is a complex call composed of combinations of introductory 'chick' notes, and subsequent 'D'

notes (dee notes, churr notes), with the number and presence of each note type being variable

(Latimer 1977; Hailman 1989). The broadband structure of the D notes in the mobbing calls

causes them to be easily localizable, while the pure tone structure of seet calls causes them to be

difficult to locate (Marler 1955). When mobbing calls are given in response to predators, many

birds are attracted to the area and may harass the predator, whereas seet alarm calls result in the

cessation of movement (freezing) by the caller and nearby birds or in rapid escape to cover

(Gaddis 1980; Ficken 1989; Baker and Becker 2002; Howell 2006; S. A. Hetrick, pers. obs.).

Although many studies have focused on aspects of the vocal repertoire of birds in the

family Paridae, surprisingly little attention has been devoted to the vocalizations of Tufted

titmice, especially in the anti-predator context (Dixon 1955; Gaddis 1979, 1980, 1983; Hill

1986). It is not known whether titmice have specific vocalizations for different predators,

although it is commonly stated that members of the family Paridae give 'seet' or 'hawk' calls to

raptors flying overhead (Marler 1955; Latimer 1977; Harrap and Quinn 1995). Titmice could be

giving seet calls to (1) denote the predator as a hawk or accipiter (predator-specific calls with

respect to species); (2) denote the class of predator as avian (predator-specific calls with respect

to class); (3) denote the class of predator as aerial (predator-specific calls with respect to class);










(4) signal the immediate degree of danger (risk-based calls); or (5) any combination of these. No

study has yet to confirm whether titmice use the seet call to reliably denote accipiters, aerial or

avian predators, or to denote a situation where the titmouse is in immediate risk of attack. As

Templeton et al. (2005) noted, previous studies have presented aerial and terrestrial predators in

different ways (Greene and Meagher 1998; Blumstein 1999b; Le Roux et al. 2001), which

confounds the interpretation of responses given to predator type versus risk situation (i.e.,

predator proximity, location, or behavior). In order to determine if the titmice were able to

distinguish between predator species or between avian, mammalian, and reptilian predator

classes, all predators were presented in the same manner in this experiment.

Research Design

The purpose of this study was to investigate the anti-predator vocal behavior of the Tufted

titmouse; specifically, whether titmice have specific vocalizations for different predators

(predator-specific calls). This study dealt with the vocal responses of titmice to both stuffed

predator mounts and live predators- avian, reptilian and mammalian. Wild-caught birds were

tested in a captive situation with both the bird and the predator being in cages. I presented

individual adult titmice with 4 different predators: an owl, a hawk, cat, snake and a control, all in

the same manner.

Hypothesis

I tested the hypothesis that titmice have predator-specific calls (call type uniquely covaries

with predator species or class). In order to have predator-specific vocalizations, the titmice must

give different call types that are predictably associated with different predator classes or species

(Blumstein 1999a). Titmice could separate the predators into classes in various ways. For

example, they may separate them into classes according to whether the predator is aerial or

terrestrial; whether the predator is avian, mammalian, or reptilian; or, they could be more










specific and separate them into species such as hawk, owl, cat, and snake. In this experiment, I

tested the latter two of these possibilities.

I conducted a true experiment under controlled conditions by presenting four different

predator species in a similar manner to individual adult titmice to obtain vocal recordings of

titmouse calls in response to the presentations. Subsequent analyses of the recordings were

conducted to determine whether titmice give specific vocalizations in response to the different

predators. All predators were presented in a clear cage directly across from the cage containing

the titmouse. Iftitmice give predator-specific vocalizations that label either predator species or

predator class (avian, mammalian, or reptilian), then the manner in which the predators are

presented should not matter. For example, iftitmice have specific 'hawk' calls that denote

hawks, they should respond with 'hawk' calls to aerial as well as perched hawks. On the other

hand, if titmice respond according to the degree of risk, one or more of their calls would likely

vary in response to an aerial versus a perched hawk (assuming that the two situations represent

different degrees of risk).

Predictions

I predicted that the titmice would respond with different call types to the different types of

predator being presented. I predicted that they would give predator-specific vocalizations that

are associated either with predator class (avian, mammalian, or reptilian) or with individual

predator species. In the first case, the titmice would have different vocalizations for the hawk

and owl (avian) than for the cat (mammalian) and snake (reptilian) and in the second case, the

titmice would have unique calls for one or more of the four different species.

Methods

I examined the responses of adult titmice to four different predator species and a control.

Fifteen adult titmice were caught between November 6, 2004 and January 21, 2005 at various









locations in Florida--the Ordway-Swisher Biological Station in Melrose, the

USDA/APHIS/WS/NWRC Florida Field Station (United States Department of Agriculture,

Animal & Plant Health Inspection Service, Wildlife Services, National Wildlife Research

Center; USDA lab) in Gainesville, FL and various residences in the city of Gainesville. The birds

were captured using mist-nets and transferred to the USDA lab. The birds were housed and

tested in 0.5 x 0.5 x 0.5m cages containing several branches for perching. They were fed an ad-

lib diet of mixed seed, mealworms, suet, and chopped fruit and vegetables.

In order to be sure of acquiring appropriate anti-predator responses, I used only adult

titmice in the experiment. Chickadees acquire information about predator identity through

learning from older birds in their social groups (reviewed in Smith 1991). Since titmice have a

similar social system to chickadees, I expected that young-of-the-year may not give informative

responses in this experiment. To determine if the birds were adults, I used the molt limit criteria

in Pyle (1997) for aging.

At least 24 hours before the first predator presentation, the cage containing the bird was

brought into the testing environment in order for the bird to acclimate. The testing environment

was an outdoor aviary (9 x 3 x 2.3m) at the USDA lab that contained numerous branches and

snags and was adjacent to forest, providing a semi-natural environment. Within the large aviary,

the cage containing the test subj ect was placed on a 0.5m platform with the predator presentation

cage 0.75m away from the test cage on a 0.5m platform (Fig. 2-1). The predator presentation

cage was made of clear plexi-glass and a sheet covered it at all times except during the tests. A

camouflage blind containing the researcher and a microphone were also within the aviary, 4.5-

6m away from the test and predator cages.









Predator Presentations

Each bird underwent a series of 5 different treatments after the acclimation period. Each

test consisted of one of 4 predator treatments a stuffed Sharp-shinned hawk (Accipiter striatus),

a live Eastern screech-owl (M~egascops a~sio), a live domestic house cat (Felis domesticus), and a

live red rat snake (Elaphe guttat) or a control (an empty cage) being presented in random

order (hereafter the treatments will be referred to as hawk, owl, cat, snake and control). The tests

began at 0800 and were conducted every 2 hours until 1600. Each test consisted of one of the

predators being placed into the cage 20 minutes before testing began. For the control, the

researcher went through the steps as if they were placing a predator in the cage. Care was taken

to ensure that the bird in the test did not view the predator until the test began. The test began

when the sheet covering the predator cage was removed, allowing the bird to view the predator

or empty cage. Recordings were made for 5min pre-stimulus and 7min post-stimulus; however,

only the first 5min post-stimulus were included in the analysis. A Sennheiser shotgun

microphone (1VE 66) was used to record the vocalizations directly onto a laptop computer using

Raven Interactive Sound Analysis Software Version 1.1 with a sampling rate of 44100 at 16-bit

resolution.

Spectrographic Analyses

I analyzed vocal responses for 72 out of the 75 presentations for the adult titmice (n=14 for

cat, n=15 for control, n=15 for owl, n=15 for snake, and n=13 for hawk); the titmice did not

vocally respond in 2 of the presentations and one of the hawk recordings was lost due to

equipment failure. Spectrographic analyses were performed on the vocal recordings using

Avisoft SASLabPro 4.39. To edit out noise, each sound Eile was FIR low-pass filtered at 12k
and high-pass filtered at 1.4k
Size=75%, Window=Hamming, Overlap=87.5%. I classified and labeled all vocalizations given










in response to the treatments, which included 'chick-a-dee' calls, songs, seet notes, and chip

notes. The notes in each call, or 'chick-a-dee' complex, were visually classified as introductory

'chick' notes or subsequent 'D' notes. In titmice, the various introductory notes grade into each

other and are not easily distinguished into natural categories; therefore, the introductory notes

were classified together as 'chick' notes (Fig. 2-2a). On the other hand, D notes can be reliably

classified (Bloomfield et al. 2005) due to their harmonic-like structure and little frequency

modulation. D notes also have a higher entropy and lower frequency than the introductory chick

notes, and D notes always occur at the end of the call, or are the only notes comprising a call,

making them easily distinguished from introductory notes (Fig. 2-2a). Songs, seet notes and chip

notes were also visually classified. Songs are unique and can be easily distinguished from other

vocalizations (Fig. 2-2b). Seet notes were recognized by being high-pitched (around 8-10k
for titmice) whistles with a narrow bandwidth that have no sharp onset or ending and cover only

a narrow frequency range (Fig. 2-2c; Apel 1985; Marler 1955). Chip notes were recognized as

being single-syllable notes that are typically chevron-shaped with a shorter duration than seet

notes (Fig. 2-2d).

I measured several aspects of the notes and calls, including some that were based on

measures used in previous studies of parids (Baker and Becker 2002; Freeberg et al. 2003;

Templeton et al. 2005). For each treatment, I averaged the number of each type of vocalization

(chick-a-dee, song, seet, chip). For the 'chick-a-dee' calls, I averaged the number of chick and D

notes overall, the number and proportion of chick and D notes per call, and the duration of each

chick and D note.

Spectrum-Based Measures

I also measured several fine-scaled acoustic parameters on the D notes in which I had high-

quality recordings (not overlapping with outside noise). I measured the spectrum-based










parameters of the notes using a power spectrum with FFT length=5 12. For the D notes, the

parameters were computed at the maximum spectrum of the entire D note (maxpeakhold) and are

similar to those used by Nowicki (1989) and Templeton et al. (2005). The parameters were

minimum and maximum frequency where the amplitude goes last below -30dB and where the

amplitude goes last below -10dB (min and max frequency with the total option activated in

Avisoft SASLabPro 4.39), bandwidth (calculated with min and max frequency described

previously) and entropy.

For each factor that was measured, I used univariate analysis of variance (ANOVA) with

the least significant difference (LSD) post-hoc test to conduct pairwise comparisons among the

treatments. I transformed the data when appropriate to meet the assumptions of the analysis

using arcsin(sqrt(n)) and log(n+1) transformations (Sokal and Rohlf 1995). All statistical

analyses were performed using SPSS 11.5 for windows and significance in all statistical tests

was set at the 0.05 alpha-level.

Results

Contrary to my predictions, the titmice did not give specific vocalizations to denote the

different predator species or predator classes (avian, mammalian, reptilian). The titmice

typically gave a combination of vocalizations in response to the predator and control treatments,

usually consisting of chip notes and possibly 'chick-a-dee' calls, seet notes, or both. Their vocal

responses were highly variable (see Fig. 2-3 a-e). Chip notes were by far the most common

vocalization: 68 of the 74 five-min recordings contained chip notes (n=14 for control, n=15 for

snake, n=13 for owl, n=13 for hawk and n=13 for cat). About an eighth to a quarter of the birds

responded with 'chick-a-dee' calls to the control, snake and cat treatments and about half the

birds responded with 'chick-a-dee' calls to the owl and hawk treatments (n=2 for control, n=3 for

snake, n=7 for owl, n=8 for hawk and n=4 for cat). More titmice gave seet calls in response to









the hawk and cat than to the control, snake and owl treatments (n=4 for control, n=2 for snake,

n=5 for owl, n=9 for hawk and n=7 for cat). Only a few birds responded with singing when

presented with the predators (n=1 for control, n=1 for snake, n=3 for owl, n=2 for hawk and n=1

for cat). The means of each type of vocalization given to the different treatments are shown in

Table 1.

Six out of 11 of the general spectrographic variables and all of the 7 Eine-scale acoustic

variables of the D notes varied significantly with treatment (ANOVA p<0.05). In no case did the

LSD pairwise comparisons indicate that titmice discriminated each predator from the other and

the control. In most cases, the cat, hawk and owl were significantly different from the snake and

control ((cat, hawk, owl) / (snake, control)) or the cat and hawk were significantly different from

the owl, snake and control ((cat, hawk) / (owl, snake, control)).

Titmice did not give a significantly different number of 'chick-a-dee' mobbing calls to the

different treatments (ANOVA F4, 74=0.41, p=0.798), although the number of introductory chick

notes per call and D notes per call differed significantly with treatment (F4, 133=44.34, p<0.001;

F4, 133=21.96, p<0.001, respectively). The control and snake treatments elicited more chick notes

per call and fewer D notes per call than the owl, cat and hawk. The cat treatment elicited the

fewest chick notes and most D notes overall followed by the hawk and then the owl. The

relationship can be clearly seen in Figure 2-4. The proportion of chick notes per call also

differed significantly with treatment (F4, 133=56.34, p<0.001), with the relationships between the

treatments the same as above. The number of seet, song, chip, chick and D notes did not differ

with respect to treatment (p>0.157 in all cases, Fig. 2-5).

The duration of the chick and D notes was significant with respect to treatment (F4,

134=3 .69, p=0.007; F4, 299=20.29, p<0.001). The pairwise comparisons revealed that the chick









notes in response to the cat were of greater duration than those given in response to the control

and owl; the chick notes in response to the snake were of greater duration than those given in

response to the control and owl; and the chick notes in response to the hawk were of greater

duration than those given in response to the owl (with no other comparisons significant). The

pairwise comparisons revealed that the D notes in response to the hawk were of greater duration

than those given in response to all other treatments and the D notes in response to the cat were of

greater duration than those given in response to the owl (with no other comparisons significant).

Of the 7 Eine-scaled acoustic parameters that were measured, all were significant. The

parameters that had multiple comparisons with a clear relationship between the treatments were

maximum frequency where amplitude goes last below -10dB, bandwidth at -10dB and entropy.

The titmice gave D notes with a larger bandwidth and higher maximum frequency at -10dB to

the cat and hawk treatments than to the owl, control and snake treatments. Additionally, the D

notes that the titmice produced in response to the hawk had a higher entropy than in response to

the cat; both hawk and cat treatments elicited D notes with a higher entropy than the owl, control

and snake treatments (Fig. 2-6).

Discussion

Contrary to my predictions, the results indicate that the titmice did not use different call

types to label different predator species or predator classes (avian, mammalian, reptilian);

therefore, they do not possess predator-specific calls in these contexts. They gave a variety of

notes and call types during the predator presentations (chick-a-dee, seet, song, and chip), none of

which were reliably associated with a particular predator. All of the predator presentations and

occasionally the control presentation elicited seet notes from some of the titmice. Dixon (1955)

and Marler (1955) have called the titmouse seet call the 'hawk' call or 'flying predator' call. But

in this study titmice gave seet notes in response to other predators besides hawks and to predators









that were perched (not flying). The results from this experiment show that titmice do not vocally

discriminate (by giving different call types) between predator species when they are presented in

similar ways. In addition, the results indicate that seet alarm calls are given in other

circumstances besides just in response to aerial predators.

Although the titmice didn't give predator-specific vocalizations, they did vary the call and

note structure of one specific vocalization, the 'chick-a-dee' mobbing call, in response to the

different predators, which is characteristic of risk-based calls (Blumstein and Armitage 1997a).

One way in which they varied their calls was by altering the note composition of the 'chick-a-

dee' mobbing call. The cat treatment elicited the fewest chick notes and the most D notes,

followed by the hawk and owl treatments. The snake and control treatments, which did not

differ significantly from each other, elicited the most chick notes and the fewest D notes.

Gompertz (1961) and Latimer (1977) noted that some parids decrease or drop the introductory

notes of the mobbing call while extending the D note section as the level of risk increases. If this

were true for Tufted titmice, then it would lead us to believe that the cat treatment represents the

greatest risk, followed by the hawk and owl treatments, with the snake treatment representing the

same amount of risk as the control. Evans et al. (1993) suggested that, under some conditions,

terrestrial predators pose a greater threat than avian predators. Given that all birds used in this

study were captured in or near suburban neighborhoods where outdoor cats are common around

bird feeders frequented by titmice and other species (S. A. Hetrick, per. obs.), all test subj ects

would likely have experience with this species. Unlike a perched raptor (species dependent upon

aerial attack to capture prey), a crouching cat in close proximity to a titmouse poses a very high

risk to the adult bird. Therefore it is quite likely that the titmice in this study could have viewed

the cat treatment as representing the greatest risk.









Encoding of Risk in Parid Anti-Predator Calls

Note composition may be particularly important in encoding information about predators

and degree of risk. Studies of other birds in the family Paridae, namely chickadees, suggest that

different variants of the 'chick-a-dee' call might encode information about the presence of

different environmental stimuli (including predators) or the motivational state of the caller

(Smith 1972; Gaddis 1985; Ficken et al. 1994). Hailman et al. (1985, 1987) suggested that note

composition variation may encode information related to many factors, including potential

predators, and the results from Ficken, Hailman, and Hailman's (1994) study with Mexican

chickadees supports that view. Black-capped chickadees altered the number of introductory A

and B syllables in their 'chick-a-dee' calls in response to a predator mount presented at different

distances, which represented different risk levels (Baker and Becker 2002). They also altered the

number of D notes in the 'chick-a-dee' call in response to raptors of different sizes, with the

smallest raptors eliciting the most D notes and the largest raptors eliciting the least (Templeton et

al. 2005). The small raptors represented higher-risk situations to the chickadees and the large

raptors represented lower risk situations, thus it is likely that the chickadees responded according

to the degree of risk that they encountered.

Besides varying the note composition of the 'chick-a-dee' calls, I found that titmice also

altered the structure of the D notes within the calls. Titmice gave D notes with a larger

bandwidth and higher maximum frequency in response to the hawk and cat, while the owl, snake

and control elicited D notes with a smaller bandwidth and lower maximum frequency. Other

animals change their call structure in different predator situations. Yellow-bellied marmots vary

several frequency characteristics of their calls, including bandwidth, as a function of distance to

certain predators (Blumstein and Armitage 1997a). White-browed scrubwrens (Sericornis

fr~ontalis) vary the call structure of their aerial trill call by increasing the minimum frequency









(pitch) according to the distance from a suddenly appearing predator (Leavesley and Magrath

2005). In both of the aforementioned studies, the distance from the predator to the subject

represents the amount of risk, with the closer distances representing higher risks; thus, we can

conclude that both the marmot and the scrubwren vary the structure of their calls according to the

degree of risk that they are presented with. In the present study, the hawk and the cat treatments

elicited D notes from the titmice that had higher entropy than the D notes elicited from the owl,

snake and control treatments. Entropy is a measure of the amount of randomness or 'noise' a

note contains, with pure tones having no entropy and white noise having the most entropy;

subsequently, entropy can be used as an indication of note harshness. Latimer (1977) noted that

many birds in the family Paridae give calls that are harsher (higher entropy) as the level of

aggression rises. Additionally, Morton (1977) documented that the motivational state of the

caller influences signal structure and that harsh (high entropy), broadband (large bandwidth),

low-frequency sounds are associated with aggressive behavior, whereas more tonal, high-

frequency sounds are associated with non-aggressive or fearful behavior. Because the hawk and

cat treatments elicited D notes with larger bandwidth and higher entropy (harshness), we can

speculate that the titmice perceived a higher level of risk and had a higher level of aggressiveness

in response to the hawk and cat than to the owl, snake and control treatments. Thus, I conclude

that the level of risk according to titmice may be associated, at least in part, with the type of

predator and not solely the proximity or location of a predator with respect to the test subj ect.

The results of both the note composition and D-note structure suggest that the titmice

perceived the cat and hawk as having the most risk, followed by the owl and the snake and

control had the least risk. In addition, more individual titmice responded with seet notes to the

high-risk species; and the mean number of seet notes was highest in response to the cat and hawk










presentations. This evidence further supports the idea that the titmice viewed these two

predators as posing the most risk. Apel (1985) concluded that Black-capped chickadees

recognized the difference in the degree of danger between Sharp-shinned hawks and American

kestrels by responding with more seet notes ('high sees') to the higher-risk Sharp-shinned hawk.

It is interesting to note that the titmice did note respond differently to the snake and control

in terms of note composition and D-note structure, thus suggesting that the titmice did not

perceive the snake as more of a risk than the control. There are several explanations as to why

this occurred. The time of year that the study was conducted was winter, which could have

affected the titmouse's response to the snake, as red rat snakes prey on birds and their eggs

typically during the summer when most birds are nesting (Jackson 1978; Halliday and Adler

1986; S. A. Hetrick, pers. obs.). Another reason that the titmice may not have reacted differently

to the snake and control could be due to the activity level and movements of the snake. The

snake was stationary for the most part during the presentations and was not actively foraging,

probably due to the cold outside temperature. On the other hand, the hawk in this study did not

move because it was a stuffed mount and the titmice still responded as if it was a high-risk

situation, suggesting that movement was not an important factor in determining their response.

Potential Biases

The responses of the individual titmice to the 4 predator treatments and control treatment

were highly variable (as shown in Fig. 2-3). Some of this variation may be due to the existence

of social dominance hierarchies within wild titmouse flocks (Brawn and Samson 1983). In these

flocks, the alpha (most dominant) male possibly contributes the maj ority of the vocalizations in

response to predators. This idea is supported by evidence in Pale-winged trumpeters (Psophia

leucoptera), where the dominant male in the group gives the maj ority of the anti-predator

vocalizations (Seddon et al. 2002). Older, and most likely more dominant, male Willow tits










(Parus montanus) give alarm calls more frequently than females or young males (Alatalo and

Helle 1990). In the present experiment, all of the individuals were adults, but the sex could not

be determined in most of the cases.

The mean number of 'chick-a-dee' mobbing calls given by titmice in response to the

predators was lower than expected based on observations of natural encounters between titmice

and their predators (S. A. Hetrick, pers. obs.). A factor that could have contributed to this could

be that the titmice were housed alone and away from other titmice during the predator

presentations. A few of the hypotheses as to why animals give anti-predator calls include

alerting others and transmitting cultural information about predator characteristics (Klump and

Shalter 1984). For either of these to occur, the caller must have an audience. Evans et al. (1993)

found that domestic chickens (Gallus domesticus) rarely give aerial alarm calls unless they have

an appropriate audience. My results show that not all of the titmice gave anti-predator

vocalizations (seet or chick-a-dee mobbing calls) in response to the predator presentations. The

most common and numerous vocalization given was the single syllable chip note, which is not

typical of natural predator encounters (S. A. Hetrick, pers. obs.). Moreover, some individuals

sang during trials, but primarily in response to the cat, hawk and owl, which likely represent the

three highest-risk predators (Fig. 2-3). In the present experiment, the birds were held in a cage in

an outdoor aviary. There were sometimes other birds such as cardinals and towhees in the area

outside of the aviary that could have provided an audience for the titmice and wild titmice could

occasionally be heard in the far distance. Even so, the titmice may have felt a reduced

motivation to vocalize because they were not around their natural flock members.

Summary

In summary, my results clearly indicate that titmice do not produce predator-specific calls

in response to different predator species or classes of predators (avian, reptilian, and mammalian)










when the predators are presented in the same manner. However, results do indicate that titmice

may give risk-based calls in response to predators due to the fact that they varied their call

structure and note structure in response to the different predators. Given that this experiment

controlled for the situation under which predators were encountered, it is possible that the type of

predator affected the amount of risk encoded in the calls of titmice; house cat and hawk elicited

call structures typically associated with greater risk than either an owl or a snake. In chapter 3, I

investigate further the nature of the risk-based communication system of titmice in response to

predators.



































Figure 2-1. Outdoor testing aviary showing the test cage containing the test bird on the left and
the predator presentation cage on the right (with the sheet cover removed).
























0.2 0.4 0.6 0.8 1 0 1 2 1 4 1 6
Time (sec)


song notes











0.2 0.4 0.6 0.8 1 0 1 2 l e4 1 6
Time (sec)


seet notes











0.2 0.4 0.6 0.8 1 0 1 2 1 4 1 6
Time (sec)


chip notes


chick notes


D notes
I


20

S15


20

S15

10


20

S15



51


20

S15


0.2 0.4 0.6 0.8
Time (sec)


1.0 1.2 1.4 1.6


Figure 2-2. Examples of the variation within the main vocalizations of the Tufted titmouse in
response to the 4 predator treatments and control. A) 'Chick-a-dee' call complex
showing the variation in introductory chick notes and subsequent D notes. B)
Variations of song notes. C) Variations of seet notes. D) Variations of chip notes.





O
C

O


100
-o


Song notes

Sseet notes

Schick-a-dee calls

chip notes


Individual titmouse


V I Isong notes
10 seet notes


Schick-a-dee calls

0 I -,IIIIIICCIIIIIichip notes
1 3 5 7 9 11 13 15
2 4 6 8 10 12 14


Individual titmouse



Figure 2-3. Number of each type of note or call given in the first 5min following presentation by
the fifteen individual Tufted titmice in response A) to control, B) to snake, C) to owl,
D) to hawk, E) to cat treatments.


3 5 7 9 11 13 15
2 4 6 8 10 12 14

















O 20

O
C

1 00
C,


1 3 5 7 9 11 13 15
2 4 6 8 10 12 14


Individual titmouse


Juu


O 200-


O

O
100-





0
1




D.


Figure 2-3. (cont.)


3 5 7 9 11 14
2 4 6 8 10 13 15


Individual titmouse


Song notes

Sseet notes

Schick-a-dee calls

Ship notes


I song notes

Sseet notes

Schick-a-dee calls

chip notes
















O 20

O
C
O 0


# O song notes
100-
I I O seet notes

Schick-a-dee calls

0l ---- i ~_~~ chip notes
1 3 5 7 9 11 13 15
2 4 6 8 10 12 14


Individual titmouse
E


Figure 2-3. (cont.)





3-
O,


CZ








Mean # of chick
notes per call
a 0 Mean # of D notes
per call
control snake owl hawk cat

Treatment type




Figure 2-4. Number of chick and D notes per 'chick-a-dee' complex call given by Tufted titmice
in response to predator treatments and control in the 5min following presentation.
Both variables were significant with respect to treatment (ANOVA, p<0.001). All
pairwise comparisons were significant (LSD, p<0.05) except for snake and control
(p=0.647, p=0.442 for chick and D notes per call, respectively). Error bars: +/- 1 SE.


















Song notes
Sseet notes
Schick notes
D notes


control snake owl hawk


R


I


Treatment type
Figure 2-5. Mean number of overall song, seet, chick, and D notes that Tufted titmice gave in
response to the predator treatments and control in the first 5min following
presentation. None of the differences are significant with respect to treatment
(ANOVA, p>0.05).














.58-


.56 1

O



.50 1








.46
N= 15 21 34 88 141
control snake owl hawk cat


Treatment type


Figure 2-6. Mean entropy of the D notes of Tufted titmice in response to predator treatments and
control in the first 5min following presentation. All pairwise comparisons were
significant (LSD, p<0.05) except between control and owl (p=0.868), control and
snake (p=0.166), and snake and owl (p=0.061). Error bars: +/- 1 SE.










Table 2-1. Mean number of each note type given by individual Tufted titmice in response to
predator and control presentations in the first 5min following presentation.

Note type Predator N Mean SD SE
Chick Control 15 2.93 7.53 1.94
Snake 38 2.53 8.49 2.19
Owl 24 1.60 4.87 1.26
Hawk 21 1.50 3.78 1.01
Cat 7 0.47 0.99 0.26


D Control 7 0.73 2.15 0.56
Snake 21 1.40 4.17 1.08
Owl 34 2.27 3.41 0.88
Hawk 21 6.29 10.06 2.69
Cat 141 9.40 21.44 5.54


Seet Control 43 2.87 5.78 1.49
Snake 9 0.60 1.84 0.48
Owl 27 1.80 3.43 0.88
Hawk 43 3.79 7.60 2.03
Cat 54 3.60 6.44 1.66


Song Control 4 0.27 1.03 0.27
Snake 3 0.20 0.78 0.20
Owl 41 2.73 8.49 2.19
Hawk 36 2.57 9.06 2.42
Cat 50 3.33 12.91 3.33


Chip Control 942 60.80 52.43 13.54
Snake 1138 70.40 80.44 20.77
Owl 713 45.60 45.71 11.80
Hawk 809 51.36 58.32 15.59
Cat 448 25.67 35.72 9.22









CHAPTER 3
INTERSPECIFIC RISK-BASED CALL SYSTEM OF TUFTED TITMICE
(BAEOLOPHUS BICOLOR) INT RESPONSE TO PREDATORS

Many animals give alarm vocalizations in response to predators but little work has

focused on characterizing responses to such signals by sympatric heterospecifics. Tufted

titmice (Baeolophus bicolor) and other members of family Paridae respond to predators

that do not pose an immediate risk (e.g., perched predators) with complex mobbing calls

that have been described as 'chick-a-dee' calls. Black-capped chickadee (Poecile

atricapilla) mobbing calls vary with the degree of risk represented by different species of

perched predators, and conspecifies, isolated from predator stimuli, will respond to these

calls with risk-appropriate behaviors. Mobbing calls of the Tufted titmouse attract many

bird species and generate vigorous interspecific mobbing flocks that harass and scold

predators. I wanted to determine if titmice also vary their mobbing calls according to the

degree of risk that predators pose and if other species respond to mobbing and other anti-

predator calls of Tufted titmice with risk-appropriate behaviors. I presented captive

flocks of titmice with live high- and low-risk predators and controls under semi-natural

conditions to acquire vocal recordings. I then played these recordings and recordings of

titmouse 'seet' alarm calls (given when the bird is startled or in a state of fear) to captive

pairs of Carolina chickadees (Poecile carolinensis) without a predator stimulus. To the

high-risk predator presentations, titmice approached the predator more closely and gave

significantly more mobbing calls with different note composition and shorter note

intervals and note bandwidths compared to vocalizations given to low-risk predator

presentations and controls. In response to the playback of titmouse mobbing

vocalizations to the high-risk predator, chickadees approached the speakers more closely,

gave significantly more mobbing calls with different note composition and longer note









duration than calls given to the low-risk predator and control vocalizations of titmice and

they froze and became silent in response to titmouse seet calls. Thus, titmice did vary

their mobbing calls according to the degree of risk they experience, and chickadees

responded appropriately to the various titmouse mobbing calls and alarm calls.

Possession of an interspecific risk-based call system provides one explanation for the

socially dominant role that parid species play in interspecific associations (e.g., winter

foraging flocks and predator-mobbing aggregations) involving multiple bird species.

Interspecific risk-based calls like those characterized here may underlie ecological

facilitation in vertebrate communities more generally

Introduction and Background

Anti-Predator Vocal Signaling

Many vertebrate species respond to predator encounters by giving anti-predator vocal

signals. Passerine birds typically have two main types of anti-predator vocalizations: alarm calls

and mobbing calls. Typical passerine alarm calls are difficult to locate and are usually given

when the birds are in a state of fear, such as when a predator poses an immediate threat of attack

(Marler 1957; Ficken and Witkin 1977; Morton 1977; Apel 1985). Responses to alarm calls

usually involve either the cessation of movement or abrupt flight to cover by the caller and other

birds nearby (Marler 1955, 1957; Gompertz 1961; Ficken and Witkin 1977; Latimer 1977;

Ficken 1989; Evans et al. 1993). In contrast, mobbing vocalizations are easily localizable and

are usually given to perched predators posing little immediate risk, and these calls attract other

species that often harass the predator, sometimes encouraging it to leave the area (Klump and

Shalter 1984; Ficken and Popp 1996; Naguib et al. 1999; Baker and Becker 2002). The

production and description of anti-predator vocalizations has commonly been presented as part

of broader analyses of species' modes of communication and overall vocal complexity. Specific









focus on the production and context of individual species' anti-predator calls has shown that such

calls can communicate information about predators, including the level or type of risk

(Struhsaker 1967; Seyfarth et al. 1980b; Macedonia 1990; Dasilva et al. 1994; Blumstein and

Armitage 1997a; Greene and Meagher 1998; Zuberbiihler 2001). However, understanding when

and how conspecifies and heterospecifies receive and respond to risks conveyed in anti-predator

signals is just now coming to light (Naguib et al. 1999; Baker and Becker 2002; Templeton et al.

2005).

Predator-Specific and Risk-Based Call Systems

Different species' vocalizations in response to predators vary in complexity. Some

vocalizations contain detailed information about the type of predator (predator-specific calls),

while some contain information about the immediacy of threat that the caller faces (risk-based

calls; Macedonia and Evans 1993; Greene and Meagher 1998). These two calling systems differ

with respect to their production specifieity. In production, both predator-specific and risk-based

calls are referred to as situationallyy specific' because the call structure in some way varies with

distinct situations. If the vocal response uniquely (or categorically) covaries with the stimulus

type, as in predator-specific calls where a different type of call is associated with different

predator species or classes, then there is a high degree of production specifieity (Blumstein and

Armitage 1997a). On the other hand, if the vocal response varies continuously, i.e., the same call

type is produced with graded frequency or intensity according to the degree of risk, then there is

not a high degree of production specifieity. However, such a system is still considered to be

situationally specific because higher and lower risk situations can be distinguished (Blumstein

and Armitage 1997a,b; Blumstein 1999a). Risk-based (also called urgency-based) call systems

have been found in several taxa including ground squirrels, marmots, scrubwrens, babblers and









chickadees (Robinson 1980, 1981; Sherman 1985; Ficken 1989; Blumstein 1995a,b; Blumstein

and Arnold 1995; Naguib et al. 1999; Baker and Becker 2002; Leavesley and Magrath 2005).

Perception specifieity refers to the nature of the signal-receivers' reaction to the immediacy

of threat conveyed in anti-predator vocalizations. If a vocal signal produced under different

situations elicits contextually appropriate responses from conspecifies and/or heterospecifies, in

the absence of other cues, the vocal signals are said to generate 'perception specifieity' (Evans et

al. 1993; Macedonia and Evans 1993; Blumstein 1999a). Intraspecific perception specifieity is

common in vertebrates but work addressing interspecific perception specifieity of signals is rare

relative to the number of systems with sympatric heterospecifics that associate with one another

(Fichtel and Kappeler 2002).

Species in the family Paridae are known to produce risk-based calls. Ficken (1989)

demonstrated that Mexican chickadees (Poecile gamnbeli) vary their alarm or 'high zee' calls

according to the degree of risk and Baker and Becker (2002) showed that Black-capped

chickadees (Poecile atricapilla) vary their mobbing calls according to the immediacy of threat.

Neither study conducted playbacks to address intra or interspecific communication to see if

others were able to recognize the variation in the calls and respond appropriately. This is an

important distinction because birds may produce situationally specific calls, but unless others are

able to recognize them and understand their meanings, successful communication does not occur.

Recently, Templeton et al. (2005) clearly demonstrated intraspecific communication among

Black-capped chickadees. They found that the chickadees gave risk-based calls to different

predators and that these calls elicited appropriate responses from conspecifics. In the present

study, I tested for the presence of interspecific communication of predation risk between two

species of parids that co-occur in the southeastern United States.









Study System

The Role of Tufted Titmice in Mixed-Species Foraging and Mobbing Flocks

Tufted titmice (Baeolophus bicolor) and (potentially) Carolina chickadees (Poecile

carolinensis) play the role of the 'nuclear', or focal species, around which mixed-species

foraging flocks form during the winter months. In these flocks, chickadees are socially

subordinate to titmice (Waite and Grubb 1988) and are frequently found outside of foraging

flocks (Contreras unpubl. data; Farley et al. in review). Other species (more than 12-15) that

associate regularly with titmouse flocks play the role of 'satellite' species (Gaddis 1983; Grubb

and Pravosudov 1994; Greenberg 2000). Regular satellite species in North-central Florida

include a diverse set of species: Black-and-white warblers (M~niotilta varia), Downy

woodpeckers (Picoides pubescens), Ruby-crowned kinglets (Regulus calendula), Blue-headed

and White-eyed vireos (Vireo solitarius, y. griseus) and Blue-gray gnatcatchers (Polioptila

caerulea; Farley et al. in review). Contreras (unpubl. data) has shown that heterospecific flock

members follow titmice around the flock territory, providing direct evidence that titmice are

likely to play an active leadership role as the dominant nuclear species in this system. Nuclear

species of mixed-species bird flocks are generally characterized by behavioral traits that lend

themselves to interspecific communication; including interspecific dominance, a high level of

vigilance, and intraspecific sociality (Munn and Terborgh 1979; Hutto 1994). While not often

tested effectively, nuclear species in bird flocks are thought to facilitate flock formation and

sometimes food-finding, initiate and guide flock movements, and reduce predation risk for

satellite species (Moinkkoinen et al. 1996; Dolby and Grubb 2000). In addition, parid nuclear

species act as sentinels by giving anti-predator vocalizations in response to predators, thereby

alerting other flock members to danger (Gaddis 1983; Dolby and Grubb 1998).









Tufted titmice serve as nuclear species in mobbing aggregations of birds as well as in

foraging flocks. Mobbing aggregations, where one or more bird species gather around and

harass a predator, are relatively common in North-central Florida. These aggregations, that can

include more than 20 or 30 species of forest birds (Sieving et al. 2004), are formed when a

predator is spotted that is not an immediate mortality threat (e.g., a perched predator). Titmice

are the most vigilant and aggressive species in mobbing aggregations and these aggregations

appear to form around them (Greenberg 2000; S. A. Hetrick, pers. obs.). In North-central

Florida, up to half of the forest bird community responds to titmouse mobbing calls (more than

to other common local species' mobbing or alarm calls in this system) by approaching the

sounds and engaging in mobbing behavior (Sieving et al. 2004). Thus, it is likely that a complex

interspecific communication system exists between the Tufted titmouse and sympatric

heterospecifics involving both the production of situationally specific anti-predator calls and

contextually appropriate responses to these calls by others (Morse 1973; Sullivan 1984;

Zimmerman and Curio 1988).

Anti-Predator Calls of the Tufted Titmouse

The Paridae have two main anti-predator vocalizations in their repertoire: the 'seet' alarm

call (Fig. 3-la; also known as the high zee, high see, 'aerial' predator call) and the mobbing or

scold call (known variously as, churring, seejert, chick-a-dee; Smith 1972; Ficken and Witkin

1977; Gaddis 1979). Mobbing calls are variants of the 'chick-a-dee' call, which is a complex

call composed of combinations of introductory 'chick' notes and subsequent 'D' notes (dee

notes, churr notes), with the number and presence of each note type being variable (Fig. 3-1b, c;

Latimer 1977; Hailman 1989). The 'chick-a-dee' call complex (or portions of it) is produced in

many non-predator situations in addition to being the dominant mobbing vocalization (Latimer

1977; Hailman 1989; Grubb and Pravosudov 1994). Which call is given depends on the









situation; for example, an alarm call (seet) will typically be given to a raptor in flight that poses

an immediate threat of attack, while mobbing calls are typically given to perched raptors

representing much less risk of attack (Ficken 1989, S. A. Hetrick, pers. obs.). Mobbing calls are

easily localizable and are given when other birds are attracted to harass the perched predator,

whereas alarm calls are difficult to locate and result in the cessation of all vocalizations and

movement (freezing) by the caller and nearby birds or in rapid escape to cover (Ficken 1989;

Baker and Becker 2002; Howell 2006).

While it seems clear that interspecific risk-based signaling is likely to be quite common in

vertebrate communities, based on the diverse studies showing perception specificity among

heterospecific receivers of anti-predator calls (Nuechterlein 1981; Sullivan 1984; Seyfarth and

Cheney 1990; Hurd 1996; Shriner 1998; Windfelder 2001), the degree to which most of these

signaling systems are risk-based and/or predator-specific is unknown.

Research Design

I conducted two experiments to address whether Tufted titmice possess an interspecific

risk-based call system in response to predators. Experiment 1 involved presenting high- and

low-risk predators and controls to titmouse flocks to address the situational specificity of the

titmouse' s vocal responses. Experiment 2 was a playback study that addressed the perception

specificity of Carolina chickadee pairs to titmouse anti-predator vocalizations in response to the

high- and low-risk predators (obtained in Exp. 1) and titmouse seet vocalizations.

Hypotheses

I propose that Tufted titmice have an interspecific risk-based call system in response to

predators. This type of call system is composed of 2 parts, which represent 2 distinct hypotheses

and I conducted a separate experiment for each hypothesis. The first experiment tested the

situational specificity hypothesrr'i\ that titmice will vary the structure of their mobbing calls









according to the situation (Blumstein 1999a; Blumstein and Armitage 1997a,b). One way to do

this is to vary the structure of the calls according to the degree of risk that they are presented

with. Titmice can vary their mobbing calls in many ways. They can vary the number of mobbing

calls given, and within a mobbing call they can vary the type of notes given (introductory chick

notes or D notes), how many times a note is given, the temporal parameters of the notes and calls

(e.g., note duration, interval between notes and calls) and the acoustic structure of notes (e.g.,

bandwidth, entropy). Titmice may use all or some of these ways to vary their mobbing calls with

respect to threat level. Several researchers have found that parids respond to higher threat levels

by increasing their call rate and increasing the number of D notes within the calls (Apel 1985;

Baker and Becker 2002; Templeton et al. 2005). Some parids decrease or drop the introductory

notes of the mobbing call while extending the D note section as the level of risk increases

(Gompertz 1961; Latimer 1977). In addition, Templeton et al. (2005) found that Black-capped

chickadees (Poecile atricapilla) vary certain temporal measures and the acoustic structure of

certain notes when presented with predators of different risk levels. I tested for all of these

possible variations across different risk situations.

I also conducted an analysis of the behavioral responses of the titmice to the predators and

controls as a standard for determining whether the chickadee responses in the next experiment

were 'appropriate'. Parids are known to mob perched predators, which do not pose an

immediate predation risk (Langham et al. in press, S. A. Hetrick, pers. obs.). To characterize a

mobbing response in Black-capped chickadees, Templeton et al. (2005) observed the number of

chickadees that came within certain distances of the stimulus and the closest distance that any

bird approached the stimulus. Since one of the accepted hypotheses for mobbing is to drive the

predator out of the territory (Shedd 1982; Klump and Shalter 1984), it would make sense that










birds would have a greater motivation for mobbing higher-risk predators more intensely. One

way to mob more intensely would be for more birds to approach the predator and to approach

closer.

In the second experiment, I tested the interspecific perception specificity hypothesis that

titmouse anti-predator calls should elicit appropriate responses from heterospecific flock

members who hear the calls in the absence of the original stimulus (Macedonia and Evans 1993).

For chickadees to give appropriate responses, I would expect them to respond to the playback of

titmouse vocalizations to predator stimuli in the same way they would if the stimuli were

present- in this case, indicated by the responses of titmice to the actual predator situations in

Exp. 1.

Predictions

Experiment 1: Titmice produce risk-based mobbing calls that are situationally specific.

I predict that Tufted titmice will produce situationally specific mobbing calls that vary

according to the degree of risk to which they are exposed. Specifically, I predict that titmice will

vary their mobbing calls according to one or more of the parameters listed in Table 1.

To the Eastern screech-owl presentation, which represents a high-risk predator situation, I

predict that the titmice will respond with greater mobbing intensity than to the Great horned owl

presentation, which represents a low-risk predator situation. To respond with greater mobbing

intensity to the Eastern screech-owl, I predict that the titmice will increase their mobbing call

rate, approach the owl closer, and a greater proportion of titmice will come within Im and 3m of

the owl. I also predict that titmice will change their note composition by decreasing the number

of chick notes and increasing the number of D notes as the risk level increases (Latimer 1977).

In addition to call rate and note composition, variation in mobbing calls according to degree of









risk will be identified using the parameters listed in Table 1 (Apel 1985; Baker and Becker

2002).

Experiment 2: Chickadees exhibit interspecific perception specificity to titmouse anti-
predator calls.

In addition, I predict that Carolina chickadees will respond to the titmouse anti-predator

calls with some degree of perception specificity. I predict that the behavioral and vocal

responses of chickadees will vary in response to playbacks of titmouse seet calls and calls for

high- and low-risk predator situations. More specifically, in response to the playback oftitmouse

vocalizations given to a high-risk predator situation, I predict that the chickadees will respond

with more intense mobbing by approaching the speaker more closely and giving a relatively

larger number of 'chick-a-dee' calls (higher call rate) compared to the response to the playback

of titmouse vocalizations elicited from a low-risk predator situation. A similar response was

elicited from Black-capped chickadees when they were played conspecific mobbing calls in

response to different predators (Templeton et al. 2005). Mobbing calls, in general, attract birds

to the area of the caller to participate in mobbing (Hurd 1996; Baker and Becker 2002; S. A.

Hetrick, pers. obs.); therefore, playbacks of titmouse mobbing calls would likely attract

chickadees to the area of the speaker. I also predict that chickadees will alter their note

composition as described for the titmice in Exp.1, as well as possibly varying other parameters

listed in Table 1 in response to the different playback treatments. In response to titmouse seet

calls, I predict that the chickadees will not generate mobbing, but will dive to cover and freeze

while remaining silent without approaching the area of the speaker (Gaddis 1980; Ficken 1989).

Methods I: Situational Specificity Hypothesis

Five flocks consisting of 3 Tufted titmice were captured in Gainesville, Florida between 13

October 2005 and 5 January 2006. All 3 birds in each flock were captured at the same time from









the same location to ensure that the three birds knew each other. The birds were captured around

suburban seed feeders using mist-nets and/or baited walk-in potter traps and then all birds were

banded with uniquely colored leg bands. Immediately following capture the birds were

transferred to a 12 x 8 x 4m outdoor aviary containing numerous live trees and snags, providing a

semi-natural habitat at the USDA/APHIS/WS/NWRC Florida Field Station (United States

Department of Agriculture, Animal & Plant Health Inspection Service, Wildlife Services,

National Wildlife Research Center; USDA lab) in Gainesville, Florida (Fig. 3-2). The aviary

was constructed of V/2 inch plastic mesh attached to 4 x 4in. posts. After the 24 hr habituation

period, during which they were monitored for normal feeding activity and general health, flocks

were tested for each of 4 mornings in a row while being fed ad-libitum from a feeder in the

aviary. Birds were held for up to 7 days before being released back at their original capture site.

Predator Presentations

Each titmouse flock was presented with 4 treatments- a live Eastemn screech-owl (high-risk

predator, M~egascops a~sio), a live Great homed owl (low-risk predator, Bubo virginianus) and 2

controls- a procedural control with a live Northemn bobwhite quail (Colinus virginianus) and an

experimental control with no stimulus (an empty perch). The owls were non-releasable,

rehabilitated owls that were borrowed from Florida Wildlife Care, Inc. The 4 presentations were

made in randomized order for each flock and spaced approximately 24 hours apart. Most

hunting by both owls is nocturnal, but both owls occasionally hunt during the day (Packard 1954;

Spendlow 1979; Gehlbach 1994). Diet studies have shown that both the Eastemn screech-owl

(screech owl) and the Great horned owl (great horned) prey on birds, but small songbirds

comprise a much greater proportion of the diet of the small, maneuverable screech owl than the

larger, less maneuverable great horned (reviewed in Gehlbach 1995 and Houston et al. 1998;









Gehlbach 1994; Turner and Dimmick 1981), thus making the screech owl a higher risk predator

to the titmice (reviewed in Templeton et al. 2005).

I placed one of the stimuli in the aviary on a 1.2 m perch or platform (for the quail) under a

removable cover and for the experimental control, I placed an empty perch under the cover,

approximately 10min before the trial began and then retreated to a camouflaged blind just

outside the aviary. Care was taken to ensure that the titmice did not view the stimulus until it

was uncovered. The flock was given 5min or longer after the observer exited the aviary to

resume normal behavior. Audio and video recordings and behavioral observations were then

made for 5 minutes pre-stimulus and 7 minutes post-stimulus; however, only the first 2 minutes

post-stimulus were included in the analysis. A Sennheiser omni-directional microphone (1VE 62)

was used to record vocal responses of the titmice directly onto a laptop computer using Raven

Interactive Sound Analysis Software Version 1.1 with a sampling rate of 44100 at 16-bit

resolution. Behavioral responses within the first 2min post-stimulus were characterized using

the following behavioral variables adapted from Templeton et al. (2005): a) the closest distance

any bird approached the stimulus (in m); b) the proportion of birds that came within 3m of the

stimulus; c) the proportion of birds that came within Im of the stimulus; and d) whether the birds

were frozen in place during the entire treatment (Table 1). For the behavioral variables (a-c), I

used Kruskal-Wallis non-parametric test with one-tailed Mann-Whitney U (1VWU) post-hoc tests

to conduct pairwise comparisons among the treatments.

Spectrographic Analyses

I analyzed vocal responses in 13 of the 20 presentations (n=5 for the screech, n=4 for the

great horned, n=4 for the controls) because titmice did not vocally respond in 7 of the

presentations. Due to the low instance of vocal response to the procedural control (n=2) and

experimental control (n=2), the controls were lumped together in the analyses, resulting in n=4









control samples. Spectrographic analyses were performed on the vocal recordings using Avisoft

SASLabPro 4.39. To edit out noise, each sound file was FIR low-pass filtered at 12k
high-pass filtered at 1.8k
Size=75%, Window=Hamming, and Overlap=87.5%.

The notes in each call, or 'chick-a-dee' call complex, were visually classified as

introductory 'chick' notes or subsequent 'D' notes. The various introductory notes grade into

each other and are not reliably distinguished into natural categories; therefore, they were

classified together as 'chick' notes (Fig. 3-3). On the other hand, D notes can be reliably

classified (Bloomfield et al. 2005) due to their harmonic-like structure and little frequency

modulation. D notes also have a higher entropy and lower frequency than the introductory chick

notes, and D notes always occur at the end of the call, or are the only notes comprising a call,

making them easily distinguished from introductory notes (Fig. 3-3). Seet calls given by

responding titmice were omitted from analysis because the recording equipment could not pick

them up due to their extremely low amplitude. Single A or 'chip' notes were omitted in the

analysis due to their prevalence and predominantly low amplitude and 'peter' songs were

omitted in order to include only vocalizations used in the anti-predator context (Latimer 1977;

Gaddis 1979).

For the predator treatments, there was only one titmouse contributing to the maj ority of the

vocal mobbing in 4 out of the 5 flocks. This may be explained by the existence of social

dominance hierarchies within the flock, with the alpha male possibly contributing the maj ority of

the anti-predator vocalizations (Brawn and Samson 1983), as occurs in the closely related

Willow tit (Parus montanus; Hogstad 1993) and in Pale-winged trumpeters (Psophia leucoptera;

Seddon et al. 2002). In 3 of the 5 flocks, it was a known adult male that responded; in 1 flock, it









was a hatch year male; and in 1 of the flocks it was an adult of unknown sex (likely a male). Age

and sex were determined according to the molt limit and wing chord criteria established by Pyle

(1997). If more than one titmouse responded vocally to the treatment, the calls of the dominant

titmouse responding were isolated and measured.

I measured several aspects of the notes and calls in the 2min post-stimulus recordings,

including some that were based on measures used in previous studies of parids (Baker and

Becker 2002; Freeberg et al. 2003; Templeton et al. 2005). For each treatment, I averaged the

number of calls, the number of chick and D notes overall, the number of notes per call, the

number and proportion of chick and D notes per call, the duration of each chick and D note, the

call duration, the duration of the 1st D note per call, the interval between notes, the interval

between the chick and D section, and the interval between calls (Table 1).

Spectrum-Based Measures

I also measured several fine-scaled acoustic parameters on a sub-sample of 10 D notes that

were randomly chosen from each 2min post-stimulus recording. If there were fewer than 10 D

notes in the 2min post-stimulus recording, as was frequently the case for the controls, I chose as

many D notes as possible from the recording. In all cases, the D notes were chosen from high-

quality recordings. I measured the spectrum-based parameters of the D notes in the sub-sample

using a power spectrum with FFT length=512. The parameters were computed at the maximum

spectrum of the entire D note (maxpeakhold) and are similar to those used by Nowicki (1989)

and Templeton et al. (2005). The parameters were minimum and maximum frequency where the

amplitude goes last below -30dB and where the amplitude goes last below -10dB (min. and max.

frequency with the total option activated in Avisoft SASLabPro 4.39), bandwidth at -30dB and

-10dB (calculated with min and max frequency described previously), entropy, and the number

of peaks above -10dB.









For each acoustic factor that was measured, I used univariate analysis of variance

(ANOVA) with the least significant difference (LSD) post-hoc test to conduct pairwise

comparisons among the treatments. I transformed the data when appropriate to meet the

assumptions of the analysis using sqrt(n), arcsin(sqrt(n)), and log(n+1) transformations (Sokal

and Rohlf 1995). All statistical analyses were conducted using SPSS 11.5 for windows.

Because many of the variables were correlated with each other, I performed a Principal

Components Analysis (PCA) on all 14 general spectrographic measures and 3 behavioral

measures (excluding the measure of whether the birds froze in place). I used the uncorrelated

composite variables generated from the PCA in a Discriminant Function Analysis (DFA) to

determine if the discriminant functions could correctly classify the flock responses to one of the

three treatments. Significance in all statistical tests was set at the 0.05 alpha-level.

Results I

The results of the univariate ANOVAs showed that all of the behavioral variables and 13

out of 14 of the general spectrographic variables that were measured varied with treatment

(p<0.05). I determined which treatments differed from the others with the LSD pairwise

comparisons test. In some cases, the Eastern screech-owl treatment was significantly different

from the Great horned owl, which were both significantly different from the lumped (procedural

and experimental) controls (hereafter referred to as control). All transformations and the results

for all pairwise comparisons are shown in Appendix A (Table A-1, A-2). Means, SD and SE for

all measures taken are presented in Appendix B (Table B-1).

Titmice exhibited more intense mobbing behavior when presented with the screech owl

than with the great horned or controls. Titmice approached the screech owl more closely and

more of the titmice came within Im and 3m of the screech owl than of the great horned or

control (Kruskal-Wallis X 2=7.057, p=0.029; X22=8.670, p=0.013; X 2=7.355, p=0.025,










respectively, Fig. 3-4, 5), although the pairwise comparisons between the great homed and

control were not significant. None of the titmouse flocks froze in place during any of the

treatments.

The titmice gave different numbers of mobbing calls to the different treatments (ANOVA

F2, 20=12.2, p=0.001; Fig. 3-6), with the high-risk screech owl eliciting a higher call rate than the

low-risk great homed or control treatments (LSD p=0.006, p<0.001, respectively) and the great

horned eliciting a higher call rate than the control, although the difference was not significant

(LSD p=0.200). In particular, the high-risk screech owl elicited fewer chick notes per call (F2,

266=378.4, p<0.001; Fig. 3-7), more D notes per call (F2, 266=837.8, p<0.001; Fig. 3-7), and a

greater number of D notes overall (F2, 20=16.2, p<0.001; Fig. 3-8) than the great homed or

control treatments. The overall number of notes per call (F2, 266=3 59.9, p<0.001) and the

duration of the entire call (F2, 266=280.0, p<0.001) were greater and the interval between notes

(F2, 1313=7.7, p<0.001) was shorter for the screech owl treatment than the great horned or control

treatment.

Of the fine-scaled acoustic parameters that were measured, bandwidth at -30dB was

different between the D notes in each treatment, with the screech owl treatment eliciting titmice

to give D notes with a lower bandwidth than the great horned and control (F2, 105=10.7, p<0.001;

all pairwise comparisons were significant). The screech owl treatment also elicited D notes with

a higher minimum frequency (where amplitude goes last below -30dB) than the great horned or

control (F2, 105=13.1, p<0.001; all pairwise comparisons were significant except between the

screech and great homed treatment, p=0.065).

The results of the PCA on the 17 measures are summarized in Table 2. Four components

with eigenvalues >1 were extracted from the data set. The first principal component (PC1) was









determined mostly by behavioral variables, call rate, and note composition variables. The

second, third, and fourth principal components (PC2-4) were determined mostly by temporal

features of the notes and calls. After varimax rotations, the first four principal components

explained 85.3% of the variance, with PC1 accounting for 48.6% and PC2, 3 and 4 accounting

for 15.7%, 10.8%, and 10.2% of the variance, respectively. The 17 variables are depicted in a

bivariate plot that shows their respective values for PC 1 and PC2 (Fig. 3-9). The DFA using the

PCA factor scores led to 92.3% correct classification of the variables with the eliciting stimuli.

The first of the two discriminant functions (DFl1) accounted for 91.9% of the variation, and the

second (DF2) accounted for 8. 1 % of the variation. The resulting graph (Fig. 3-10) shows the

separation of flock responses into three basic groups with clearly separated group centroids

between the three treatments. The only misclassifieation was one of the flock' s response to the

great horned treatment classified as a response to the screech owl treatment.

Methods II: Interspecific Perception Specificity Hypothesis

Ten pairs of Carolina chickadees were captured in Gainesville, Florida between 10 January

2006 and 5 March 2006. The birds were captured, banded, and housed in the same manner as

the titmice in Exp. 1. In the study region, observations of normal group sizes at feeders indicated

that while titmice normally travel in groups of 3 or more, chickadees are nearly always in pairs.

Thus the number of birds per 'flock' was different for the two species but was determined to be

the most natural combinations of individuals likely to be related to, or at least familiar with, each

other (flock members were always captured within a few minutes of each other at the same

feeder in the same mist net or in adj acent or same potter traps).

Playback Presentations

Playback recordings (2min duration) were constructed from recordings of titmouse

mobbing calls, alarm (seet) calls and control vocalizations to make 5 independent replicates of









each of 4 playback treatment types. Two of the playback treatment types were of titmouse

vocalizations acquired in Exp. 1 in response to the screech owl and the great horned. The control

playback was constructed from pre-stimulus recordings of titmouse vocalizations (containing

mostly contact vocalizations) acquired in Exp. 1. For each playback treatment type, there were 5

unique exemplars (5 different titmouse flock responses, recorded in Exp. 1). The fourth playback

treatment type was of titmouse 'seet' calls acquired from other sources. Five unique seet call

recordings (2min each, 123 to 217 seet calls per min) were made using seet calls recorded from

one flock of free-living Tufted titmice responding to unknown stimuli (by Lang Elliot;

http:.//www.naturesound. com/) and from a previous experiment with captive titmice responding

to presentations of predators in close proximity (hawk, owl or cat; n=4 different birds, by S. A.

Hetrick; see Chapter 2). Each of the 20 playback recordings (5 variants of each of 4 treatment

types) was used twice for a total of 40 playbacks (n=40). Each of ten chickadee pairs received 4

of the playbacks (one each of the 4 treatments). The 4 treatments were presented in random

order for each pair and spaced approximately 24 hours apart.

At the beginning of each playback treatment, SAH placed a pair of camouflaged

RadioShack speakers (Model 40-1431) in the aviary on a 1.2m platform approximately 10min

before the trial began and then retreated to a camouflaged blind just outside the aviary. For each

treatment, the speakers were randomly placed in one of three locations in the aviary in order to

reduce habituation to a particular direction. The pair was given 5min or longer after the observer

entered the aviary to resume normal behavior. Recordings and behavioral observations were

then made for 5min pre-playback, during the 2min playback of the titmouse vocalizations and for

5 additional min post-playback. The methodology from Exp. 1 was followed to record and









video-tape the chickadees' behavioral and vocal responses. This methodology was also used to

analyze the behavioral variables associated with the chickadees' response.

Spectrographic Analyses

For each treatment, I analyzed 3min of behavioral responses and conducted spectrographic

analyses of 3min of recordings. Some chickadee pairs gave strong responses during the 2min

playback while some waited until the playback had finished to begin responding, so analysis of

3min total (2min during the playback and 1min post-playback) was deemed the most appropriate.

Chickadee behavioral responses and acoustic parameters were characterized as in Exp. 1. I did

not analyze fine-scale acoustic measures on the chickadee vocalizations because most of the

chickadee vocalizations overlapped with the titmouse vocalizations in recordings of the playback

trial. The acoustic data were analyzed using univariate ANOVA with the LSD post-hoc tests to

conduct pairwise comparisons among the treatments. The 'seet' playback was only considered

in the analysis of the behavioral responses and excluded from the spectrographic analyses of

vocalizations due to the low instance of calling during the playback treatment. Significance in

all statistical tests was set at the 0.05 alpha level.

Results II

The results of the univariate ANOVAs showed that all of the behavioral variables and 9

out of 10 of the general spectrographic variables that were measured varied with playback

treatment (p<0.05). The LSD pairwise comparisons showed that in some cases, the chickadee's

response to the screech mobbing playback (playback of the titmouse vocalizations elicited from a

screech owl) differed from the response to the great horned mobbing playback (playback of

titmouse vocalizations elicited from a great horned), which both differed from the response to the

control playback (playback of pre-stimulus titmouse vocalizations). In other cases, one or more

of the playback treatments did not differ from the other playback treatments. The details of the









transformations and the results for all pairwise comparisons are shown in Appendix A (Table A-

3, A-4). Means, SD and SE of all measures taken are presented in Appendix B (Table B-2).

Chickadees exhibited more intense mobbing behavior when they heard the screech

mobbing playback than when they heard the great horned mobbing and control playbacks.

Chickadees approached the speakers that broadcast the vocalizations more closely and both

members of the pair were more likely to come within Im and 3m of the camouflaged speaker in

response to the screech mobbing playback than to the great horned mobbing, control, or seet

playbacks (Kruskal-Wallis X 3=23.666, p<0.001; X 3=24. 118, p<0.001; X 3=16.687, p=0.001,

respectively; Fig. 3-11, 3-12), although some pairwise comparisons were not significant (see

Appendix A, Table A-3). The chickadee pairs did not freeze in place during any of the

playbacks except the seet playback in which they each froze for 100% of the 3min experimental

periods (and for long periods afterwards).

The screech mobbing playback elicited an overall greater number of notes per call

(ANOVA F2, 338=9.1, p<0.001) with a fewer number of chick notes per call (F2, 338=32.3,

p<0.001; Fig. 3-13) and a greater number of D notes per call (F2, 338=27.6, p<0.001; Fig. 3-13)

than the great horned mobbing or control playbacks. All of the pairwise comparisons were

significant for these three variables except for the number of chick notes per call for the great

horned mobbing playback and control playback (LSD p=0.108). The call duration and the

average duration of each D note were greater for the screech mobbing playback than the great

horned mobbing or control playbacks (F2,335=27.8, p<0.001; F2, 775=1199, p<0.001,

respectively).









Discussion

Situational Specificity of Titmouse Mobbing Calls with respect to Risk

The results of the predator presentation experiment confirmed that Tufted titmice have

situationally specific mobbing calls in that their calls varied according to the situation. These

calls are risk-based because the calls varied according to risk. More specifically, the results

confirmed my prediction that titmice would exhibit a stronger mobbing response to the higher

risk predator. Behavioral responses of titmice to the predator presentations showed that titmice

clearly distinguished high and low degrees of risk by exhibiting a more intense mobbing

response to the high-risk Eastern screech-owl. They approached it more closely (within Im) and

gave a greater number of 'chick-a-dee' mobbing calls than to the low-risk Great horned owl (Fig.

3-4, 3-6). These results are consistent with the results of other researchers who found that other

species in the family Paridae increase their mobbing call rate as the level of risk increases

(Latimer 1977; Apel 1985; Baker and Becker 2002; Templeton et al. 2005). Black-capped

chickadees (Poecile atricapilla) altered their rate of calling in response to a stuffed falcon at near

and far distances (Baker and Becker 2002) and in response to different predator species

(Templeton et al. 2005), by calling at higher rates in the higher risk situations. Many species of

rodent also increase their rate of calling as risk increases (Nikol'skii and Pereladova 1994;

Blumstein and Armitage 1997a; Randall and Rogovin 2002).

Titmice also varied their note composition in response to the different predators by

decreasing the amount of introductory chick notes per call and increasing the amount of D notes

per call as risk increased (Fig. 3-7). These Eindings agree with the observations of past authors

who noted that as the level of fear or risk increases, titmice drop the prefix (chick) notes and

increase the churr (D) notes (Odum 1942; Gompertz 1961; Latimer 1977). Variation in note

composition was previously thought to encode information related to many factors, including










information about predators (Hailman et al. 1985, 1987; Ficken et al. 1994). Black-capped

chickadees alter the number of introductory A and B notes in their 'chick-a-dee' mobbing calls

in response to predators presented at different distances (Baker and Becker 2002) and vary the

number of D notes in the calls in response to raptors of different sizes, with the smaller, higher

risk raptors eliciting the most D notes (Templeton et al. 2005). According to my findings and the

findings of the authors listed above, it appears that call rate and note composition, in particular,

are important for titmice and chickadees in communicating about risk.

I also found that titmice varied the note interval and the bandwidth of the D notes

according to the level of risk, which the chickadees could be cueing in on to help them interpret

the content of the titmouse calls. Other birds and mammals have also been documented to

change the structure of their call notes in different predator situations. White-browed scrubwrens

(Sericornis frontalis) vary the structure of their aerial trill call by increasing the minimum

frequency (pitch) of their calls according to the distance from a suddenly appearing predator

(Leavesley and Magrath 2005). Yellow-bellied marmots also vary several frequency

characteristics of their calls, including bandwidth, as a function of distance to certain predators

(Blumstein and Armitage 1997a). In both of these studies, the distance from the predator to the

subj ect represents the amount of risk, with the closer distances representing higher risk of being

caught or attacked by a predator; thus, we can conclude that both the scrubwren and the marmot

vary the structure of their calls according to the degree of risk that they perceive. As Templeton

et al. (2005) noted, variation in the structure of calls in response to different predators, reflecting

the degree of risk perceived, likely occurs in many species, but few researchers have tested for

this.









The results of the DFA indicate that the high- and low-risk predator treatments and control

treatment can be reliably (in 92.3% of the cases), and potentially uniquely, distinguished by the

combined behavioral and general spectrographic variables that were measured in this experiment

(Fig. 3-10). The combination of variables measured clearly discriminate between the titmouse

responses to the treatments, raising the possibility that the production of anti-predator signals

could be predator-specific. Davis (1991) made a similar argument that yellow-bellied marmots

had predator-specific calls based on results from a multivariate DFA. If the titmouse's response

were predator-specific, this would mean that while individual acoustic and behavioral parameters

may vary in a graded fashion across predator situations (e.g., more and fewer D notes), the

combination of multiple acoustic characteristics encoded in titmouse anti-predator signals could

(together) uniquely identify distinct predator species or classes. It is impossible to conclude this,

however, without further study of the nature of call production in Tufted titmice.

Some variation was evident in the vocal responses between the different titmouse flocks,

and this could be attributed to the predator' s movement and behavior at the time of presentation.

Many authors have observed that the behavior of a predator affects birds' reactions to it.

Increased call rates and more intense mobbing as a result of predator movement were reported in

several species of birds including Carolina wrens (Thil yesthesi us ludovicianus; Morton and Shalter

1977), Eurasian blackbirds (Turdus merula; Frankenberg 1981), and Pied flycatchers (Ficedula

hypoleuca; Shalter 1978; reviewed in Apel 1985). It has also been suggested that some

passerines can even detect differences in posture and behavior that are associated with how

hungry a predator is, and mob more frequently if the predator is hungry (Hamerstrom 1957). In

the present study, we did not detect that the two owls exhibited differences in posture or that

their movement differed at different times, but we did not examine this, since they were tethered









to the perch and were generally calm during experiments. I did observe that the owls were

sometimes looking in the direction of the titmice and sometimes looking in the opposite

direction. On one occasion, the great horned owl jumped down from the perch to the ground and

this seemed to evoke a momentary increase in titmouse call rate.

In general, it is well accepted that parids give alarm calls in response to predators that pose

an immediate threat, such as an aerial predator in a low attack flight, and that they give mobbing

calls to predators that do not pose an immediate threat, such as a perched predator (Gaddis 1980;

Ficken 1989). Having different calls for different classes of predators suggests that titmice may

be giving predator-specific calls, although this argument is weakened because there are many

exceptions as to when titmice give these calls (see below). Here, I show that mobbing calls

given by titmice to high- and low-risk (but all perched) predators and no-risk controls are also

clearly situationally specific. Whether they are categorically different (predator-specific) or

graded into one another (risk-based) may depend on how the calls are perceived by receivers.

For example, if a receiver is listening to only the number of D notes given per minute (Fig. 3-8)

then the level of risk being communicated will be graded across the range of numbers of notes

given and these can vary among individuals presented with the same stimulus. In these cases,

the communication would be risk-based and not predator-specific by standard definitions

(Macedonia and Evans 1993; Blumstein and Armitage 1997a; Blumstein 1999a). However, if a

receiver is basing its assessment of risk on more than one parameter characterizing the titmouse

mobbing calls, then 3 distinct (non-overlapping) classes of risk could be represented in the

titmouse vocal responses to the three types of stimuli (2 owls vs. control). In either case, titmice

are giving vocal signals that distinguish between high- and low-risk predators (both perched and

not of immediate threat) and no-risk controls (quail or empty perch). More work is needed to









determine whether single or multiple call characteristics generate the most 'appropriate'

responses in receivers.

Some bird species possess predator-induced calls that are functionally referential- that is

the calls are predator-specific with respect to predator species or class (e.g., high degree of

production specificity) and they produce appropriate responses in the individuals hearing them

(e.g., high degree of perception specificity; Klump and Shalter 1984; Evans et al. 1993;

Blumstein 1999a; Seddon et al. 2002). Within this framework, titmouse anti-predator calls

appear to be functionally referential in that they usually give seet alarm calls to aerial predators

and 'chick-a-dee' mobbing calls to perched and terrestrial predators (Gaddis 1980; Ficken 1989).

In purely experimental situations used here (Chapters 2 and 3), I have found that this is not

always the case, as perched predators and terrestrial predators presented to titmice in captivity

sometimes elicited seet calls. This would imply a lower degree of production specificity for the

seet call than is generally supposed. Seet call use may be more related to the level of fear or

surprise that an individual experiences (Marler 1957; Apel 1985) and in captivity, fear levels

could be generally higher than in the wild. 'Chick-a-dee' calls also appear to have a low degree

of production specificity because they are given in a variety of non-predator situations as well as

being the main mobbing vocalization (S. A. Hetrick, pers. obs.). Because one of the main

requirements for functional reference is high production specificity, there is weak evidence that

titmouse anti-predator calls are functionally referential because both calls appear to have low

production specificity.

Perception Specificity of Chickadees Exposed to Titmouse Anti-Predator Calls

The results of the playback experiment confirmed my prediction that chickadees would

respond with greater mobbing intensity to titmouse calls given in response to the high-risk

screech owl and with freezing and silence to the titmouse seet playback. In other words,









chickadees that heard the titmouse calls were able to interpret them and respond appropriately.

When the chickadees heard the screech mobbing playback, they responded (as the titmice did in

response to the screech owl) by approaching the stimulus more closely (within Im) than to the

great horned mobbing playback (Fig. 3-11). Black-capped chickadees respond in a similar way

to conspecific vocalizations elicited by high- and low-risk predators by altering their approach

distance to the stimulus (Templeton et al. 2005).

In the present study, chickadees responded to the screech mobbing playbacks by altering

the note composition of their 'chick-a-dee' mobbing calls. They decreased the amount of

introductory chick notes and increased the amount of D notes per call (Fig. 3-13). I was unable

to examine more subtle variations in call and note structure of the chickadee calls in this study

(see methods) but it is likely that these varied in other ways similar to those of Black-capped

chickadees and titmice (e.g., bandwidth, frequency; Templeton et al. 2005; Exp. 1). But by

documenting that the relative frequency of chick versus D notes per call varied by treatment, my

conclusions about risk-appropriate responses by chickadees are supported by the findings of

others (Apel 1985; Baker and Becker 2002; Templeton et al. 2005). And responses of the

chickadees clearly paralleled those of the titmice (Exp. 1), indicating that chickadees responded

to the titmouse calls, in the absence of other cues, in much the same way that the titmice

responded to the actual predators.

When the chickadees heard playbacks of titmouse seet alarm calls, they froze in place

every time and were almost totally silent for 5min or longer. This is similar to the observations

of Ficken and Witkin (1977) who found that Black-capped chickadees immediately became

motionless or moved to cover if in the open and froze upon hearing the alarm (e.g., high zee)

calls of another chickadee. Several other authors have observed that recipients stop moving and









become silent upon hearing these calls (Ficken et al. 1978; Gaddis 1980; Waite and Grubb 1987;

Ficken 1989). These responses make sense because seet calls are thought to be given when birds

are fearful and/or perceive imminent attack, as with the sudden emergence of any potential

predator including (especially) flying or low-cruising raptors (Latimer 1977; Ficken et al. 1978;

Apel 1985; Ficken 1989). Smith (1972) suggested that seet calls probably function to alert

recipients to danger. If the message that titmice are giving with their seet calls is extreme

danger, it seems appropriate that the chickadees would respond by freezing in place and

becoming silent. These behaviors would allow an individual to remain inconspicuous in the face

of potential immediate danger, for example, to lessen the risk of being detected by a cruising

predator that had not been located by the prey. In sum, responses to the situationally specific

anti-predator calls of the titmouse suggest a high degree of perception specifieity in chickadees.

Titmice Give Interspecific Risk-Based Mobbing Calls in Response to Predators

Although many researchers have demonstrated that anti-predator calls can communicate

specific predator information to conspecifies (intraspecific communication), few have shown that

these calls can also have meaning to heterospecifics interspecificc communication). Mixed

species groups of lemurs, (Eulemur fulvus rufus and Propithecus verreauxi verreauxi) have

evolved an interspecific functionally referential alarm system for diurnal raptors where both

species respond to the calls of conspecifics and heterospecifics in the group (Fichtel and

Kappeler 2002). Western grebes (Aechmophorus occidentalis) that nest in association with

Forster' s terns (Sterna forsteri) respond to the alarm calls of the terns by leaving their nests and

swimming to open water (Nuechterlein 1981). Other primate, rodent, and avian groups may

have interspecific anti-predator calling systems as well (Marler 1957; Francis et al. 1989; Shriner

1998; Windfelder 2001; Langham et al. in press). Compared to the large proportion of

vertebrates that join mixed species groups (reviewed in Greenberg 2000), and the pervasive









evidence that most animal groups form, at least in part, to gain anti-predator advantages (Caro

2005), a lack of interspecific communication related to predator avoidance may be more

surprising than its presence but more work is needed to survey for the prevalence of these

communication systems.

Most researchers that have found risk-based (or urgency-based) call systems in response to

predators have focused on typical alarm calls as opposed to mobbing calls (Robinson 1980;

Blumstein 1995a; Blumstein and Armitage 1997a; Leavesley and Magrath 2005). This

distinction is important because alarm calls are associated with a flight or freeze response while

mobbing calls are associated with approach and harassment of the predator. Many species'

mobbing calls, as well as their alarm calls, may also be able to communicate differences in the

risk environment.

It is likely that perception specifieity of titmouse risk-based mobbing calls occurs among

many other species, besides the closely related chickadees, that associate with titmice in foraging

and mobbing flocks and that simply share habitats with titmice. Downy woodpeckers, for

example, respond with risk-averse behaviors when they hear titmouse alarm calls (Sullivan

1984) and this makes sense because they spend a great deal of time with titmice in winter

foraging flocks (Farley et al. in review). More telling are the results of an experiment by Howell

(2006) in which the same playback recordings of titmouse calls that I used in Exp. 2 (for

chickadees) were presented to free-living Northern cardinals (Calrdinalis cardinalis) that were

feeding at platform feeders in the open. The cardinals responded with risk-appropriate behaviors

(e.g., freezing to seet calls, and diving for cover more often in response to mobbing calls to high-

risk predators than to mobbing calls to low-risk predators). Northern cardinals will join mobbing

flocks but not titmouse-led foraging flocks, so are not classified as a mixed-flock j oiner.









However, they are sympatric with titmice in many habitats throughout their common range and

so are exposed to titmouse anti-predator calls very frequently. Most surprising yet, is the

evidence found by Schmidt (unpubl. data) that squirrels feeding in trays in the open also respond

appropriately to titmouse alarm (seet) and mobbing calls. These results suggest that a wide range

of species that are and are not associated with, or taxonomically related to, titmice may perceive

titmouse anti-predator calls with a high degree of perception specifieity.

Finally, though I conclude based on findings presented here that both production and

perception of titmouse anti-predator calls are risk-based (graded, or continuous), the near-

complete separation in multivariate space achieved by the DFA between the titmouse vocal and

behavioral responses to control, great horned and screech owl presentations (Fig. 3-10) raises a

second possibility. If receiving species are able to discern multiple acoustic characteristics of

titmouse anti-predator signals (i.e., hear the calls in a multivariate fashion), they could be

discriminating among predator-risk situations according to class, as in a predator-specific

(categorical) call system rather than a risk-based (continuous) system of perception (Davis

1991). If so, then the anti-predator signaling system of titmice may actually be interspecifieally

functionally referential (Evans et al. 1993; Blumstein 1999a). However, it is impossible to draw

this conclusion without further study of the nature of call perception in species that exhibit

appropriate responses to titmouse anti-predator calls.

Potential Functions of Interspecific Risk-Based Mobbing Calls of Titmice

My study revealed characteristics of risk-based mobbing calls of titmice that could be used

by chickadees and other species to assess situation-specific risks, and clearly shows what risk-

appropriate behaviors are when mobbing and alarm calls are given. But why should titmice

possess a complex communication system with interspecific risk-based mobbing calls? Alarm

calls (for extreme danger) are produced in many species that are intraspecifically social, and









have been reasonably explained on the basis of kin selection (Hamilton 1963, 1964; Maynard

Smith 1965; Sherman 1977;Woolfendon and Fitzpatrick 1984). Researchers have formulated

many hypotheses as to why birds give mobbing calls to perched predators (Curio 1978, 1980;

Smith 1991). One hypothesis that seems to make sense for our particular system is the "move-

on" hypothesis where individuals seek to drive the predator from the area (Curio 1978).

Titmouse flocks hold stable winter territories and therefore would benefit by moving a predator

out of their territory to decrease the risk of future predation (Brawn and Samson 1983). By

giving interspecific risk-based mobbing calls, titmice can alert others to the specific situation and

induce them to j oin in the mobbing of the predator. Risk-based calls in response to higher risk

predators would likely generate a more intense mobbing response because titmice and the

receivers would have a greater motivation for driving these predators out of the area. A more

intense response would likely result in a greater chance of moving the predator out of the area,

which would benefit all flock members.

Most of the birds that participate in foraging and mobbing flocks with titmice share the

same predators. Shriner (1998) noted that when species have predators in common, they might

be able to obtain important information about predation risk from the anti-predator calls of the

other species, and so natural selection would drive the evolution ofinterspecific perception

specificity. It has been shown that more birds are attracted to titmouse mobbing calls than other

local forest species' mobbing calls (Sieving et al. 2004) and my work shows that this could

happen, in part, because the titmice are providing detailed information about the risk

environment in their calls that may help reduce the predation risk for other species that evolve to

use that information. The possession of such calls by titmice suggests a mechanism underlying

their socially dominant role in mixed-species foraging flocks. Tufted titmice give information-









laden vocalizations according to the risk environment and Carolina chickadees and potentially

other flock members are able to interpret and exploit this information, which may be an incentive

for them to join these mixed-species flocks (Gaddis 1980; Sullivan 1984; Howell 2006). Other

flock members may also benefit from the titmouse' high rates of vigilance, which allows them to

reduce their own vigilance and put more of their energy into foraging (Cimprich and Grubb

1994). Therefore, there are many reasons to associate with titmice, and the specific information

about predation risk that titmice provide, in addition to their high vigilance and aggressiveness

towards predators, may each play a role in decreasing the predation risk for other species that

associate, or merely live, with titmice.

Summary

In summary, this study clearly demonstrates that titmice possess an interspecific risk-based

call system with respect to predators. I have described many call characteristics that could be

used to communicate situational specifieity and I have shown that chickadees respond to

titmouse anti-predator calls with a high degree of perception specificity. Other work is showing

that the risk-based anti-predator calls of titmice are perceived by, and generate situationally

specific responses in, a wide range of unrelated species. If most of the species that share

predators with titmice participate in an interspecific risk-based communication system, then this

suggests a mechanism that could underlie interspecific facilitation via predation-risk reduction

among diverse sympatric vertebrate species. The majority of species in the family Paridae, that

are distributed throughout the Holarctic, exhibit highly conserved (similar) mobbing calls and a

high proportion of passerine species that live with titmice respond to titmouse mobbing calls by

exhibiting typical predator-mobbing behaviors (Langham et al. in press). This study (and

others) suggests that heterospecifics respond appropriately to parid anti-predator calls with a high

degree of perception specificity. Therefore, parid anti-predator vocalizations support a system of









interspecific communication about risk in which many species that share their predators may

participate. The existence of such a broad-based communication system suggests that

interspecific facilitation within bird communities of the Holarctic (where parids are distributed)

may be as common as other, more widely-studied, ecological interactions in organizing bird

communities (e.g., predation and competition).











kHz











A 0.2 0.4 0.6 0.8 1.0 1.2 s




chick notes D notes
kHz




20 ~ ynI







B 0.2 0.4 0.6 0.8 1.0 1.2 1.4 s




chick notes D notes

20

15

10





0.2 0.4 0.6 0.8 1.0 1.2 s


Figure 3-1. Examples of the major anti-predator vocalizations of the Paridae. A) Seet call of the
Tufted titmouse. B) 'Chick-a-dee' mobbing call of the Tufted titmouse with
introductory chick notes and subsequent D notes. C) 'Chick-a-dee' mobbing call of
the Carolina chickadee with introductory chick notes and subsequent D notes.


































































Figure 3-2. Outdoor aviary at the USDA/APHIS/WS/NWRC Florida Field Station in
Gainesville, Florida. Top picture: Platform with removable cover is in the center of
the picture with camouflaged blind in background. Bottom picture: Another view of
aviary with cover removed, revealing the Great horned owl on the perch.


IJ BI ~rt!
t~F~p~r 4 r --
.
II .---~'
C-:l
it; i. ..

















e- ~~ 7~ ~F~-P ~~ IEL-L


chick notes


D notes


0.2 0.4 060 0.8
Time (sec)


1.0 1 2 1 4 16.


Figure 3-3. Examples of the variation in the chick notes and the less variable D notes in the
'chick-a-dee' call complex of the Tufted titmouse. The chick notes grade into each
other and are not reliably distinguished into natural sub-categories.





g 3.0-





2.0-








1.0-



control great horned screech
Treatment Type





Figure 3-4. Closest approach distance of Tufted titmice to the stimuli during the predator and
control treatments in the first 2min following presentation. All pairwise comparisons
were significant (MWU, p<0.05) except between the great horned and control
treatments (p=0.594). Error bars: +/- 1 SE.























O 0.6-


o

0.4-



Prop. of titmice
0.2-- 1 Within 1m of
stimulus
Prop. of titmice
O within 3m of
stimulus

control great horned screech

Treatment type




Figure 3-5. The proportion of Tufted titmice that approached within Im and 3m of the stimuli
during the predator and control treatments in the first 2min following presentation.
For the proportion of titmice within Im and 3m of the stimuli, respectively, all
pairwise comparisons were significant (MWU, p<0.05) except between great horned
and control (p=0.953, p=0.953). Error bars: +/- 1 SE.





40-




0 30-




20-




1 O





control great horned screech

Treatment type




Figure 3-6. Number of 'chick-a-dee' complex calls given by Tufted titmice in response to the
predator and control treatments in the first 2min following presentation. All pairwise
comparisons were significant (LSD, p<0.05) except between the great horned and
control treatment (p=0.200). Error bars: +/- 1 SE.





III


Mean # of chick
notes per call
Mean # of D notes
O
per call


control


great ho0rned


screech


Treatment type




Figure 3-7. Number of chick and D notes per 'chick-a-dee' complex call given by Tufted titmice
in response to the predator and control treatments in the first 2min following
presentation. All pairwise comparisons were significant (LSD, p<0.05). Error bars:
+/- 1 SE.





250-



I 200
O















control great horned screech
Treatment type




Figure 3-8. Mean number of D notes given by Tufted titmice in response to the predator and
control treatments in the first 2min following presentation. All pairwise comparisons
were significant (LSD, p<0.05). Error bars: +/- 1 SE.














1.0- 14 12


15

0.5-

1 1
O0 16
O O11 O 1
cV 6
O0.0 _7 10 175 O
1 OO O 7 09

80 3 OO 04


-0.5-2




-1.0-

-1 .0 -0.5 0.0 0.5 1 .0

PC1

Figure 3-9. Plot of 17 behavioral and general spectrographic variables of Tufted titmouse calls
in two-dimensional space defined by two principal components. PC1 is determined
mostly by behavioral variables, call rate, and note composition variables, and PC2 is
mostly by temporal features of notes and calls. The variables corresponding to the
numbers in the plot are listed in Table 3-2.





1-I A O





OO

-1 Treatment

MC Group Centroids

-2 -1 I 2- screech

1- great horned

-3 O 0- control
-4 -3 -2 -1 0 1 2 3


DF1


Figure 3-10. Graph of the results of the Discriminant Function Analysis for Tufted titmice with
4 PCA factor score input variables generated from 17 original behavioral and acoustic
variables listed in Table 3-2.





6.0-



S5.0-







c. 3.0-



oj 2.0-




1.0-


control greathorned screech seet

Playback treatment



Figure 3-11. Closest approach distance of Carolina chickadees to the speakers during the
playback treatments of Tufted titmouse vocalizations in the first 3min after the start
of each playback. All pairwise comparisons were significant (MWU, p<0.05) except
between great horned and control (p=0.538). Error bars: +/- 1 SE.





0.6-






S0.. .4 -- -






0.2-

Prop.of chickadees
within 1m of speaker
Pro of chickadees
O
within 3m of speaker

control great horned screech seet

Playback treatment



Figure 3-12. The proportion of Carolina chickadees that approached within Im and 3m of the
speakers during the playback treatments of Tufted titmouse vocalizations in the first
3min after the start of each playback. Not all pairwise comparisons were significant
(see Appendix A, Table A-4). Error bars: +/- 1 SE.





O










Mean # of chick
notes per call
Mean # of D notes
O
per call
control great horned screech
Playback treatment


Figure 3-13. Number of chick and D notes per 'chick-a-dee' call given by Carolina chickadees
in response to the different playback treatments of Tufted titmouse vocalizations in
the first 3min after the start of each playback. All pairwise comparisons were
significant (LSD, p<0.05) except for the number of chick notes per call between the
great horned mobbing and control playback (p=0.108). Error bars: +/- 1 SE.










Table 3-1. Acoustic (first two columns) and behavioral parameters (3rd COlumn) used in
analyzing the response of Tufted titmouse flocks to high-risk and low-risk predator
presentations and control presentations.
General spectrographic Measures of acoustic
measures structure of D notes Behavioral measures


*call rate- number of calls



*number of chick notes
overall

*number of D notes overall



*number of notes per call



*number of chick notes per
call
*number of D notes per call

*proportion of chick notes
per call
*duration of each chick note

*duration of each D note

*call duration

duration of 1st D note of
each call
interval between notes

interval between chick and
D sections in each call
interval between calls


minimum frequency where
amplitude goes last below
-10dB
maximum frequency where
amplitude goes last below
-10dB
minimum frequency where
amplitude goes last below
-30dB
maximum frequency where
amplitude goes last below
-30dB
bandwidth at -10dB

bandwidth at -30dB

entropy

number of peaks above
-10dB


*closest distance any bird
approached the stimulus

*proportion of birds that
came within 3m of stimulus


*proportion of birds that
came within Im of
stimulus
*whether the birds were
frozen in place during the
entire treatment


* parameters also used in Experiment 2 with Carolina chickadee pairs










Table 3-2. Factor loadings of the 17 behavioral and general spectrographic parameters on the
four principal components after varimax rotation. Eigenvalues and amount of
variance explained by the respective components are given at the bottom of the table.
Parameter PC1 PC2 PC3 PC4
1 close approach -0.884 0.295 -0.142 0.230

2 prop. in Im 0.777 -0.457 0.150 -0.021
3 prop. in 3m 0.700 -0.253 -0.046 0.071
4 call rate 0.833 -0.178 0.177 0.151
5 number of chick notes overall -0.407 -0.055 0.274 0.616
6 number of D notes overall 0.957 0.023 -0.151 -0.047

7 number of notes per call 0.760 -0.152 -0.557 -0.014
8 number of chick notes per call -0.944 -0.145 0.000 0.186
9 number of D notes per call 0.880 -0.065 -0.395 -0.190

10 prop. of chick notes per call -0.932 -0.051 0.168 0.250
11 duration of each chick note 0.038 0.148 0.015 -0.861
12 duration of each D note 0.022 0.936 0.017 -0.227
13 call duration 0.850 0.184 -0.429 -0.099
14 duration of 1st D note of each call -0.203 0.935 0.157 -0.060
15 interval between notes 0.028 0.609 0.308 0.515
16 interval between chick and D sections -0.075 0.215 0.889 0.145
in each call
17 interval between calls -0.854 -0.042 0.275 0.193

Eigenvalue 8.26 2.67 1.83 1.73
% variance explained 48.6 15.7 10.8 10.2









CHAPTER 4
CONCLUSION

The Tufted titmouse is a vocally complex species that possesses a sophisticated

anti-predator call system (Gaddis 1979, 1980, this study). I conducted several

experiments to investigate characteristics of this system and summarize the results in

Figure 4-1. In the Eigure, I present titmouse vocal and behavioral responses to high- and

low-risk predators and heterospecific responses to playbacks of the titmouse

vocalizations (in the absence of the original predator stimulus). In response to high- and

low-risk perched predators, titmice exhibited mobbing behavior (Fig. 4-la) and produced

'chick-a-dee' mobbing calls with the high-risk Eastern screech-owl (screech owl)

eliciting more intense mobbing (closer approach to the predator and more mobbing calls)

than the low-risk Great horned owl (great horned) or control (see Fig. 3-4, 3-6). When

Carolina chickadees (a heterospecific associate of titmice in the wild) heard playbacks of

these calls in the absence of the predator stimuli, they responded in a similar manner as

the titmice by exhibiting mobbing behavior and producing 'chick-a-dee' mobbing calls.

Playbacks of titmouse mobbing calls produced in response to the high-risk screech owl

elicited more mobbing calls and closer approaches by the chickadees than playbacks of

titmouse calls given in response to the great horned (Fig. 4-la, second box; see Fig. 3-11;

Chapter 3). In response to aerial predators (Gaddis 1980) or when startled by the sudden

emergence of a potential predator (Chapter 2), which represent extremely high-risk

situations, titmice sought cover or froze in place and produced 'seet' alarm calls (Fig. 4-

lb). When chickadees heard playbacks of titmouse seet calls in the absence of the

original stimuli, they also responded by seeking cover or freezing in place and becoming

silent (Fig. 4-1b, second box; Chapter 3).










It appears that titmice give more mobbing calls and exhibit more intense mobbing

behavior as risk increases up to a point. But if the risk is too great, as is the case in an

aerial predator encounter, mobbing is no longer appropriate and titmice exhibit fearful

behavior (giving alarm calls, freezing in place and becoming inconspicuous). The fact

that titmice, in general, give mobbing calls and exhibit mobbing behavior in response to

high- and low-risk perched predators that do not pose an immediate threat and give alarm

calls and become still and inconspicuous in response to aerial predators that do pose an

immediate threat, can be explained by Morton's (1977) motivation- structural rules. The

rules state that low-frequency, broadband sounds (like titmouse mobbing calls) will be

produced when the caller is in an aggressive state and is likely to attack; whereas high-

frequency, pure tones (like titmouse alarm calls) will be produced when the caller is non-

aggressive or fearful.

In my first experiment, I found that titmice do not produce predator-specific

vocalizations that denote predator type or predator class (Chapter 2). They gave a

combination of different vocalizations in response to the control and predator treatments,

which included avian, mammalian and reptilian predators (see Fig. 2-3). Seet alarm calls

were occasionally given upon the removal of the cover from the presentation cage, and

were most often given when the cage contained a hawk or a cat, which were likely the

highest risk predators presented (explained in Chapter 2; see Fig. 2-5). The sudden

emergence of these predators in close proximity to individual titmice in this experiment

represents an extremely high-risk situation. It makes sense that titmice would produce

seet alarm calls to all the predators presented in the manner used, but more of them in

response to the predator species that represent the most risk. In this experiment the cat









and hawk likely represented the highest risk (see Chapter 2) and these were also the

predator treatments that elicited the highest mean number of seet calls. Additionally,

'chick-a-dee' calls were given by titmice in response to all the treatments, but were most

often given in response to the cat and hawk. According to Morton's (1977) motivation-

structural rules, these combined results likely indicate that titmice were most fearful

(more seet calls were produced) in response to the cat and hawk treatment at first, as the

seet calls were mostly elicited at the very beginning of the presentation. After their initial

response, the titmice were most aggressive (more mobbing calls were produced) in

response to the hawk and cat treatments. Overall, the results of this experiment clearly

show that titmice do not produce predator-specific vocalizations in response to predator

species or predator class (avian, mammalian, reptilian), but instead may be producing

risk-based anti-predator calls.

Therefore, in the second experiment, I tested for and found that titmice produce

situationally specific risk-based mobbing calls that vary according to the degree of risk

that a predator represents (Chapter 3). In addition to producing more 'chick-a-dee'

mobbing calls (see Fig. 3-6) and approaching the stimulus closer in response to the high-

risk screech owl (see Fig. 3-4), titmice also varied the note composition of their mobbing

calls (see Fig. 3-7) and varied several temporal and frequency characteristics of their

vocal response with respect to risk. Chickadees that heard playbacks of titmouse

mobbing calls in response to high- and low-risk predators responded in much the same

way that titmice responded to the actual predators (Fig. 4-la). In addition to producing

more mobbing calls and approaching closer (see Fig. 3-11) when they heard titmouse

calls in response to the high-risk screech owl, chickadees also varied the note










composition of the mobbing calls (see Fig. 3-13) and varied a few temporal note and call

characteristics as a function of playback type. These combined results indicate that

titmice possess an interspecific risk-based call system in response to predators (Fig. 4-

l a).

The 'chick-a-dee' mobbing call and the seet alarm call appear to be functionally

referential in terms of the type of predator encounter, as noted by Templeton et al. 2005,

because, in general, titmice give mobbing calls in response to perched and terrestrial

predators and seet calls in response to aerial predators (Gaddis 1980; Evans et al. 1993;

Blumstein 1999a). For Tufted titmice, the evidence for this claim is weak because

neither of these call types has a high degree of production specifieity (a requirement for

functional reference) because both calls are given in a variety of situations, including

some non-predator situations. For example, the seet call is often given in any situation in

which the bird experiences alarm and can be evoked by the sudden emergence of any

potential predator, either aerial or terrestrial (Ficken and Witkin 1977; Latimer 1977), as I

found in Chapter 2. In this study, titmice sometimes gave seet calls during the control

treatment in response to the removal of the cover from an empty cage (Chapter 2). Even

though no predator was present, the movement of the cover likely startled the titmouse

and resulted in the titmouse producing seet calls. Additionally, the 'chick-a-dee' call of

titmice is multifunctional and is given as a contact call, in coordinating group

movements, and is also the primary predator mobbing call (Gaddis 1979; S. A. Hetrick,

pers. obs.). Therefore, both call types have low production specifieity and there is weak

evidence for functional reference.









In summary, I present my four central conclusions. First, I documented that

titmouse anti-predator calls are not predator-specific with respect to predator species or

predator class (avian, mammalian, reptilian; Chapter 2), and this dispels some confusion

in the literature about the production of seet alarm calls and 'chick-a-dee' mobbing calls.

Seet calls have been described as 'hawk' or 'aerial predator' calls, and while they usually

are given in response to a flying hawk, they are also given in other situations. Therefore,

they are not reliably associated only with flying predators. Second, I documented that

titmouse 'chick-a-dee' mobbing calls should be classified as risk-based in their

production because the number and quality of the this call varied in a graded fashion as a

function of risk. The third conclusion is that the nature of risk encoded in titmouse calls

can be used by other species, as evidenced by appropriate behavioral responses by

heterospecifics upon hearing the calls. This study was the first to document interspecific

communication about such fine-scale differences in predation risk; i.e., that chickadees

discern differences between calls given in response to two different species of perched

owls. Finally, the last two conclusions taken together indicate that titmice possess an

interspecific risk-based call system- a more sophisticated interspecific communication

system with respect to predation risk than was previously known. Previous work showed

that many species respond to titmouse mobbing calls, but we have improved

understanding of these interactions by documenting that both the production and the

perception of the calls by other species are risk-based. Titmice give specific information

about different risk situations and heterospecific responses indicate that other species

tailor their responses according to the specific level of risk being communicated. Many

species that are sympatric with titmice respond to their anti-predator calls and may be









benefiting from the specific information that they contain. The positive benefits gained

by receivers of predation risk information suggest that titmice may be playing an

important facilitative role in animal communities.














Heterospecific
Chickaclee
Response


Eliciting Stimulus Stimulus


E wRisk

hc~ned owl I rh Cp low-riul-






Spernhed Ehstm high-ris); I-rk
sl achr-owl I








serialraptr or



rewdator


Titmo~use Behaviorala
Response







mobbirb4~


Titmo~use Call


less itt~rens
mobbing,mobing
calls- fewer





ittbense mobing,
mobbing calls- lots


-rt __ ~


Figure 4-1. The flow diagram summarizes the vocal and behavioral responses of Tufted titmice to predators representing varying
degrees of risk and heterospecific responses to playbacks of titmouse anti-predator vocalizations. The two shaded rows
demonstrate the production of and responses to A) titmouse 'chick-a-dee' mobbing calls, and B) titmouse seet alarm calls.


fleeze siant


see}; covr,
freez, silst









APPENDIX A
SUMMARY TABLES OF STATISTICAL TESTS FOR RESPONSES OF TUFTED TITMICE
TO PREDATORS AND RESPONSES OF CAROLINA CHICKADEES TO PLAYBACKS










Table A-1. Behavioral responses of Tufted titmouse flocks to high- and low-risk predators and controls in the 2min following
presentation. One-tailed Mann-Whitney U tests were used to generate pairwise comparisons. scr = Eastern screech-owl,
gh= Great horned owl, cont = control
Measure Kruskal- df Asymp. screech screech great Multiple Comparisons at p<0.05
Wallis Sig.* versus versus horned
X2 p-value great control versus
horned p-value control
value p -value
closest approach (m) 7.057 2 0.029 0.053 0.019 0.594 scr<(gh=cont)
prp in 1m 8.670 2 0.013 0.016 0.013 0.953 scr>(gh=cont)
prp in 3m 7.355 2 0.025 0.053 0.008 0.953 scr>(gh=cont)


*- Asymp. Sig.


Asymptotic significance














Measure Transformation ANOVA *Fdf Adjusted screech screech great Multiple
p-value rZ versus versus horned Comparisons at
great control versus p<0.05
horned p-value control
value value
call rate (#/2min) sqrt 0.001 12.22, t0 0.540 0.006 <0.001 0.200 scr>(gh=cont)
chick notes overall (#/2min) log(n+1) 0.384 1.02, t0 0.001 0.706 0.198 0.381 scr-gh=cont
D notes overall (#/2min) lo(n+1) <0.001 16.22, t0 0.616 0.045 <0.001 0.007 scr g>cont
notes percall (#/2min) <0.001 359.92, 266 0.726 <0.001 <0.001 <0.001 srh>cont
chick notes percall (#/2min) sqt<0.001 378.42 26 0.740 <0.001 <0.001 <0.001 scr g D notes per call (2min) sqrt <0.001 837.82, 266 0.863 <0.001 <0.001 <0.001 scr>gh>cont
po.of chick notes percall acisrt 0.001 16.32, 14 0.702 0.260 <0.001 0.002 scr g duration of each chick note (s) <0.001 17.823 131 0.205 0.001 0.034 <0.001 gh>scr>cont
duration of each D note (s) sqt<0.001 63.72 1492 0.078 <0.001 0.499 0.001 h>cont=scr)
call duration (s) <0.001 280.02 26 0.679 <0.001 <0.001 <0.001 srg>cont
duration of 1st D note in each <0.001 52.02, 298 0.256 <0.001 <0.001 0.199 scr<(gh=cont)
call (s)
interval between notes (s) srt <0.001 7.7 ,1313 0.010 0.011 0.001 0.049 scr interval between chick and D sqrt <0.001 13.82 sq 0.229 0.409 <0.001 <0.001 (scr=gh) sections in each call (s)
interval between calls (s) srt 0.000 597.02 293 0.803 <0.001 <0.001 <0.001 scr min. freq. where amplitude 0.044 3.22~ 105 0.041 0.778 0.025 0.019 (scr=gh)>cont
goes last below -10dB (Hz)
max. freq. where amplitude 0.021 4.03~ 105 0.054 0.552 0.019 0.007 (scr=gh)>cont
goes last below -10dB (Hz)
min. freq. where amplitude <0.001 13.12, 105 0.189 0.065 <0.001 0.001 (scr-gh)>cont
goes last below -30dB (Hz)
max. freq. where amplitude 0.734 0.32~ 1n, -0.013 0.530 0.821 0.480 scr-gh=cont
goslast below -30dB (Hz)
bandwidth at -10dB (Hz) 0.279 1.32. 105 0.006 0.668 0.195 0.119 scr-gh=cont
bandwidth at -30dB (Hz) <0.001 1. y 0. 157 0.047 <0.001 0.006 scr g

Table A-2. General spectrographic measures and measures of acoustic structure of D notes of Tufted titmouse calls elicited in


response to high- and low-risk predators and controls in the 2min following presentation. scr


Eastern screech-owl, gh


Great horned owl, cont


control










Table A-2 (cont)
Measure Transformation ANOVA *Fdf Adjusted screech screech great Multiple
p-value r2 VeTSus versus horned Comparisons at
great control versus p<0.05
horned p-value control
p-value p-value
entropyI 0.607 0.52, 105 -0.010 0.524 0.619 0.329 scr-gh=cont
peaks above -10dB (#/D note) 0.988 0.02, 105 0.019 0.915 0.944 0.879 scr-gh=cont
*- Fdf = Fdf treatment, df total











Table A-3. Behavioral responses of Carolina chickadee pairs to playbacks of Tufted titmouse vocalizations in response to high- and
low-risk predators and controls and titmouse seet alarm calls in the 3min following the start of each playback. One-tailed
Mann-Whitney U tests were used to generate pairwise comparisons. scr = Eastern screech-owl, gh = Great horned owl,
cont = control
Measure Kruskal- df Asymp. screech screech great screech great control Multiple
Wallis Sig.* versus versus horned versus horned versus Comparisons at
X2 p-value great control versus seet versus seet p<0.05
horned p-value control p-value seet p-value
value p -value p -value
closest approach (m) 23.666 3 <0.001 0.001 <0.001 0.579 <0.001 0.043 0.004 scr<(gh=cont) prop. in 1m 24.118 3 <0.001 0.005 <0.001 0.684 <0.001 0.280 0.481 scr>(gh=seet=cont)
prop. in 3m 16.687 3 0.001 0.011 0.105 0.436 <0.001 0.247 0.052 scr>gh
scr-cont
gh=cont
scr>seet
gh=seet
cont>seet


*-
Asymp. S
=
ig.


Asymptotic significance