<%BANNER%>

From Pinus elliottii Plantation to Pinus palustris Ecosystem

Permanent Link: http://ufdc.ufl.edu/UFE0021874/00001

Material Information

Title: From Pinus elliottii Plantation to Pinus palustris Ecosystem The Role of Herbicide in Longleaf Pine Restoration
Physical Description: 1 online resource (85 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: diversity, fire, flatwoods, herbicide, hexazinone, imazapyr, longleaf, palustris, pine, pinus, restoration, sulfometuron
Interdisciplinary Ecology -- Dissertations, Academic -- UF
Genre: Interdisciplinary Ecology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Under a natural fire regime, the longleaf pine (Pinus palustris Mill.) ecosystem is characterized by an open, longleaf pine-dominated canopy and a diverse, grass-dominated understory. In the years since European colonization, the vast majority of longleaf pine acreage has been cut over, fire-suppressed, and fragmented, which has resulted in extensive understory invasion by shrubs and a concurrent decline in herbaceous species diversity. One of the biggest challenges to successful restoration of this ecosystem is the persistence of shrubs in the understory, which suppress longleaf pine seedlings as well as native herbaceous plants. Herbicide can be used as a supplement to fire in order to enhance shrub control, but must be studied carefully because of the potential for negative impacts on native plants. Information is lacking about the effects of herbicide on natural longleaf pine flatwoods communities. We used a banded application of three herbicides and one tank mix as shrub control treatments following complete removal of the slash pine canopy and replanting with containerized longleaf pine seedlings in a mesic-wet flatwoods. The herbicides tested were Arsenal? (imazapyr), Oust? (sulfometuron methyl), Velpar L? (hexazinone), and an Oust? + Velpar L? tank mix. Four years after application, no negative impacts on understory species richness, diversity, evenness, or community composition were evident in any of the herbicide treatments. Oaks (Quercus spp.), one of the dominant shrub genera on the study site, were resistant to sulfometuron methyl, and this herbicide was therefore ineffective both as a pine release treatment and for enhancing herbaceous species growth. Imazapyr was the most effective treatment overall, significantly improving longleaf pine seedling growth and also enhancing herbaceous species cover. Both hexazinone and the hexazinone + sulfometuron methyl tank mix provided some seedling growth and understory enhancement as well, though these effects were not as pronounced as those in the imazapyr treatment. Shrubs resprouted aggressively after a dormant-season prescribed fire in the fifth year after treatment, indicating that herbicide-related increases in cover of wiregrass and other herbaceous species may be lost in future fire cycles. Retention of overstory pines for shrub competition and as a source of fine fuels is recommended in addition to herbicide for sustainable silviculture in longleaf pine flatwoods.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Jose, Shibu.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021874:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021874/00001

Material Information

Title: From Pinus elliottii Plantation to Pinus palustris Ecosystem The Role of Herbicide in Longleaf Pine Restoration
Physical Description: 1 online resource (85 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: diversity, fire, flatwoods, herbicide, hexazinone, imazapyr, longleaf, palustris, pine, pinus, restoration, sulfometuron
Interdisciplinary Ecology -- Dissertations, Academic -- UF
Genre: Interdisciplinary Ecology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Under a natural fire regime, the longleaf pine (Pinus palustris Mill.) ecosystem is characterized by an open, longleaf pine-dominated canopy and a diverse, grass-dominated understory. In the years since European colonization, the vast majority of longleaf pine acreage has been cut over, fire-suppressed, and fragmented, which has resulted in extensive understory invasion by shrubs and a concurrent decline in herbaceous species diversity. One of the biggest challenges to successful restoration of this ecosystem is the persistence of shrubs in the understory, which suppress longleaf pine seedlings as well as native herbaceous plants. Herbicide can be used as a supplement to fire in order to enhance shrub control, but must be studied carefully because of the potential for negative impacts on native plants. Information is lacking about the effects of herbicide on natural longleaf pine flatwoods communities. We used a banded application of three herbicides and one tank mix as shrub control treatments following complete removal of the slash pine canopy and replanting with containerized longleaf pine seedlings in a mesic-wet flatwoods. The herbicides tested were Arsenal? (imazapyr), Oust? (sulfometuron methyl), Velpar L? (hexazinone), and an Oust? + Velpar L? tank mix. Four years after application, no negative impacts on understory species richness, diversity, evenness, or community composition were evident in any of the herbicide treatments. Oaks (Quercus spp.), one of the dominant shrub genera on the study site, were resistant to sulfometuron methyl, and this herbicide was therefore ineffective both as a pine release treatment and for enhancing herbaceous species growth. Imazapyr was the most effective treatment overall, significantly improving longleaf pine seedling growth and also enhancing herbaceous species cover. Both hexazinone and the hexazinone + sulfometuron methyl tank mix provided some seedling growth and understory enhancement as well, though these effects were not as pronounced as those in the imazapyr treatment. Shrubs resprouted aggressively after a dormant-season prescribed fire in the fifth year after treatment, indicating that herbicide-related increases in cover of wiregrass and other herbaceous species may be lost in future fire cycles. Retention of overstory pines for shrub competition and as a source of fine fuels is recommended in addition to herbicide for sustainable silviculture in longleaf pine flatwoods.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Jose, Shibu.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021874:00001


This item has the following downloads:


Full Text
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 E20101110_AAAAKC INGEST_TIME 2010-11-11T04:01:27Z PACKAGE UFE0021874_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 3334 DFID F20101110_AACJED ORIGIN DEPOSITOR PATH freeman_j_Page_39thm.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
153a9300cde6b8618a0ad48ba9f95171
SHA-1
a5c4f949caca2618b70425e0365caf21ea2bcb5a
9697 F20101110_AACIXG freeman_j_Page_68.pro
816a388fa250ce2750dee53240c38177
6031bd0f52843fd4026dc9b53d0a199e11dadc89
25271604 F20101110_AACISJ freeman_j_Page_21.tif
9448a954db896ab7734c4c564d40c7de
e7f00d81281189ab3ed2a29ef3b22a7e4dd0d5f7
90196 F20101110_AACINL freeman_j_Page_60.jpg
e7e5458511dc4e8e3e8cebc8fbd45b68
74c3f938f4d7c8dd693ae35239f47cbe34d5a865
5483 F20101110_AACJEE freeman_j_Page_40.QC.jpg
fb8ada6192bcbb9dc4b49ccf57107389
4597ac96020d1f31265092b67f439f4907f1d979
10873 F20101110_AACIXH freeman_j_Page_69.pro
6df83a0f30d3fff0ed80b15ff253e1d3
5f513bc48b92f3182689e96c4eaa951639927088
F20101110_AACISK freeman_j_Page_22.tif
92393ad0d64cdb3b9969fb09c0e23bca
9eb3aeb5b3df08cd8a9f55ed8214bfcf4572ded9
82047 F20101110_AACINM freeman_j_Page_61.jpg
d652cb40d57a3bd6b633d129a29c3d20
f441330575bb01d8ddbdeb601a349a959fce2185
1416 F20101110_AACJEF freeman_j_Page_40thm.jpg
efae8072d5958d3b53a5178e1e797658
e7843b6ab69ed0f76cbd9445cde9305b86a9c6ed
9032 F20101110_AACIXI freeman_j_Page_70.pro
8db9b9e443a3cf29ce34e30950382ddb
906cb5512ca67b851bc10a5f31b68f2cd646eae4
F20101110_AACISL freeman_j_Page_23.tif
8ced35d721c94b1f51000a3876b80424
83ad6a2c42db65f938ab7b5a0fd7ad9a4bce2653
84988 F20101110_AACINN freeman_j_Page_62.jpg
179902452285658b5d486ef7430ffd6e
b122e7c766535d1d4f9e080402b82bb1e8f56066
6729 F20101110_AACJEG freeman_j_Page_41.QC.jpg
dd87a6076fed9994eddc05217cea50cf
8b7b47664e74e847b39cba1923e6982710087baa
F20101110_AACISM freeman_j_Page_24.tif
3768eb47eae82fb0923cc24835199ee0
49a01e81b2d05ae8106bd0ed6b46f7eeb68cc4b6
91314 F20101110_AACINO freeman_j_Page_63.jpg
dc80aebee2d6c1412655ea026bcc9c09
86954dbd03783d752a5b055d5becd110bb12581c
11652 F20101110_AACIXJ freeman_j_Page_71.pro
2620ec089212aa9ff563815c0a2f7310
517edc3cb412cd71eebfebb7fb7fcdc1651e2dc7
2044 F20101110_AACJEH freeman_j_Page_41thm.jpg
e6d123852d8c1160b8fa3c99d1199189
5830ecde4837f1c68b89b4105c5e077c4cac4e9a
F20101110_AACISN freeman_j_Page_25.tif
80126476285b2fdb6fafb4e1e6071a7d
b6d337352b0a88952cab603f7ec711a2117b5671
91185 F20101110_AACINP freeman_j_Page_64.jpg
69d6f03ffa622140dcf14366ffb5e196
04ae1cc6d2799bda601d472810f0ee502cfe6923
53721 F20101110_AACIXK freeman_j_Page_72.pro
8c937a6b76376d0730a6576d7eeafef5
fa063cef387714df6c9586dff396fe3b7fa06480
10654 F20101110_AACJEI freeman_j_Page_42.QC.jpg
5b20bcc7e2a268054a52992fc6997bb6
afdc478e39ce7307da6a7fca442386a02690c9bc
F20101110_AACISO freeman_j_Page_26.tif
95090031f291886417d66f7c5635be96
30766a9beb3fe4144439bee5a09f0b9a1d624d6f
82212 F20101110_AACINQ freeman_j_Page_65.jpg
8e1720305c9f2886741eabe4f3c7bc65
20c78038ff63dd553bb3bae44f2a879d941848ad
54550 F20101110_AACIXL freeman_j_Page_73.pro
9a28af56efccbe5f320b8882ee6d011c
03c0f5cab1977a2fc4fbac8c52548ca109cf5262
3247 F20101110_AACJEJ freeman_j_Page_42thm.jpg
d1dee4ecff03aabb3c9b17cf0b012139
685987657e3737eea9fdcfb34ed1d8baf2cc6af3
38155 F20101110_AACINR freeman_j_Page_66.jpg
9aaac26621ce0844bd8c2c82adf6a97d
434b50f8f1dd0a9f759be198badd868907326e6c
53984 F20101110_AACIXM freeman_j_Page_74.pro
fd314b75115755de3c96438e4dd6eab3
f1a1382991f5a098f78ba5c6cacc18b7dd9110c9
9028 F20101110_AACJEK freeman_j_Page_43.QC.jpg
b6c08959dca87a49c3a79ec35fd080a2
aca7fbd5264c70d3ae86430508c1aba653ed339c
F20101110_AACISP freeman_j_Page_27.tif
18fc0e70f5c395dea891e8b61858aa0a
cda29f07a33bb38755374e09e42165c82240f8ef
20265 F20101110_AACINS freeman_j_Page_67.jpg
aaa6579325f00b1b0d3f3a5f2fac17ef
38d12ece257aa9908e80064b90d361597ca8d21a
3604 F20101110_AACIXN freeman_j_Page_75.pro
841d3cca67e89dee81f92806c3b33ad0
b627582545049c34a52475ccdbd495035287974c
3004 F20101110_AACJEL freeman_j_Page_43thm.jpg
da50d3981f8a4e7c8c1423d688aa3e8f
ef1a804744b318d03ff2c7e78c63539307a52e05
F20101110_AACISQ freeman_j_Page_28.tif
fb3f950cbc9e33f8229f3b623d17682a
280de50d24dae212c7fdcb08c0cd7dc91893f227
20079 F20101110_AACINT freeman_j_Page_69.jpg
66b1abcdc70e6adf475e792df0afc8a6
d5f1a68a518a12e51b929187aba6167149dc2a9e
75769 F20101110_AACIXO freeman_j_Page_76.pro
79bdca7874147b90c0c9e33842d31c62
332c6ed314306820eae32676bdd80b5805ce7648
8202 F20101110_AACJEM freeman_j_Page_44.QC.jpg
076365fbe914fe1c74acb860c0c1bfa7
3caa3c017700b60198a4d9aaa038cc90cda989f9
F20101110_AACISR freeman_j_Page_29.tif
cf548defc9221bdc8ff658b9e0951465
c559add771d9711d079092cf9834028683b37b30
17196 F20101110_AACINU freeman_j_Page_70.jpg
adeec1e334c9bdb55762a24ffd38a2bc
605cea7125e251d0f956de4630143535d8dea398
76037 F20101110_AACIXP freeman_j_Page_77.pro
6b01659724e3a47d030728cd6c6a1636
d74895b63662b79aeae53b60389f2d507d8f9550
2408 F20101110_AACJEN freeman_j_Page_44thm.jpg
cff9aaec462f56603ec80334af35c2f4
c86cbc7ea4c11858c20e30dac010031cbd2290ee
F20101110_AACISS freeman_j_Page_30.tif
559443df85cda0cfa010bb6257ea6171
5a450da343372af1ab6c57227aa0b186468c15a3
29067 F20101110_AACINV freeman_j_Page_71.jpg
ab6c7744898050545b017409c230de73
615f31d0ec94325d398a26e39110ca55a1d1fa04
63942 F20101110_AACIXQ freeman_j_Page_78.pro
3c882560cde05a1afae0ec3ad0b3a530
b9f4549075e0ae8614116ced74a06776d8bdd678
8692 F20101110_AACJEO freeman_j_Page_45.QC.jpg
3a7079725c8b3303a468514ac34a9fe4
c9ef532477f0ab9ab6aca83c93a131edd8db6a31
F20101110_AACIST freeman_j_Page_31.tif
9d70053c5efd9598dbd92224bccfca8b
71638f8c4577b914b5a624696f946622c256b59c
89666 F20101110_AACINW freeman_j_Page_72.jpg
e5651234e23355b699bb1eaf8bd003b9
825eee03b4c87f02224a21c0554d621926f88140
66648 F20101110_AACIXR freeman_j_Page_79.pro
8dd91f659d8e2feedc190c9574675ba3
84e86c831885918fe5fefe8ca3ad50efd63fcf7b
2859 F20101110_AACJEP freeman_j_Page_45thm.jpg
454c824f94d2678b9bc435e53b903e18
791939605589ca0b2c398ef0466bd2625011aecf
F20101110_AACISU freeman_j_Page_32.tif
cbffb6e700cd92a95d60400b831a4d2e
9f528d6c3c7bbf5d621404cfc27905184c08bc20
89860 F20101110_AACINX freeman_j_Page_73.jpg
00f06ccdd41dfb6bcb9f26c333fcb4e5
3b9013711c51cf473526b36d005addfba34c9aa1
64947 F20101110_AACIXS freeman_j_Page_80.pro
75c460771e7c5f821279454989266aa3
4b278c39616c6384121610fff9ffd04ae7e8b8cb
12084 F20101110_AACJEQ freeman_j_Page_46.QC.jpg
e2a2ac371da22c78da2bfdfb8fec7e56
d07b149ecd05684ba87b2b1156834dce1c2a3a1c
F20101110_AACISV freeman_j_Page_33.tif
316e77312d22041e212d6678756119b6
eae15ffa21151e00c8a363f535ef41fb84eaedcf
F20101110_AACILA freeman_j_Page_07.tif
866bd98e79b512bc5e34c401330db9ce
b0c6e4a5ba2958ab73921ed00be9586e5492ee10
89960 F20101110_AACINY freeman_j_Page_74.jpg
4f9ef8ab824ad444dd63bc73a0bc408c
afeba0dc93e4a79333c1bf815be59ebd946d9149
64118 F20101110_AACIXT freeman_j_Page_81.pro
434626f571f7cb0bf1d2fd55af0292e4
70bb161e8af9110abe1498445921c0c97e33f541
3748 F20101110_AACJER freeman_j_Page_46thm.jpg
b25adb77abb1984689488380e4fa6f2b
4b2441ba350625ca0dc160a72f45fa003c8ae971
F20101110_AACISW freeman_j_Page_34.tif
c801b0026747c3e30558aca865fe8110
4b95cbaa33b5e947407deb2f1c79999d16d69e2c
129092 F20101110_AACILB UFE0021874_00001.xml FULL
4b3661e8400c8385a7a53e5c11e8f49f
530d1a2de659e278ec40233615a50d87ed27df70
8435 F20101110_AACINZ freeman_j_Page_75.jpg
d260881eaaf1104f19192e33f1ba6ded
14d62c60a6a7f4872cbdc3da94e2ddb3191fe57d
3872 F20101110_AACJES freeman_j_Page_47thm.jpg
563c06089679c812bc969342bde06341
7f102a591954d9cdbb42ddc1317f34078c3d349f
F20101110_AACISX freeman_j_Page_35.tif
11c620c6539d0f65f1e8896f07f24451
56c6da3c3275c07a4db788f579b4600aeb8d425c
62964 F20101110_AACIXU freeman_j_Page_82.pro
1882944dcafde05cfcf6673d1fe11bc7
badab2e8e710781228e2cc343668512a6dc0dcae
13686 F20101110_AACJET freeman_j_Page_48.QC.jpg
bcf7a8d8256bf3ed23763760f9739266
6f25fba1ea3ec49436d3c738fcb6bfe12c1492d6
F20101110_AACISY freeman_j_Page_36.tif
499b3379c23e9eba49ab740968a3b4e9
b0429ce1830c2c9ab146e0f4336064c915f92967
67746 F20101110_AACIXV freeman_j_Page_83.pro
bb405d074a1c5ee87319308b6f55d7fb
d5aacc6d4e04002db3bddfe00e3f0ce76cc15341
311693 F20101110_AACIQA freeman_j_Page_43.jp2
00228d3409101483043288378d01a3a9
b8fc3ddf5c62ca9d1fa52f8a3fbf5bfe84fa330e
4008 F20101110_AACJEU freeman_j_Page_48thm.jpg
03e91854c9a2fe9187220b8fb4dc7d23
9c1563346f9610ff894386bd61ec1484f5a10d9f
F20101110_AACISZ freeman_j_Page_37.tif
7ebf03afab50c8ddaec836aee5cbdff8
d60e44061fad90d2d66d69714d321aceabb55e1d
22262 F20101110_AACILE freeman_j_Page_01.jpg
801668edec13b439ed4f279895e8cf7f
65c7f9b9f8ed223bb5f0bc17b60e5831b872bb94
9473 F20101110_AACIXW freeman_j_Page_84.pro
6292e70a22bfd062ba64f8c3fe4c291f
62e1997936265d9f5ccd26bcd4d83a73bc6f3173
304475 F20101110_AACIQB freeman_j_Page_44.jp2
fcef5f8f982f91ea0c33db301e5ed233
5bd840811352fbc79f6ba125326754223b02022d
10110 F20101110_AACJEV freeman_j_Page_49.QC.jpg
9f40a61f3b63caf087be4eb6f6e33fb4
8ac18a82bc1ea1441e6dbf581e11383d10a92af2
3643 F20101110_AACILF freeman_j_Page_02.jpg
1234a0af21c4f338cbed53493c5ea92b
1853f910ee63a19e655bcb31b3a623d34efa384f
11308 F20101110_AACIXX freeman_j_Page_85.pro
31b2a01ddba0a9cd22a9706269fbeb2a
4948dbdfa9fa848f3597d93cf042562a9c75c7c6
319581 F20101110_AACIQC freeman_j_Page_45.jp2
e17fd3bbd82195ab3c2656b7981438ab
f3d6187899137a10ea8ae6de5aa63bd936b1b7d4
3356 F20101110_AACJEW freeman_j_Page_49thm.jpg
92a725820ad902a8757c4934f8bc1044
ba67863b6227e5697933f6c2015b50dd345c4a3b
58389 F20101110_AACILG freeman_j_Page_03.jpg
a86031b028d1ca0a276e0aeeac1049e0
f4a5d3aec900027c6e2e287d0e965666a26afff5
49436 F20101110_AACIVA freeman_j_Page_07.pro
e08ff29a6dec4bfa3b85d34bd0d215fb
78911f54c3b7c297f38b45efd7a77c1057ab7c9f
489 F20101110_AACIXY freeman_j_Page_01.txt
1323ff6ebdb3dd60378c1013e1c9fa0e
c44019bfdc16ce1a8d11fbfdc1c849694ed952ee
532402 F20101110_AACIQD freeman_j_Page_46.jp2
0168ec3700c4625e1b76e8878cb9136e
2b957782b6992aec55fe5a91ceefb5238e9f480b
10437 F20101110_AACJEX freeman_j_Page_50.QC.jpg
51fb5a9d97ace95d40ea5a95e77b47f4
70641bc1cca2007399bb36858a1b3f2a2cbc8b4c
75848 F20101110_AACILH freeman_j_Page_04.jpg
91df0337858565acab6293b967557563
5202b8ae272d6136579ed4f052034a58d9c86988
47246 F20101110_AACIVB freeman_j_Page_08.pro
403d64ca120bc5c8f8d46b8ec260ceb2
4f048e2c14e96104e91b245f8879790d99d351dc
89 F20101110_AACIXZ freeman_j_Page_02.txt
05cd8104865c4a6b9fc14e939eb047d2
923509f033b5cc149383207305b4acd70bf9fbed
532830 F20101110_AACIQE freeman_j_Page_47.jp2
320d1e25d2e6f98d68d1c08c2adfb9f9
d1946a67461e3bc7830663c57e5f433ec7e20ccc
2868 F20101110_AACJEY freeman_j_Page_50thm.jpg
e0f3cf82a6acb224f531faff8894097e
e13e199b736518e2356ae45986d5f5f34dc39f52
27831 F20101110_AACIVC freeman_j_Page_09.pro
2f0c350468d0c2930c7cff124866dc2c
cc06a16d6078dbebd76e4536c0e501f962583b6f
541277 F20101110_AACIQF freeman_j_Page_48.jp2
1a276da1d27644f25295179afccb8892
264fd8f673fcee46f68541460598f86f82424376
28705 F20101110_AACJEZ freeman_j_Page_51.QC.jpg
451c40867ce80b64741615c8e864202b
821dd2ab92ff5f226eea88f6390b5b1b89e710ec
6704 F20101110_AACJCA freeman_j_Page_11thm.jpg
91f91fe0eec0210ceb4474b79078489f
6be2b1be20e660a500a7ec77a4e8ddee247db160
29813 F20101110_AACILI freeman_j_Page_05.jpg
89a396328fc4005bc342ba31cc0ed273
50128def0dbc90b90a4beb9b7862830746496913
54244 F20101110_AACIVD freeman_j_Page_10.pro
16d8bbb27e04fd1fdccdece6d940ffe0
828c6275004cb6687cbe4d7401dcfce2b3765343
355383 F20101110_AACIQG freeman_j_Page_49.jp2
563c0c35be23b379a0229878838307fe
877396d31a7babfbf2f4c9e69aa21c22e46a8071
14706 F20101110_AACJCB freeman_j_Page_12.QC.jpg
87f9275703423adff5daa97db20c6036
b9982b7a87e50cfa9470c5d7e4021e7dc563f185
21553 F20101110_AACILJ freeman_j_Page_06.jpg
88250297478777f38f6ac2c65ee43059
f449994871aac1f03f4df4efd2938994e42371ed
54868 F20101110_AACIVE freeman_j_Page_11.pro
9de1caf0ee27263e563cf2d646fa449a
3c9b39822359d23bfe51d8a26eab0693867fbcd0
341327 F20101110_AACIQH freeman_j_Page_50.jp2
8f91113b900043cbd4539a02641341b5
5eef18c0ae3976063dd06cb4ef286f9b985e91d3
5388 F20101110_AACJHA freeman_j_Page_77thm.jpg
9bf5ed55eeb9a03b61270b3e58579f68
4d9585be1c24baa7583fe57d4388cc7ae3fd222d
3597 F20101110_AACJCC freeman_j_Page_12thm.jpg
8ce864a571d5476236e7e7fc425b64ea
761b6a7393d1ca7da65ab9f4be9f1abfa3d755de
48461 F20101110_AACILK freeman_j_Page_07.jpg
9e86e9273c58dc9f8e0629a3c862c2e5
e004b7c8fae66588112b138773aac5b11bf616c4
30263 F20101110_AACIVF freeman_j_Page_12.pro
fe695b477f70889e5f3ac2e623b1057c
2924dbb451ad431efb92cc32a5968847edfcd529
1051940 F20101110_AACIQI freeman_j_Page_51.jp2
bb7cc9daf4bf044013aceb3f4a386b42
6e40a3204aa4f045790e4b8004f42bbc32810ac5
29468 F20101110_AACJHB freeman_j_Page_78.QC.jpg
9d4539d296096a9a1c70394a101c8c2c
0661b1abadaa3fe01e1659795b74b56636dc69ed
28563 F20101110_AACJCD freeman_j_Page_13.QC.jpg
bfb1d81a1c2012f1696fdd0ed6294d92
f28759d8d8313cf2cc7064ecdc345083e0d9bb27
81277 F20101110_AACILL freeman_j_Page_08.jpg
ab677a31d5d63c4d79df0bd12763131d
40368db48ca9f96529efeb9036eb127451bffee0
53194 F20101110_AACIVG freeman_j_Page_13.pro
987b99e6ea52348607858aab46c7435c
f819d44a172e243254ea159dd7fcd3f8e0258ae5
1051949 F20101110_AACIQJ freeman_j_Page_52.jp2
2cc15f7a903bbaf48e15a7eb3288b9a7
aae840cb1fe2b4cd84d73c594231bdea4f256689
7172 F20101110_AACJHC freeman_j_Page_78thm.jpg
06435a93d222198c17415611e218e018
39c3851f6c1838e3324498e074f50db4e0c11840
6661 F20101110_AACJCE freeman_j_Page_13thm.jpg
b029b74c32f2fd7bae50417ab2c94500
9e61e682eb494a7f6f09661520a68f025866008b
48349 F20101110_AACILM freeman_j_Page_09.jpg
727160cb7ac65cc0e12fa6fe3b0b7ddf
dd665b171052a750ae702fed81638ab1ceedac7b
51664 F20101110_AACIVH freeman_j_Page_14.pro
2f7d4ae03ac07d2e42de019491684713
615ff815c6500601fd04eeb4bbffb744bb25e4c8
1051979 F20101110_AACIQK freeman_j_Page_53.jp2
428afded20e748403975f77e5097a7e5
827f344a51d8dd9d6ffb749e4521462d074f2cf3
26855 F20101110_AACJCF freeman_j_Page_14.QC.jpg
7289ab8b9fb4ab5bb975266738bd73ca
2803560422c8d144910c8918f43fbf3dd06625b8
90885 F20101110_AACILN freeman_j_Page_10.jpg
4cb684a1e4977b9426ba03ca40dac7f0
15d21dc2256c767ed3750c3a5b8597501c3ac3cd
50764 F20101110_AACIVI freeman_j_Page_15.pro
f83f1cd195064d437373eba61c8ad3cc
f4ac94ff9a2fd8d7dc4bc0fff2a5f4613e97b3e5
1051964 F20101110_AACIQL freeman_j_Page_54.jp2
88c785a3197d2a9f2ac7976ffa6b43da
138206746992b289f2c4e81e99cbee2a56486a89
29991 F20101110_AACJHD freeman_j_Page_79.QC.jpg
ed8cabe6d592fcca777a057807b5e6b6
956ab17ef03d365a2001c861e08599c28b1e00dd
6470 F20101110_AACJCG freeman_j_Page_14thm.jpg
faee5ca81dc85079b54d9cd548662196
0233f149be2fc274f7b80c59957aaf51bad5ea60
89513 F20101110_AACILO freeman_j_Page_11.jpg
e001b1039de9ce757002b1cb3d0ab9ea
daeef4a2eb53336df8f096db76b6582ad67096ea
55146 F20101110_AACIVJ freeman_j_Page_16.pro
b94d767d54484d5503812df4b4bd5ae4
e1e13c2ee6d6ce9586e8801110be8e8b87554f49
1051971 F20101110_AACIQM freeman_j_Page_55.jp2
da58b6de6f855f218d8970800ed544f1
f4e028189e725de50971b6c3d56f837c0fdbe26f
7170 F20101110_AACJHE freeman_j_Page_79thm.jpg
e6e05244df3f5a8e98a0b3bb2b3941fb
a63ab69eb4607ab192523df32261691933e81def
26100 F20101110_AACJCH freeman_j_Page_15.QC.jpg
e0043a686119e96a37225694b22ad2ac
044d09c8f10294581a84d3db5fe432e13c16a352
49877 F20101110_AACILP freeman_j_Page_12.jpg
f499bd4ce8c006bf3197d99c2a1c8953
13b03475382a93e923b21d17824136f4edb361a5
47854 F20101110_AACIVK freeman_j_Page_17.pro
933cac9362cf0d59fa9ab5a795ac41fd
e44bd2e0a15e9288cc7a24499246126df200f76e
29293 F20101110_AACJHF freeman_j_Page_80.QC.jpg
7fa6c3eaabd305e46dd059ad158e146a
353d18f38d680e05c04e7ad8f9ad32240ee45ed6
28294 F20101110_AACJCI freeman_j_Page_16.QC.jpg
7b6d830efdef2c2453f7a56c478c3413
48febd797b0b1427931e6f4177f5c4821b0f4533
90632 F20101110_AACILQ freeman_j_Page_13.jpg
0ed929d0de0f8faaaa0986d719146cb1
fabbffb3bead0085c04c390ccd48999056ed3c8c
53206 F20101110_AACIVL freeman_j_Page_18.pro
d8b2962a467b7d40b280ec7c299b18ba
7b74b0392bf337a48de6f81b7296519c9415177d
1051958 F20101110_AACIQN freeman_j_Page_56.jp2
ddadb96cbc673ee5675cb5cb25bb44c1
11e59f477832a4d4eb366d30da0ffd52cd88d1a0
6954 F20101110_AACJHG freeman_j_Page_80thm.jpg
bafbf766780bd0ca200b99dff13a9a34
4665a909b5f357073be65c71882938fd34d5632f
6560 F20101110_AACJCJ freeman_j_Page_16thm.jpg
c00d9ef04be91f1c0621e5ca753ded1d
38f970c64fddc3fe42a362bf23198b611e4297eb
85401 F20101110_AACILR freeman_j_Page_14.jpg
7fd886ac238f2cf8191200abde414fd5
7a9af70822b8908319ee0f7fc65c17834244ed90
50136 F20101110_AACIVM freeman_j_Page_19.pro
4fd0e949745b108e4948e1251a14ac92
3ea309164ab274a63db89b3d7652e2547d385201
1051985 F20101110_AACIQO freeman_j_Page_57.jp2
c95a869a630d2e9d7c87373c8dc1771d
522ac6117d8d7c381965742a2bad5de24cedec7e
29871 F20101110_AACJHH freeman_j_Page_81.QC.jpg
059d0ba1e21d3777c71f627aed5b9ad8
6c928d6155729aec0321ffb8f59fca0a37d5c9e1
25561 F20101110_AACJCK freeman_j_Page_17.QC.jpg
01ff1b1ec6db0d054126c07a561908a9
5d99d621bfed5ad576cf3b6a703c416fe66f887c
83254 F20101110_AACILS freeman_j_Page_15.jpg
193acbcc13ed552efa55d8913205b03d
269b15979126e5fb726f5f57856c046e3bf730ad
49852 F20101110_AACIVN freeman_j_Page_20.pro
44a86dc9c93f6deaefd5bd591f8e8273
5aefdedb980d53aef2e46d05fb60d8f4249ab173
1051894 F20101110_AACIQP freeman_j_Page_58.jp2
02fd7ec075758eedcc9e1d565e3b1162
ab5eb13c04b1dd71395dca17bf590e3b28eea1d0
7064 F20101110_AACJHI freeman_j_Page_81thm.jpg
eb603ac2a86ddb4c052d96af5df0be76
1c5b081c3568829fb939d723dbcd8e3334ab1b8b
5950 F20101110_AACJCL freeman_j_Page_17thm.jpg
e31546f592d8679aa7dd6ffaded5c370
2a19c2aca31f452c480f3d7dae4f16eef2c7daca
91104 F20101110_AACILT freeman_j_Page_16.jpg
212ee9a2896af6240707ca162fb4c087
4663d80afa1a207a7be7da6b7fbe8319f2df6ca5
55223 F20101110_AACIVO freeman_j_Page_21.pro
15e154f95d2814ad87d4e960375d39c3
8abc97bd327d21aa939aa7676e54f694115b8f62
1051980 F20101110_AACIQQ freeman_j_Page_59.jp2
8c2c08f76e4a34563098ae7bf1ed41d4
0f382f3a8976954d1a35cef34002576687e3013b
28550 F20101110_AACJHJ freeman_j_Page_82.QC.jpg
5ded5a88a7498dd38544b40faf5d9162
064199fd859eaaed03069d5d3e9c92be31a0b692
28005 F20101110_AACJCM freeman_j_Page_18.QC.jpg
ca5e1316ca2c989f90b26b07d148e614
fe42bca658cafc8bb516466c6197d0acd334c83d
80623 F20101110_AACILU freeman_j_Page_17.jpg
0995e72ab2aa860a1a386a9057b34506
e4f06dbc9aa664915e270f93c3d928cd42c60fee
54759 F20101110_AACIVP freeman_j_Page_22.pro
13007c662e7211af2a30631cdc4caf7d
7fb5ffce5ed5dcbfcf6e157cb036cf982ef783ee
1051968 F20101110_AACIQR freeman_j_Page_60.jp2
697e2b45aa3bb683e1c4a943ca7b1dcc
76c8b0a933e768d1af55207c1b99611a6cdc1170
6692 F20101110_AACJHK freeman_j_Page_82thm.jpg
5e0628f72284e0feca94544d397cecd4
ff9462357a796cf32467677955260f436a379160
6737 F20101110_AACJCN freeman_j_Page_18thm.jpg
f04b6596902787b9dfaf3305ac3b7b95
77d6a951d19d85f935998162dd3ad74f4d08d2fc
89350 F20101110_AACILV freeman_j_Page_18.jpg
37453c5c53a8526141b89b5f99af36b0
3436e6d162afa01ebb8c49e0e3a381b8a10c059b
54186 F20101110_AACIVQ freeman_j_Page_23.pro
37e326bcf6d908c42cae13f0b9f31600
247c0d1198a650ccb2997f2d0424c1073b76ebcb
1051942 F20101110_AACIQS freeman_j_Page_61.jp2
e0de653bcc8986abc0fb4f91d081cba3
2a3f663f473a80b5b228bb7113c0c1a10fb9b9ce
29654 F20101110_AACJHL freeman_j_Page_83.QC.jpg
35a1578f44a11a065d1c7b049cae192b
f4a54201c00d020caee3fdda6762dedec1798c91
26883 F20101110_AACJCO freeman_j_Page_19.QC.jpg
6b74f4239bcd8ebdd0baba524a086ec0
c329491b14b45b42c673f42006f8b8c3a6bcd194
85062 F20101110_AACILW freeman_j_Page_19.jpg
fe26699f3d18cf3a9bc5593367a32037
868c6600f91dc05c48283c17be6fd6aa0b022662
52202 F20101110_AACIVR freeman_j_Page_24.pro
4620bddebff5351c8b020ee2d0c4b32f
b118ec0c5e0274e2c7343cb9a58885d6c5d5e383
1051908 F20101110_AACIQT freeman_j_Page_62.jp2
6c51f8530dbee70621c22413184aa2d6
70cd122e64e0ff8ef9f9382d12799f794de87204
7193 F20101110_AACJHM freeman_j_Page_83thm.jpg
50cd637f4412acb647554975beef0513
ead2572d445f7e28d7ef81adfb2d39529653de4b
6556 F20101110_AACJCP freeman_j_Page_19thm.jpg
3e3d6138aa0cf9460ba8f0b909acb27d
27e0b405da4948457bc31c78bb61863bcbb1c18d
82724 F20101110_AACILX freeman_j_Page_20.jpg
3c158b7f2ad49dcc6beb0a6a62225f85
1c4895c32079f67d760ea269a2726d5908982a2a
1051911 F20101110_AACIQU freeman_j_Page_63.jp2
72e10d3ef3a4c4c3d49cc196a991e3b4
b4831029b3ed16d3caf4989d84f65208683d6f30
5488 F20101110_AACJHN freeman_j_Page_84.QC.jpg
b695c8a8fcdccd553df7ccec64eeb474
db6bf57633c1265754ba17e9b849e3262365747a
25729 F20101110_AACJCQ freeman_j_Page_20.QC.jpg
1eda4d53469621d5db7ed613cf4c3dfb
92d46eb2b8e15d9a5e6b94d894ddae1c1bea48bc
51501 F20101110_AACIVS freeman_j_Page_25.pro
c8aa617aaac6b727907c43666f5a76eb
1b6554952decd057b2a6c186cbaf0566c991bc33
1051928 F20101110_AACIQV freeman_j_Page_65.jp2
dae7e64da579f4ef33a1f216c52a8163
b959cba25df925d56df52688a806cd0fefd922a9
91699 F20101110_AACILY freeman_j_Page_21.jpg
07824afefbda937082279f5cf17e4993
7affcd1413791b14e1e1b047c8566b94856bc2dc
1391 F20101110_AACJHO freeman_j_Page_84thm.jpg
a12ef0dde9df15a6a762d7409b1c04e0
82decbaeafdfdacee52360ff48ec6e24ab90254e
6386 F20101110_AACJCR freeman_j_Page_20thm.jpg
ee95387796432a81feac09eab1909f16
1cc68dd4970ca4d441b3f0a22700414982bc5b28
54647 F20101110_AACIVT freeman_j_Page_26.pro
9ced1b6ece40616e76a5bc75a419df52
9a0dd5788ec496a7bdc287c53192711ed0a51ab0
473497 F20101110_AACIQW freeman_j_Page_66.jp2
add8ba2c6ae489f4c614650e4f543413
3a6d00c92d10a8dbdf30f3997eb9213036e35566
90797 F20101110_AACILZ freeman_j_Page_22.jpg
183478f46f35869541ec257c2934bc50
1b0316de1395e3ab76ff86b94a2581494f48ecff
7015 F20101110_AACJHP freeman_j_Page_85.QC.jpg
74faa87aa42776ca1bad1f2262feae94
f5d528ecbacaed35552ff1127dda587bdff6f3ef
28661 F20101110_AACJCS freeman_j_Page_21.QC.jpg
f4fbf992ba3ecb17b04406319e3b67cf
adafb0b507891f554994904785e89ce09f5e453c
52892 F20101110_AACIVU freeman_j_Page_27.pro
dc70174a2c2cf05a3297a99ddd3db1cc
279c641adad425bb7c0322c1130deebcd29ee39c
217083 F20101110_AACIQX freeman_j_Page_67.jp2
f7a25017d86f26937066cd7f6e6370d1
7bbecca973bfdc3cceaae28d54de9dc188196927
1662 F20101110_AACJHQ freeman_j_Page_85thm.jpg
c68154ef8c79e2583c8db77d88a46ca9
2e8c8d582c2a36d59f6a4d9e3e0123849750b277
6914 F20101110_AACJCT freeman_j_Page_21thm.jpg
ba7b257fa12cf48bb9efad52c2e1c676
f6fe339e59266194de080b036cb8b60bc3a08f8c
54697 F20101110_AACIVV freeman_j_Page_28.pro
3612e582f9ccf424b8b1ce9638416e6e
f487a8f5d25161a6e782d74a118f6aa2fefd67ee
102113 F20101110_AACIOA freeman_j_Page_76.jpg
bbcd8f7a0c221a8f57f802c365693efd
2326f32fedecedb7ac1ff6095f167c4dac9d7622
224369 F20101110_AACIQY freeman_j_Page_68.jp2
a6d7afa2bd30b5654b61651ba1a28503
a7095c079f9d3ddc4802d23d16e03a031ec2c088
99936 F20101110_AACJHR UFE0021874_00001.mets
faf355dbf248f22156dc71c421ba03e4
0c7d023fe406370afcf6e81f6da2b19d3694f90d
28397 F20101110_AACJCU freeman_j_Page_22.QC.jpg
85fe88c0e42101d86e6111f88dc96180
52ba51c1ebefb0664425c329f8a0e7065b2be2b8
47798 F20101110_AACIVW freeman_j_Page_29.pro
9213926768676c9585bf37239efaaa85
b60f0c8f5db31558e9d1839e7c6c8826bb15d8d5
87881 F20101110_AACIOB freeman_j_Page_77.jpg
dc5f05b73776f26909e338aa706ea18f
0c91df8464a4bce5b5c6bee14736618d5db7963b
219537 F20101110_AACIQZ freeman_j_Page_69.jp2
81b17b9260b86c19c357b201c2022fe7
4584b8c5e11b79d5fdbaf1b074c43930279738a6
6836 F20101110_AACJCV freeman_j_Page_22thm.jpg
0bb19717af58e01406ea2fff3ae9909f
a626eddfc378921ec6031eb83ecf6e6213065e6d
54514 F20101110_AACIVX freeman_j_Page_30.pro
8e43ff430e0522bdd6fa77a50e91f50f
dca577e337ad6bcab26a99c2ee6dea8ef9ce40b4
105138 F20101110_AACIOC freeman_j_Page_78.jpg
19580ab9ba6a5b5651125000b3f6e4f3
db1a063a1b975d5ea9fb456192551373f19b331e
28570 F20101110_AACJCW freeman_j_Page_23.QC.jpg
2e75ae65049b56730a43711cca1e4a6c
37df1a470cccce8e89962add7a36ad9873708b3a
51924 F20101110_AACIVY freeman_j_Page_31.pro
98f3ae94775f044739e184a47e6ea520
ddeda9942018efb34cf8c1689086ffd069a06ff6
110024 F20101110_AACIOD freeman_j_Page_79.jpg
ae2a61a761ef71ddbf93e82b3fc088b0
3a5762531ed2c105240cf52c38cfc3c9c7a38bec
F20101110_AACITA freeman_j_Page_38.tif
d8f60665a8390b2ef8d52ca2e5922f13
defeea42d4f40a225b588a0ddfd9c9b974b35045
6756 F20101110_AACJCX freeman_j_Page_23thm.jpg
da945ccf185309f74e2e8706e85064df
cbab4be5b634bd4a43c054b28b72a2b27aef5d29
54862 F20101110_AACIVZ freeman_j_Page_32.pro
bdff78cd1d4800d64ff4faa585d2a9d8
b6456536a441a704b25cd8d238ffbf6c6945fe23
109484 F20101110_AACIOE freeman_j_Page_80.jpg
c70ff3ddc42b6b654bb68412f1181215
583ec15142c8314e8aa6da8fc3e52ad4171b5638
F20101110_AACITB freeman_j_Page_40.tif
74262ad88211887990f0459a3b10641a
10e9cc69f32441803b727c85a642289069c4e2d9
27330 F20101110_AACJCY freeman_j_Page_24.QC.jpg
bcbee7c59c2e9b14c4b22501e646dc1e
4194b9c287da3818914d2e346bfe37849ca71433
107867 F20101110_AACIOF freeman_j_Page_81.jpg
0972b1ed8bd3a4980c8cd77d11b8b849
d9e922731ac17b315ca75b2eba2a3a59e4ffd7a1
F20101110_AACITC freeman_j_Page_41.tif
83afccf1d88855a78cddd036fa082344
7e1a22eb86f18c902f8fc2aa0cd6ec0428a20641
6394 F20101110_AACJCZ freeman_j_Page_24thm.jpg
9816d149f6f041608c295a00d4dad970
94fdf15e292c839e90edeac483175dbcf92c1bc4
105339 F20101110_AACIOG freeman_j_Page_82.jpg
c485f3541bcb0d714bfc57a615164520
09754c96345a2d5f930a40af2389ae2f5e3fc3f8
1363 F20101110_AACIYA freeman_j_Page_03.txt
4636178656e74f58c010bd78eaa2806b
bfc8064f603520139c0e94ab180e410e6f0210f7
2159 F20101110_AACJAA freeman_j_Page_56.txt
6c8cce77c9dfc33c9227e9b848fd8bc4
79026b3cdfd915607474fd318bd854d3e5a3579f
F20101110_AACITD freeman_j_Page_42.tif
3a945ad96763531d9f1bd754a16c823e
d39f0795c20f74139f6b39af21ae4247581734b1
107537 F20101110_AACIOH freeman_j_Page_83.jpg
aa3f0614a50f1cf7c6e1bd940e61e993
4fb0d67ebada1c8fb8e4dd50ab37bcb7ccf91598
3474 F20101110_AACIYB freeman_j_Page_04.txt
20a9b092936370d7083666a6ebf99ac8
e16f2d88c14a5f17e9a8c3a24a711867c2e0fe90
2065 F20101110_AACJAB freeman_j_Page_57.txt
5abd77bc1bc73bd71c9ef33526bb437d
d8e45561e062b4ad9aedaa1e03cd3ab76e60dfa0
8423998 F20101110_AACITE freeman_j_Page_43.tif
5a841c2f424477f25e746890922b53fe
36cb3c4cca7ae5e0f177d1c83aba9ac94c8dea83
6676 F20101110_AACJFA freeman_j_Page_51thm.jpg
fb0d23f996559a4ad43870377bb0c1de
9d973024188b44eccf7fbf2dbc76b568391ccf1e
2130 F20101110_AACJAC freeman_j_Page_58.txt
8f24c6938a4409fef26e07b69c7be476
4fd8a53c59f615d34b109de0b4c146ea93c6ddcf
19631 F20101110_AACIOI freeman_j_Page_84.jpg
a8c5a84382806156ebe73f8a68d108c5
20f99be8e06af34381ca1fc3a857823a837226fe
1247 F20101110_AACIYC freeman_j_Page_05.txt
51ecfe07c99ace2d1acea8d95b70f9fb
6651fb83f52670880a1148ceec995f6c9921f234
F20101110_AACITF freeman_j_Page_44.tif
b13a6b45d710678626329ff5ee7b2d26
a3a2d21e6378c0725fefa8edbb74b960dbafb7c7
2081 F20101110_AACJAD freeman_j_Page_59.txt
a80a54b4694deb10cfa35c1ad3e4413f
20abb8e4060178f74870a5e82801d1d284272832
22901 F20101110_AACIOJ freeman_j_Page_85.jpg
143112cab910131965d763df5d8a852f
140d07198934dc3723819cbc682ae821ee8d5d4c
761 F20101110_AACIYD freeman_j_Page_06.txt
cd6c3799675c06bb6c263bc9df10c479
2a031d5ffe49524a5389f4ffe5e9e86c40c93849
F20101110_AACITG freeman_j_Page_45.tif
71be511e8524b7058f3185cfbbc4ecad
bc09840c5b87fadd5b5fa7480a7ddca76b146965
28178 F20101110_AACJFB freeman_j_Page_52.QC.jpg
6a7a87a547bdf6c9717a0cfb1cf7bac4
22b3e8943d15b2fb8d5c43bd0800fd2f744c8815
2146 F20101110_AACJAE freeman_j_Page_60.txt
a80173285fede8d839010622cafef517
527892c408e17531251071f4a61ad1688a70f6c7
260675 F20101110_AACIOK freeman_j_Page_01.jp2
c7e4eb7e658aedbb6b8ab313b53f29be
ce81c88e60f549cecc000af729319aa3968d68bb
2019 F20101110_AACIYE freeman_j_Page_07.txt
bf2606ee24659814c51ccfb8db21da0c
fa9460c60766342f2cef2fb04ad244626013b289
F20101110_AACITH freeman_j_Page_46.tif
4bc379fcf99438121f349a5fa9d953b2
61ba288b5f6e6f717e6503952130efaab35c7905
F20101110_AACJFC freeman_j_Page_52thm.jpg
34174d4d447789a581a9dc39d1ee2d52
90a79e832afacf9945be13f7b142566c0a48d956
1979 F20101110_AACJAF freeman_j_Page_61.txt
abb5d6135fdce8e2b6ca0a862b488d5d
f8f265b2148a48eb4b8141870fd809539aecb039
2068 F20101110_AACIYF freeman_j_Page_08.txt
7a481b5acf276133ae5a99ceb02958fb
e5ea2a9a6c93539daa86dc77fbacd0bb016f83cf
F20101110_AACITI freeman_j_Page_47.tif
2437d580107bfd389994d129518c2610
86c6ab1e9a9aa13e5f8142d73b946e45aa50f226
29800 F20101110_AACJFD freeman_j_Page_53.QC.jpg
90e69d9e1fd478d12ead0c2d1384523c
7aaa48ef1c633c28fd6ab7f69b3fd33d3831a016
2080 F20101110_AACJAG freeman_j_Page_62.txt
39c6ee64fff28beadc17ca928ee61ef0
81206e8df8f1015fa5043c668b00c846b8cac1c3
25597 F20101110_AACIOL freeman_j_Page_02.jp2
d5e491c91c25ae66d879a757e936c3b4
b383ff34d6e59dfe5423267470b123816105ea0c
1103 F20101110_AACIYG freeman_j_Page_09.txt
c708fe301520aeaef582ca695276176d
f77e7d26d02b5adb4805481f99daef6087a80082
F20101110_AACITJ freeman_j_Page_48.tif
a7f61b23220cee30f5f586ca223ce6ea
5c6a74783101a7f09808eacb9981946af8172c4d
7042 F20101110_AACJFE freeman_j_Page_53thm.jpg
bd841b2a48fa189ee41aa06549b44f65
87eb84e991a4ff0e92078b8b39313cb1a967ad1d
2198 F20101110_AACJAH freeman_j_Page_63.txt
c6dd8e4c0f41d96c333181a761b5a26b
83447f34c8d9c3e6da7aaff70688df027b3bc2de
773792 F20101110_AACIOM freeman_j_Page_03.jp2
4607e4f2c4b2fbdf73f3d3b887f605b1
57af25f592dcfca81ed8f5ef36f9fe6bfc4b38d1
2201 F20101110_AACIYH freeman_j_Page_10.txt
c854f4f20f6aa5fd3e8ffa59cb51b6a6
00dcbc7e10970195a3988dc3d2a82c1a3541981f
F20101110_AACITK freeman_j_Page_49.tif
5c6c1d3e9bc93c2a7bc970f1d1d9ad3d
9620351f53cc6600ef61e4bd2a28f30d7ecbba2b
26288 F20101110_AACJFF freeman_j_Page_54.QC.jpg
8b87d226165764164a3509b7a5fd92ca
6bf72bb39e7e80e8c902f62af3c9baac31b8f4f2
2186 F20101110_AACJAI freeman_j_Page_64.txt
833a407b494e3124604d34327cffe343
24500565c24d8cff208ab8d1245a0c2ca9c807a5
1051984 F20101110_AACION freeman_j_Page_04.jp2
3cda348e314f5d2752da5dde6f69bf17
2abca8e6fcedd5423c7cf888cce07dc8ca503918
F20101110_AACIYI freeman_j_Page_11.txt
1d6c8d0d822881604706d87abd5c11fe
38231dbff8046c2fadb0e98a2f930bc9f7e541f9
F20101110_AACITL freeman_j_Page_50.tif
ebb25a4fe16de570f68ea371b4c7af83
8abaf8a2ac8d06be26ab891bdc6d84963b74897d
6550 F20101110_AACJFG freeman_j_Page_54thm.jpg
51d108fbe2917f196228c6149077d744
c765bce93babeb1139a32aefb2c3e03abd03d93a
1939 F20101110_AACJAJ freeman_j_Page_65.txt
fcc192d56bb6bf66e2286ef2e27fa6b0
0dadd9990fec2c6043b29e3610dbccdf9f37ba39
514652 F20101110_AACIOO freeman_j_Page_05.jp2
a540bbf0ab47922838c05a4f21cd2a2e
9daf51d977d727db4fb3294611e2a381c9f685bc
1225 F20101110_AACIYJ freeman_j_Page_12.txt
0727704da89f6bd252149c5daeac5398
5d203546feb183de534a7f348b7de81f9a825512
F20101110_AACITM freeman_j_Page_51.tif
cb3de9f1ec0c55bf2a9ca9b8cfec64e0
c7e17c7e98d5eb02e8a6e090d6a71c49aa89ff51
27253 F20101110_AACJFH freeman_j_Page_55.QC.jpg
07d823bff6ac005b57ce8b6932734777
fc93356bec0eec473bf704d9dd86ecaa6458850c
1144 F20101110_AACJAK freeman_j_Page_66.txt
bbbde85b362aad8a4f119cc500f335cc
a3905e7bee4286d2a5d8a1362865fac9d1258dc4
294822 F20101110_AACIOP freeman_j_Page_06.jp2
8059e4a4c7f153ddb79c234ac124a0ae
a277fd4da5134e758e246eaaff5f3a7f208a486f
2171 F20101110_AACIYK freeman_j_Page_13.txt
4acb6bca04a69c17d0c8ffd2615c2e8f
22ca30e933e83ba072573096aec40dd2c9bba4a4
F20101110_AACITN freeman_j_Page_52.tif
8d1078aac0f328a8c97cefbe72317925
ab46e11d3e57c29c761c94d98e1d4d02c5addde3
6564 F20101110_AACJFI freeman_j_Page_55thm.jpg
d2f7255f6c29b3ec846564eee97c4c7a
f358d5fface407de099b2d5b1bdfd4085b219fcb
726 F20101110_AACJAL freeman_j_Page_67.txt
b5ebe8041e8e16c2b493ac21bb264282
29331616c0aaccad30f7ca458cf2be19de7b2914
733484 F20101110_AACIOQ freeman_j_Page_07.jp2
b8b1d97dc799172a0ae3815ac1f7e034
77ac37bbd230e92053ed1be7c77ea2136896984d
2037 F20101110_AACIYL freeman_j_Page_14.txt
d8cc80938fa90a2f6fd55c1145f3eef7
3abcd700cd368c26231dd98fab54a9caccdb8e73
F20101110_AACITO freeman_j_Page_53.tif
07e382c077a8a7af71f946a838e3e7e1
6be44768c689edf20337f255eb20b745214acee7
28279 F20101110_AACJFJ freeman_j_Page_56.QC.jpg
dacc456ba88aca3393e24ef787fde322
538f54fe46273feba3896c9eb9be4ff8fcf2536e
573 F20101110_AACJAM freeman_j_Page_68.txt
d3fd7503b4f0cb86d8fc7b69fb4892f3
7d2b8ac31b377de9fd4955e51250972ec2023b38
1051965 F20101110_AACIOR freeman_j_Page_08.jp2
5271f42e449279243982636a2005e978
6a56b0095ae1ce8b12237dab5ff93bbe1197db7d
2003 F20101110_AACIYM freeman_j_Page_15.txt
37f5b354610e8cbea87fe6391bdb47d4
c8d0370931d3f6e1c14d365b006fc009d76f791f
F20101110_AACITP freeman_j_Page_54.tif
74ac51e97fd36b5c9deb6256ea0b29d5
1789f14de35307a30296bc04b25304b8252909b9
6939 F20101110_AACJFK freeman_j_Page_56thm.jpg
3b675ada92c0227b500335e63fff4ff2
b76cda090333ef7f70208c811bc8a5f2df812630
752 F20101110_AACJAN freeman_j_Page_69.txt
8bcdff26d1b0ee2770e67a5963ef4a02
3823b4b05de65df400dbecc254f143037c2ddff8
633162 F20101110_AACIOS freeman_j_Page_09.jp2
826048ad7ce355a26d54f0a006ad7278
f4cec69018996867c93170d25d3a7a676d5244b9
2166 F20101110_AACIYN freeman_j_Page_16.txt
9b3262daf370c2de694d90c64982ecb6
5d887c07f7e9b264838feacb1990b95f1d7769ca
27236 F20101110_AACJFL freeman_j_Page_57.QC.jpg
c50da4ef10d4ab11b78f24c073d6206d
47e1e884e94ed066de49aeac521411c18381d99e
644 F20101110_AACJAO freeman_j_Page_70.txt
daf368c2cc17f220a724c76e1cf16a0a
50379e1f7fff7e05c3e453fc2cd349f95c8f749d
1051982 F20101110_AACIOT freeman_j_Page_10.jp2
1ad9184aa9146dc6e010f0b93f69b546
75d010511ddbe91df0e787927c101c51c60ce8d8
1903 F20101110_AACIYO freeman_j_Page_17.txt
96976ee75bdc59c87fbd2d1b98972590
fc81c9d17261653ca9a0430ff0dc21fcce3452f1
F20101110_AACITQ freeman_j_Page_55.tif
b49c8dcd902ce29f6990d7b15b5fdfc3
43afff186b4dc392be5569c6eaad98858e4eabe7
6312 F20101110_AACJFM freeman_j_Page_57thm.jpg
f53417f3260805b0809fafed69ec7fa3
2ce1869d7aa9aa59dde960f81379049e93bc3d51
657 F20101110_AACJAP freeman_j_Page_71.txt
de77fe563f8650995852df1854aa9348
95fc43b6cb39cbed7d8d74b4751efcebdf09ce01
2093 F20101110_AACIYP freeman_j_Page_18.txt
abd659c5f0c0a865da85e1ac708a1ad2
476c7b62d4db40e9b5dc32d5878f33c756c0cbda
F20101110_AACITR freeman_j_Page_57.tif
636d0952d70ddd27a47a5b870e5f5183
1a0f3084454e22dbf0bb5c1f7b79f31a526d6370
1051956 F20101110_AACIOU freeman_j_Page_11.jp2
3945d22dec05b05962520ad289f7c92c
d1455d133179afc9f4febc7e5287d22f97df01be
28820 F20101110_AACJFN freeman_j_Page_58.QC.jpg
563fd1011d0fdb6f5720affd2669b72f
5df81353971cb5fb381c404691e6f9dae9d40ede
2179 F20101110_AACJAQ freeman_j_Page_72.txt
05ff47a2b0a808ec36ead3333407aedc
0ca554bb28c3fecc858f4093ec997608d879d50c
1984 F20101110_AACIYQ freeman_j_Page_19.txt
c6ac8e65ac1570acead83bd4fc877be2
88b6e5bd0fc6defd145d9ff902c13f3f70ea11d6
F20101110_AACITS freeman_j_Page_58.tif
be0b5794c4621255f5b3e7745c2e8cb9
80467576c7635a8dfba30895e9a204b7e335712d
676588 F20101110_AACIOV freeman_j_Page_12.jp2
3032ff0a5b8d876933276f5340f6e4bb
7f75091a6d192cf2137baf94d3be331247d756be
6647 F20101110_AACJFO freeman_j_Page_58thm.jpg
cc964a7cad6c67748b060dc557bf3c6f
c489c8fb3058684d3565281c13086f19bd43d54e
2144 F20101110_AACJAR freeman_j_Page_73.txt
cec64493dac869ad9eb6fe7729d70944
23c53079984ec420b335f7c904fe15fd7ef23e0f
F20101110_AACIYR freeman_j_Page_20.txt
abff3fc971bab835ef9d22c634803937
32aa4f0fc7b4d5128fea4c3c0e3c31734a0b9ddd
F20101110_AACITT freeman_j_Page_59.tif
2211acb1b3fc5d1066769043379b5c65
b184542042f1a60513368caa700c26a133db483c
1051944 F20101110_AACIOW freeman_j_Page_13.jp2
a8c86a1afcbb35858b3efcc34dfb96ee
913992a18d9673d31ffafba54ee60013bd2855dd
27499 F20101110_AACJFP freeman_j_Page_59.QC.jpg
b50bc2581571211a40d2ea43be80203a
22ca2a902c7228dcf6f4bd1935acf7f01beaa547
2120 F20101110_AACJAS freeman_j_Page_74.txt
bc1f86f820415ae50671f90b9fd31ec6
ccef58534b74b37bcab1d079a3645a5e6770ca33
2162 F20101110_AACIYS freeman_j_Page_21.txt
ede0f2d3bf6f67b76cd293ea113cd4cb
55e268a77e023ce0332483aaf5d13c38111bcf6f
F20101110_AACITU freeman_j_Page_60.tif
a85d532633559f98f5e2a2a63294e0ab
ff4034137842e01cf4161d486671cb12acf9fbb4
1051978 F20101110_AACIOX freeman_j_Page_14.jp2
f825d5fa5d5028e8d2b02d738d441bac
77b5de904f0c7e6e9c0aad9b7c441bd92899c326
6842 F20101110_AACJFQ freeman_j_Page_59thm.jpg
e8daec7036e2e1e77e1ab4e90b04e8b5
f4cc2846823c91e8d60943153046ce618874ec1d
147 F20101110_AACJAT freeman_j_Page_75.txt
b1b4bbb29c0ae94d7cbfc5be12f0e0e3
b8886293bf6c62df75d44c617ae8690103835ff7
2165 F20101110_AACIYT freeman_j_Page_22.txt
539f608d49f698e3699542af43342145
a2c6a685998a27369764623ed64338246100e467
F20101110_AACITV freeman_j_Page_61.tif
8291681bec35a9432e0930af1fb7234b
faa0dfe9971e3c746504359fb13b7f91a5fe0577
91385 F20101110_AACIMA freeman_j_Page_23.jpg
4c5f0b8b34faab69b2c2f4493f1dd858
26b02e4baa63deb77646e3c0f248acf72389aaf0
F20101110_AACIOY freeman_j_Page_15.jp2
ec9b57f1120e12fdcee5d48eaa7d4279
188fac27b0ad7a0a949cfd37a57137b51240ffef
28956 F20101110_AACJFR freeman_j_Page_60.QC.jpg
00da5cde7fae33cbd03d5b03bbf70fba
ccd9e7ed5f0c06c632cca0713baa964b30c50232
2954 F20101110_AACJAU freeman_j_Page_76.txt
9157ab4b2eac6da994782c40b7820398
0f3b526db9f873561ab454114719327a5071a230
2123 F20101110_AACIYU freeman_j_Page_23.txt
7852609d052b11e1011c20e693a97193
99d080f4a60ad4c80a68a37d7421c5febbc13394
F20101110_AACITW freeman_j_Page_62.tif
97456b833a59765264928aee17bd4285
f918e43e78361aef883b41f5d0c30014c4ba7b58
87260 F20101110_AACIMB freeman_j_Page_24.jpg
76d97334608957789ba22f5c9db40b6f
c23e0c0769a0e6c278d4b6ab7503c0d4eb15ba37
1051967 F20101110_AACIOZ freeman_j_Page_16.jp2
05cd93603f9ac81550a992a79ae133c1
467e795ba4d471781444c01445940950518ef210
6426 F20101110_AACJFS freeman_j_Page_60thm.jpg
d84c20892c0a17a4fd36eea218de0275
401a9ca81e3f97e5ac9329d2ec643fafa66862b0
3074 F20101110_AACJAV freeman_j_Page_77.txt
118fded8e4166c362869ecd2240f1760
f218a297d017a2e25e84a920ba74e42fd4ac7544
F20101110_AACITX freeman_j_Page_63.tif
2f7e4a1554fcaea5d6c89922a505febb
f1944202191f901caf7b7ad034cac02e2aef0fa1
85592 F20101110_AACIMC freeman_j_Page_25.jpg
d7d10832d5567e4967e054e9e521aede
de879808d0c620e3ac53420182e6bbe1eed8293b
25322 F20101110_AACJFT freeman_j_Page_61.QC.jpg
3a2b7d8bcabd9155faf74c02178da3ec
22ec612de413a6fa8f29471292bccdcc6fcb9054
2537 F20101110_AACJAW freeman_j_Page_78.txt
1574e56883b7b4ecb4b165665a0724f7
d3582f51b36776339f4cad369cdd7c1277902317
2049 F20101110_AACIYV freeman_j_Page_24.txt
dbb325088955a13fcd9ad461707eddbc
06dc4af882704c7845ba4d1dbca69f3fc78ff43e
184296 F20101110_AACIRA freeman_j_Page_70.jp2
586ca68c34b027151dc5fac0fa30ca9c
5c53835020bb4fdae0d656f6b29ec8355d591f97
F20101110_AACITY freeman_j_Page_64.tif
a69ee6435cf5ff92fc34d1aac9189526
eeb373b4e38a867a73f58a12c12188fc8c47223f
89479 F20101110_AACIMD freeman_j_Page_26.jpg
fe8dee1e9e9179343194d06883536b93
33def4294ac9c47c66b0fc097a223dd96722b20f
6009 F20101110_AACJFU freeman_j_Page_61thm.jpg
d01e10346164736fea989952683b04b4
eaefd7c3b212f1ca45df12155b23e36cb1f530cb
2638 F20101110_AACJAX freeman_j_Page_79.txt
872ef18497c667b1c396ea9cf3971aee
582473761635d9fdbfd97ffc0b19cc03288126e4
2038 F20101110_AACIYW freeman_j_Page_25.txt
9b1912cc7405440ab9c78f42a89997e6
165368cf673c364c10c76a001c285b8ad48edf46
321227 F20101110_AACIRB freeman_j_Page_71.jp2
98a70c76772751e106a65bf0b8dc3cac
b243d767f243f4ba58f7fe2804eec1515e0dd608
F20101110_AACITZ freeman_j_Page_65.tif
deee9e4a6531fc4b142f51f690f3bfa5
e0b2734da834529a17107b8fd8fab482343c843d
87550 F20101110_AACIME freeman_j_Page_27.jpg
b44d85f52e77d9b969c6346bfeebd9b7
a42b24445421b0ece19d15685253197a1515c1c7
26553 F20101110_AACJFV freeman_j_Page_62.QC.jpg
a91d22cef832a2efe344d6f946ff7857
444e49b67566966aae28ea3803d10745de519068
2581 F20101110_AACJAY freeman_j_Page_80.txt
070be346a8be2ec5bced315801a85692
1b08942837a1f59e8c60279c13bd62e3cad19f7f
2145 F20101110_AACIYX freeman_j_Page_26.txt
4cd6ada26bae9d2dd27e0bb467696d99
86b4da09a9845e97ab3f44a24d245a62f2158f21
1051960 F20101110_AACIRC freeman_j_Page_72.jp2
6cbeaf4c19c1eca63798316cbf7e35d2
954dbd6cd02bfef4dbe03e5a5b93416057540571
89683 F20101110_AACIMF freeman_j_Page_28.jpg
c76ee5bd6dbee2dcedf27e3bbb226b61
3a0d4fba87525e7079f15ead3d0e13a5effeea53
6339 F20101110_AACJFW freeman_j_Page_62thm.jpg
75dc7e0b7199349a3850800efdd2ecf4
dd26bf479d4f64fe37c417f4a07025d319f0b8f9
2561 F20101110_AACJAZ freeman_j_Page_81.txt
7dd46f8b53e948ed3544bb15ec3eeeb2
a69505271db9d69ccb8d7077c4ce390bdea3df63
2086 F20101110_AACIYY freeman_j_Page_27.txt
cda73119dcc912ff0c110221309b4c65
565e8f502985309372693a28073ff3c1a70e40ce
1051906 F20101110_AACIRD freeman_j_Page_73.jp2
13c5610842ae4fe118c48b1919429f72
ac550f85d7c3a718f9c2aa03feb7cf7626e98a29
81296 F20101110_AACIMG freeman_j_Page_29.jpg
ef5919e7732d613a72a0da4bdb829414
974e2acb5c5105837ef3ee38d85e1bafa83cd74f
56366 F20101110_AACIWA freeman_j_Page_33.pro
6c578fb0209e8f2ab1237c7380413334
d83131594e4e075c129335644fa875f1f4496980
28665 F20101110_AACJFX freeman_j_Page_63.QC.jpg
c09af3f07bfb5e9ac009850aeaefb594
e94735d6aa3ff76e8993c03d91e29dcc3d84e0ef
2147 F20101110_AACIYZ freeman_j_Page_28.txt
0be12a3767880cf4a6adc0487635bd05
a5bea8143f7c9757ead6467d53bbe2a04a4b644e
1051973 F20101110_AACIRE freeman_j_Page_74.jp2
d95df5938d3e2b42c651c4dd01cc08ba
6d7800f8ef62e0ed48c6510d39fc438dc8ead4a8
88509 F20101110_AACIMH freeman_j_Page_30.jpg
d68d6aa37d5fc70a0b54c61c678c25f9
24af2aed6278b6a7ac285cf39909ff94f5934331
55898 F20101110_AACIWB freeman_j_Page_34.pro
4098dc34deaea34c6f22ec1fa3e22410
9f34d43e9b3ddf534aef83c51b4fb7bf2b47a3fa
6840 F20101110_AACJFY freeman_j_Page_63thm.jpg
fc4fc1ebec8a3a270fcb52dfd5a56043
1b65e4c306dd4fa00e2812f9bba3aa8e4eadac12
87641 F20101110_AACIRF freeman_j_Page_75.jp2
d656db87d6e2080d3ac173e9a36233e0
25b1665c4eaf95d22630ccd3b8f69a7ddc5b1821
84842 F20101110_AACIMI freeman_j_Page_31.jpg
8bd2710510cc3f66389f83dcf7b943ec
4539441e8ecfef802161ec9fe33c0f2e21a01d44
57163 F20101110_AACIWC freeman_j_Page_35.pro
4b83bd13a8a2c40bf949973e24b324b6
c9e069232e2cae63e28ed5ee58f942c7fe7d9542
28190 F20101110_AACJFZ freeman_j_Page_64.QC.jpg
356e32acc69ed45c6804564c2889f796
50727b7786ce0da7a4b354a4874b5348a8914761
26940 F20101110_AACJDA freeman_j_Page_25.QC.jpg
d71e3a516586ffb13be23904bd8d60df
ef9791e085c0361b0e27bea1859b88e7b41ca864
1051959 F20101110_AACIRG freeman_j_Page_76.jp2
5c70c8a076aba2e41d4cc2f61c750034
57bbb54cfdeaf18396a64c2e1f40c351b7043f9a
6765 F20101110_AACIWD freeman_j_Page_36.pro
3dd419e8db7a1a474bc8972ddfe7f486
b843e4de6d03bd5b104c896810223be0d23d49a8
6445 F20101110_AACJDB freeman_j_Page_25thm.jpg
00dc5fdcac37e94911cc1a159f31a4c4
b9cae0a366a1e9f8f9994959a57d1e027eb2bc77
1051912 F20101110_AACIRH freeman_j_Page_77.jp2
617955b72f481dbf9bdef298883dc554
15f6de728245941ab087b836bca244d14f4bfa38
89380 F20101110_AACIMJ freeman_j_Page_32.jpg
aa288884257254908b86623952731267
fd409640390975ef4174d3a6461ed3dd51bdc856
34052 F20101110_AACIWE freeman_j_Page_37.pro
298aa0644d5290b681e7c646bde80a9a
49fadb07ac5016afcc4246ff13ef202130a8bfd3
28553 F20101110_AACJDC freeman_j_Page_26.QC.jpg
cf901dacb1c7ec64cd013c873b6e878a
4c466484fc127ae6bce961b314e76f11342343c1
F20101110_AACIRI freeman_j_Page_78.jp2
c678f8bb65372530dcc001be05e1c980
e3a875d774957887323d32794e2553c78b168847
91721 F20101110_AACIMK freeman_j_Page_33.jpg
62541bd06a7400663985a9e278274e3e
df82f5147b2aa511635d1ee91015b36ae3e794c0
19071 F20101110_AACIWF freeman_j_Page_38.pro
115651ca0a5a5aa1b78a3c7a4213314c
f888b8cefca323c9160672368a869b8784e07d4b
6719 F20101110_AACJDD freeman_j_Page_26thm.jpg
6107b61116e534326b90a0eea249c420
27dac79b1b8effe925d550e1230494295673a547
1051966 F20101110_AACIRJ freeman_j_Page_79.jp2
c05f9a2e30885d3b29dc51eacc817f52
1d01a67f9f669cfb242722cd8c5e5aa6d4c981c0
90427 F20101110_AACIML freeman_j_Page_34.jpg
aed5d7db02b083cc4fba5b8aa0ea3368
2033819044e0f6c1217c34af6cd2957dd1470b5d
33720 F20101110_AACIWG freeman_j_Page_39.pro
63d2232e3d28bdb125b955976960f72a
ba9b1b0ae29ccd371bb3ad9cc13538c5d3553daf
28564 F20101110_AACJDE freeman_j_Page_27.QC.jpg
35f173e2fc998b1fd00553f7f9bc2280
14ad4c9361b60e24cd403ed68ffb44ea1259e25d
F20101110_AACIRK freeman_j_Page_80.jp2
1c791180e81280995ed52c0c6ef50ae6
c6524ba4c2f398b7aea6019caef15742473945ba
92240 F20101110_AACIMM freeman_j_Page_35.jpg
eabedd7e77bf608291574c97a79bf2c7
8c35bd0b3d1b0a7953f1101ee79d550b999fd680
10167 F20101110_AACIWH freeman_j_Page_40.pro
cecd3e8cc52b5ad06a7973c16a136c5a
ab8c9bcbcb08686a4205275579df2289160ea3fc
6597 F20101110_AACJDF freeman_j_Page_27thm.jpg
a3689e18599451f0d2604b1e070c71a4
3627a552c4d0be0158d723815d9ae473bfc079fa
1051983 F20101110_AACIRL freeman_j_Page_81.jp2
429887dd2d53b9d9deb19070c24c703c
73ec867ff3d5401e64584b0da57f93c63775e95c
14096 F20101110_AACIMN freeman_j_Page_36.jpg
3fa51b2fd76986a55917b40c6cb37232
25114ee91ae3604e48e9b6017034af4678a7d72d
11922 F20101110_AACIWI freeman_j_Page_41.pro
1ce781037bb94f6c5619401741043ffa
40e7c57bf7c144b7749dd54b41edca4aecf94030
28400 F20101110_AACJDG freeman_j_Page_28.QC.jpg
47bb687f374dfc93049594eb45479a38
a4cb8660ed273ae5d6bfd18f6a2449acc72b321b
1051976 F20101110_AACIRM freeman_j_Page_82.jp2
5ca8d8176b828304c403e5c2c2b66d70
aae125451a85042aa6aa85b335e1e4293c30ab63
51387 F20101110_AACIMO freeman_j_Page_37.jpg
50b75d76f0007dd1f15134721cb48077
44f62669a9b002af080bd93f66953102a41c3dff
20409 F20101110_AACIWJ freeman_j_Page_42.pro
3a1a9340e860a255639c3671dc697efa
236ddd581883c196d1809c1ed9397c0d7c5f2b6d
6468 F20101110_AACJDH freeman_j_Page_28thm.jpg
6538a96a642f010d628af4ae975730ec
2d71bab2943c6ee737a5c5b43ce7da0b70906e6d
1051962 F20101110_AACIRN freeman_j_Page_83.jp2
fc05dfc428a8fea35588f6bc3defe124
dfb94148d86513beeb10e51f145349a930752c20
24526 F20101110_AACIMP freeman_j_Page_38.jpg
abba5f61409ed1c4d7cc24274832e0f3
9ac4734aa1b8efe49db4ed858b65b4a6dceb0554
14921 F20101110_AACIWK freeman_j_Page_44.pro
5a3ac8005571ca18ec1ac6f9c322cfbb
231643932f8ce23f2ec37f9778e3e2dce6ce4b26
25628 F20101110_AACJDI freeman_j_Page_29.QC.jpg
4a8bfe345a0f6f36b921178e750dac31
66454359394cfb66c314b5d892cfd310d9638d2e
48825 F20101110_AACIMQ freeman_j_Page_39.jpg
1dd4c9ccc9ee62223ea666160a29aad2
83b020b89caa4acb57fcfde9f1cab2f407d007d0
14766 F20101110_AACIWL freeman_j_Page_45.pro
6e6c68a46081f75f9f27b83305aef086
9abf1701442d991acb30d44a5316e589ed78ef99
6209 F20101110_AACJDJ freeman_j_Page_29thm.jpg
eeca609b5f50e979612e4f09c5d1841e
51dc3fb42d5ea183979b6aa23fc8892b831a3214
221236 F20101110_AACIRO freeman_j_Page_84.jp2
378839b1bb04b6b74bc200d22ffaa4d4
29ff1f65f44980419d8de4e9f93df64f164dfc5d
18331 F20101110_AACIMR freeman_j_Page_40.jpg
cf6a0c04915f046861aa2aeb19681a14
aa226a9a6938b01f06d2e752cf4734c8eb71b51f
20919 F20101110_AACIWM freeman_j_Page_46.pro
f1551a419caba81faeca75f0df5f7f02
cd42fa8379a20d9e3ee9b899bc6c3a2ee885a366
28706 F20101110_AACJDK freeman_j_Page_30.QC.jpg
a56f65111dd234572a71e1c846aab2c8
206876c0b82af99fa8596d44903a7fae96f4e323
266588 F20101110_AACIRP freeman_j_Page_85.jp2
9337661d24ba8ee472ead778105f51f8
bc91c86b2383504026f91f50e5f1aa5dda2a0344
23520 F20101110_AACIMS freeman_j_Page_41.jpg
3fd9fdb106a1424bbbfc07ca59001d2f
6d31829ea25a7cb29a737bab099370ca1de6c44a
20094 F20101110_AACIWN freeman_j_Page_48.pro
65e14ca4b704310b6a62327f4ae59e7c
8e6d7f06ccc4c44e1529b6d52ccb8195b7085866
6482 F20101110_AACJDL freeman_j_Page_30thm.jpg
a4db0cdf8af832c708caf88b3881c327
5e14875b70e63d6a9490913072afed12bfedec95
32486 F20101110_AACIMT freeman_j_Page_42.jpg
8c170cf6334536c8d3161730632d19ad
7ffba1fc8a28a6a486bdc65a6ea1fcaadefea93e
10312 F20101110_AACIWO freeman_j_Page_49.pro
a5ec87a58627faac0e5105aeae861795
129df954669862d93717e16c6e0f3cee80b34418
F20101110_AACIRQ freeman_j_Page_01.tif
71777116f6dc82b6201e12d4808884eb
eacff02996d84b05d2ab49b580dcddbd95882fbd
26583 F20101110_AACJDM freeman_j_Page_31.QC.jpg
2054cee780a5a15623a30d93d032d19c
2e26a615e95cec9efcb000299841eab58f5d82b6
28225 F20101110_AACIMU freeman_j_Page_43.jpg
e1dc5e4f93dad01d9fa951c23c49dc15
10c6b8908320fef0caee78b4ccfc6dc62e1d2c76
53988 F20101110_AACIWP freeman_j_Page_51.pro
078752d56fbfb4d2cd878acca76951ea
6495dc489e83c39bfd61899057c98573e86d0419
F20101110_AACIRR freeman_j_Page_02.tif
bec0cadb0af38f50fada6c7273dadf43
6e387bf6d7dd53e6670ddfc5bb8f117c1b86bb70
6435 F20101110_AACJDN freeman_j_Page_31thm.jpg
ea464ab8577cf647770e7038588b282e
25c41e89e13001067fb09e2465b22a556e29eee0
26726 F20101110_AACIMV freeman_j_Page_44.jpg
968361d1da3917e92422ef152b3386bd
754823a0029bce76a0f57591549ed10aa4a72b1b
55411 F20101110_AACIWQ freeman_j_Page_52.pro
8994f342e7c8f38529aa677126bf952a
92453e3e8c86442ec4c1566e9896a89b80202b49
F20101110_AACIRS freeman_j_Page_03.tif
c0d4c7cf4919ffb16d763b456b6fd37f
bc2aeaea830c0a86615e372fff5633e1dbd9dfbb
28025 F20101110_AACJDO freeman_j_Page_32.QC.jpg
9ecc868e3955190e48060c79c152ae66
61ea214c0a39c757aaf0433dcef2e7cfe75ae05f
25828 F20101110_AACIMW freeman_j_Page_45.jpg
451a8bc4cec6da83023f93f5822eb950
3b6be7ad4f611af798c3597891d01174352421b4
58186 F20101110_AACIWR freeman_j_Page_53.pro
7d51403ed9819b9936454cc797c6bfd4
9021de1cedca50cfb3a54f9c59357e03a5578ebc
F20101110_AACIRT freeman_j_Page_04.tif
94ebc644cda74776c2d039dbca2c4f0b
35c833814b7bd58ea60a520abe67678f51bf7c8a
6710 F20101110_AACJDP freeman_j_Page_32thm.jpg
d2fa8c2eb76698d261c5b39bee2c9773
9215d7412588c5814e5b66632b10c98a19dafba8
43191 F20101110_AACIMX freeman_j_Page_46.jpg
97fb4ceab9c67e5abad39e36d895fc24
504b06d98389ba0f18b077c1530fb1fd4124a34b
52683 F20101110_AACIWS freeman_j_Page_54.pro
b8144cf70a2b7a510f034508b6265e12
d3199a884fe02b8ba6d94b451e0de8d9add796f8
F20101110_AACIRU freeman_j_Page_05.tif
640dc39366eb7f00777f2bad074da378
8e216dc7dc6b21d23376be03e77d933621f6ce79
28715 F20101110_AACJDQ freeman_j_Page_33.QC.jpg
0a51e3f7dfc95392c8b883292ab21847
9655a1e37729a08a00801cbc911480955d964e47
43984 F20101110_AACIMY freeman_j_Page_47.jpg
926a2daee3e5b85c41733fe0ed62c9cd
4d2938fbe04d63e7597e99789a4de05f3edf0e41
F20101110_AACIRV freeman_j_Page_06.tif
0930abcffe526b8b8acd82ce375adcbc
a418c51655120237e9ffd7188059689280082623
6970 F20101110_AACJDR freeman_j_Page_33thm.jpg
b857bec8740373be8ceb0563735e35ad
5c6fe5fd11cb7ef22652b7a2bda65a962332bbdd
45397 F20101110_AACIMZ freeman_j_Page_48.jpg
c0c26cbfe13b4de6806ea545474d9d93
9e19bfa21b02e7333d7f4a2ed4713407da517498
54072 F20101110_AACIWT freeman_j_Page_55.pro
9d0c1838a974c7a8572773dd63f34e1a
1537e86c7de4a6cbfb6b5c9d3d1ed1ed36c64dca
F20101110_AACIRW freeman_j_Page_08.tif
92505c7b9be9738651fb06e4ddd2b925
b5b0bed423bea9367d437575fd2695dd90580b4a
28437 F20101110_AACJDS freeman_j_Page_34.QC.jpg
144495bf889629cb6706d36ce4a93037
8584ce57b2164c54df61e15e71bcbda2d15f9f2b
54986 F20101110_AACIWU freeman_j_Page_56.pro
00a053e583b0fe82afb350761bfd7dee
ecb79b0682b7e6284c8a28f1d0f1841c0bba1caf
F20101110_AACIRX freeman_j_Page_09.tif
39cb5f6806a8da4f52908f305a3805f6
df6dabb07bb65315e1182624e78e738a8e5da8af
6901 F20101110_AACJDT freeman_j_Page_34thm.jpg
5398feab9ffcadb707087af77ff70786
a0ebb4f633417a39f57bb343c8beae038720d59c
51493 F20101110_AACIWV freeman_j_Page_57.pro
0d7cfaeddaafaec69e2e2f96b74dfdde
86cb36f7b405b31168862049304e482d14e322dd
1051931 F20101110_AACIPA freeman_j_Page_17.jp2
fc99401b4a6b89fc0d19e06dd898b8f1
1df89953d79d4aa6ace83f6c50dca68c9877d3fa
F20101110_AACIRY freeman_j_Page_10.tif
48772cefba388286a2d2252bd92069a8
2303fe95753718c91920cc06ed83e174dce3a89a
28683 F20101110_AACJDU freeman_j_Page_35.QC.jpg
266931276ab1d31385405c4bab081c8b
bcb370fcba20a9729e6213033262503dcedf9c58
53989 F20101110_AACIWW freeman_j_Page_58.pro
55a630816e2413bd49b843522d9b9fcb
372089c13f0ca0659c95794eed62fb2f609c7122
1051913 F20101110_AACIPB freeman_j_Page_18.jp2
d9b26db44ba5a50a06bba1ddddedeb62
05bbc649970fd7bd32eb9f2cbe9993cf2ae87386
F20101110_AACIRZ freeman_j_Page_11.tif
53414f1ebe1e186bfb6b81d6776dc176
f8f3970d1ce98c376931a5c96fbd04c921581792
6900 F20101110_AACJDV freeman_j_Page_35thm.jpg
471e496211a3d1d773fc8b7ca963db99
440232a7a3210dcf0d4f8c80b833084e46be6649
52598 F20101110_AACIWX freeman_j_Page_59.pro
bf2a0742e73548040f54682c949761cc
b8654ede9aebf08cc4cb6177461099640fdc0465
F20101110_AACIPC freeman_j_Page_19.jp2
f453cd8fb923f40fe20f6255b3ed78d5
1d932fa994c75108061c31b72ae262e1dc351d50
4784 F20101110_AACJDW freeman_j_Page_36.QC.jpg
55c4725a9a7f7f3ca1b8fde3a56a022b
96c114c64f5518fd618759585922d8378a30800f
F20101110_AACIUA freeman_j_Page_66.tif
e1d06d44e9d3844536acac3896ddeaf0
758070769a110c18e8b85b941a8b7a7b5ce1cbef
54525 F20101110_AACIWY freeman_j_Page_60.pro
4a2ecc781c920623f46eb82276e0d275
afba8213be3f4cda681eafb1ba57cb81db8e66e0
F20101110_AACIPD freeman_j_Page_20.jp2
37116c0398c5ab2128757161d5b3c92a
db57409d1b34b9a548840072b6d6a5e777696666
1262 F20101110_AACJDX freeman_j_Page_36thm.jpg
7287272ddda970d1140330b32222b91c
7890cf3c45bfbc0fae941ad34c47b4910a3c3826
F20101110_AACIUB freeman_j_Page_67.tif
4affad063d5071579381f1859b934073
949ee5eda0c0389dda1585f7eae21c45ce923f76
48935 F20101110_AACIWZ freeman_j_Page_61.pro
55cbaf3c2fd1f9e0651379213991b415
dd04726ce72e614bf013fa42256c7ed82684e3e2
F20101110_AACIPE freeman_j_Page_21.jp2
66f0fe380834d0f55d744d903bb7ce6f
a45ab392029e0054a4a0c437ec24fa9580d25be6
14397 F20101110_AACJDY freeman_j_Page_37.QC.jpg
8d3f1a838e7ffb01f4cc696294f2d933
6c13e7cd92f788bfe3449aae635ed84337d4239a
F20101110_AACIUC freeman_j_Page_68.tif
1b616b0b6649e1b10872de6767254c88
0785296ce59d1c3b0a2d3db20e2b3ca343feef45
F20101110_AACIPF freeman_j_Page_22.jp2
dec052d28a306624227874bb8dfe7a48
ba1f90c4c41389ac4b2da278be4f6c2720586e5e
3312 F20101110_AACJDZ freeman_j_Page_37thm.jpg
85b0090cf162a350ed23a1b5bcb5a4a5
dbce12308d1c40e53049f99ea86a5b0ab97e08ca
2520 F20101110_AACJBA freeman_j_Page_82.txt
0779001162cad9107a1af47eea949f73
dcfd6a33c24a8d7d3f23e80077f18a6ff4a7fcc3
F20101110_AACIUD freeman_j_Page_69.tif
44836f22b60d7dd819a93edf9a42519f
e038085dfa142f91d68258604a9c1cdc9fa85422
F20101110_AACIPG freeman_j_Page_23.jp2
791ab6d7056b1bb7170fbef6a098445b
e5fcd172fc2220054ecb503722a93369f85cc7f6
1899 F20101110_AACIZA freeman_j_Page_29.txt
03e7696e471941d9779b2739173759c5
eb388444e0201372ed5af4d494f7747523cf04bd
2681 F20101110_AACJBB freeman_j_Page_83.txt
19976d74831d960ff35d47f85f290e33
f1a391f52822d5938882d0c50d234cea8bf4d519
F20101110_AACIUE freeman_j_Page_70.tif
6b50d9c6081246800cdcc22f903db0fb
240d80feda7d07242c9ca2b63dc346fbc2d8c124
F20101110_AACIPH freeman_j_Page_24.jp2
0abc02807d8e60765c63971931a07f7e
66d3d9769afe51fc7cb0d0ecf0d7799fc00c7c8f
2176 F20101110_AACIZB freeman_j_Page_30.txt
8192b890e2001bb222de8a569fd1e1dd
b5350305372994cadf8730045c4a94e81dea8bfd
6887 F20101110_AACJGA freeman_j_Page_64thm.jpg
c57f2c026527267d06ae96152aabd149
5eaf5be70d767859a1b2f00dc68ffab6a457cf27
383 F20101110_AACJBC freeman_j_Page_84.txt
4feab284a8a4bd42cb3ba522c3b9229b
f89da3ea0474b2aead11de6933a13b05137119fa
F20101110_AACIUF freeman_j_Page_71.tif
116cff5d9c24e7c9a0f70a06741a6f24
3e007c700ba00ebe9a120220285993c1664b750f
1051975 F20101110_AACIPI freeman_j_Page_25.jp2
19bad14003bbd42834bc3454921f0c2b
4e9d46acaab79351793c3792ff65e2f21d495370
2048 F20101110_AACIZC freeman_j_Page_31.txt
706f8ad3928501d52ad677a36efe59cd
2334b7ac3fb033f62d941a75ca46639ec56acf43
25820 F20101110_AACJGB freeman_j_Page_65.QC.jpg
1c1328d02de939ded692eef5a543bf5a
a2a8de36211aaecc0c22ac0f0d8b5ffc8ee77bf5
491 F20101110_AACJBD freeman_j_Page_85.txt
82851fffad0bdf77e5c22e721d3d1dfa
f9d0c5ec1b284310161ca45f9ff360cb7253ae6a
F20101110_AACIUG freeman_j_Page_72.tif
86294308cae723bde718a97f60e079eb
cbf99e8d8f359a739c9e932b4ba3cd2ca7158ece
1051947 F20101110_AACIPJ freeman_j_Page_26.jp2
a7f9c2d562ebf726d4a68b89f83ab3c7
aee7894410ae98376714aa005ff8157bb488cb1f
2167 F20101110_AACIZD freeman_j_Page_32.txt
7b0c475442e15e19fdc9e277064ebfbe
9496eb75f9d2c4a64497f0379d0bf56d0c33f7b0
388697 F20101110_AACJBE freeman_j.pdf
4cd427a92461da8af71735c8aa54cdab
d9b675751a1efd3229f70b6a9b6efffc521b96cc
F20101110_AACIUH freeman_j_Page_73.tif
d60197d83557265d806ba56d385fe9fc
2b8a670d455a731cd7329e616f4c121e9d62e5e4
F20101110_AACIPK freeman_j_Page_27.jp2
289ff5f7dbf46e8a254c79b1147ce839
27b87b0726bb863edb5a5496d319954a38163528
2206 F20101110_AACIZE freeman_j_Page_33.txt
f0b4a8149fc3ea7508932168a8cb9ef8
36602969195dce14745ec3e9417a309238318ba6
5968 F20101110_AACJGC freeman_j_Page_65thm.jpg
66dfc81d21d1ebc20cdec10c2ed81ce5
db1e1313c134e72dc818c62a0d1db13e0800154b
1582 F20101110_AACJBF freeman_j_Page_01thm.jpg
6e6ab134b11fc33805fe251324263756
9fa1bc9dd925568dd63c70c87b38ff80201a43e2
F20101110_AACIUI freeman_j_Page_74.tif
ea550032af7936a717568da403bef66d
bd9f942caa3e31a2e20b3dc4eee5095632b699e9
1051943 F20101110_AACIPL freeman_j_Page_28.jp2
b891d4dbf7ee3397c12bcc0397175cc7
b74fc3742367acce47d5ccd7508bcf91e035ef26
2194 F20101110_AACIZF freeman_j_Page_34.txt
2bd4e1ae064f0b67f50781e6d4ff6b71
591233a1b30830f20317c000a4445ac84b0db502
11177 F20101110_AACJGD freeman_j_Page_66.QC.jpg
73ceb7f98c20c5ed53e258efc4fb87c2
d3f868bf719902ee678c7802bc17486ae23fedf8
6651 F20101110_AACJBG freeman_j_Page_01.QC.jpg
f5c5f9f1ddd756a0dd3c3ee4884bf08e
e586b8c5bb227325104dd6662a168fb0f5afa03e
F20101110_AACIUJ freeman_j_Page_75.tif
17a20e8a36251210d02bf5346715cfad
3793cdb8021a3e83308b2311b93dcfe26183d731
2237 F20101110_AACIZG freeman_j_Page_35.txt
3a225890081095bc0c050f089a345d25
e46e22279960c3707419453fd768ffd1cdaa8aff
2685 F20101110_AACJGE freeman_j_Page_66thm.jpg
07609077eec609da257f6e3dcd20a02f
b30b52d1b9ac86702326fe3be9e20e798723e6fb
1268 F20101110_AACJBH freeman_j_Page_02.QC.jpg
09a3332d73494227d20cb017772a960c
9e7541487df71765d8c146c90ef37e610474a1f5
F20101110_AACIUK freeman_j_Page_76.tif
76118091ad3d5caf5c3766a863306586
ffc979925f907babb2c8a541870c36a5a52dccb1
1051954 F20101110_AACIPM freeman_j_Page_29.jp2
46dcb2cfacd51fe90fa3af03a029c7dc
062fb9c1a4653eb6c6c835017f132e90d4a7c9db
271 F20101110_AACIZH freeman_j_Page_36.txt
cc911777993e105d52aa02d1037d2d15
f8278e16e646c3c196911ab67cfea631e083a0dc
6450 F20101110_AACJGF freeman_j_Page_67.QC.jpg
e062d1843b185f407048b16095d4959a
1de6201d23220620641b6823531324981bdfeec1
474 F20101110_AACJBI freeman_j_Page_02thm.jpg
d3d263396e01cdf94d43d54a2fe2112a
6beac9eb3d3d07bd51444405bf345eff7234d5b4
F20101110_AACIUL freeman_j_Page_77.tif
057222f2fef9e2af15513e4d1dbb6e0d
df8583815454eaf8ab6c3180ba0c68dca15ad27c
F20101110_AACIPN freeman_j_Page_30.jp2
6514017ffbfeb5a5cea643033b2de178
f80470942bbb42ba3c660fcdc70e77f5684db6f3
1487 F20101110_AACIZI freeman_j_Page_37.txt
a50577f44c44ca51aa9adb46fc7beec3
4fc5fbfea9c4c2dd4046da3187c43453f3dc85c5
12986 F20101110_AACIKQ freeman_j_Page_43.pro
79969cad167bfa8d634ea2b8cf0084d0
ca22eb0e5ed220da5233a4f60c4e30f3b775fd2a
F20101110_AACJGG freeman_j_Page_67thm.jpg
12d1ad11ab04d63232c03b684393e1a6
dde21d5bfb342e4275e83de90388025478c3056a
18456 F20101110_AACJBJ freeman_j_Page_03.QC.jpg
87ada128ae020a319895f7123ad48828
fa807eaa5ab11ce5810afd54f8fd0b4b1a05b3b4
F20101110_AACIUM freeman_j_Page_78.tif
be5eaebcab8a3655fdd4a57c9d759397
3e9b3158f744b0a227d8db97427643e19c0cf80f
1051969 F20101110_AACIPO freeman_j_Page_31.jp2
6baa572c0db8be6b951142b4351f86d6
4f7715d4711162ec2410e0c917d7867da244b84e
1078 F20101110_AACIZJ freeman_j_Page_38.txt
20129c593a0630bbf0200d6b8a8a4ff3
caba59f3d2de09647ed787e2853c7858cf344306
F20101110_AACIKR freeman_j_Page_64.jp2
fe1979cdd1a882a695fd6f8f19d39d65
d1a010ccbc8a96c2d807aa2b2c51e4793256f407
6767 F20101110_AACJGH freeman_j_Page_68.QC.jpg
db21f66fb9328b948c2909708dd4d620
62d31b1697baadfb6edad2bb546f62511b9b5eb3
4404 F20101110_AACJBK freeman_j_Page_03thm.jpg
8a7439ce56ba9a6448acff821a052145
0a02bc110d56e8f927a670a9b7fd592ebec03ad0
F20101110_AACIPP freeman_j_Page_32.jp2
98398c446915672dfe6eb047530c3d83
9576c0b924b2fa40ac169bd2f9f2e0e514cf0a3e
1494 F20101110_AACIZK freeman_j_Page_39.txt
9308732699091b34670ab6aab2601a65
899a0c41c919a347dba24613513334dcd8c0564e
F20101110_AACIKS freeman_j_Page_56.tif
a787b91a005824049d74dd05d61dfad5
2f2589e391b78feb0b905d22e5862dfdcff4f813
F20101110_AACIUN freeman_j_Page_79.tif
8e69fbea1b5d84da2106dbdf7fd03ad8
1e37fbe0c69c67b546f6e6edd387b14755816ecb
2227 F20101110_AACJGI freeman_j_Page_68thm.jpg
65c77cbc2bb357ba3fb4b7efce3ac393
c9227aaaf20dde2c966a1023de4eb5cb598fabe5
16290 F20101110_AACJBL freeman_j_Page_04.QC.jpg
65ddc26af769120d52ccc12b4c3f1f27
d07b3d1bb56f729ba3679ce287bdd400aefa3a83
F20101110_AACIPQ freeman_j_Page_33.jp2
e92fe0a5ceae1900353becff645e698a
ec19ba8a5101e6747faa12d3eed9cb0e8cb95171
480 F20101110_AACIZL freeman_j_Page_40.txt
b4f74d425b7483c3751b92be16625506
216fac2f93d6c4d3182150cf9194c9e566719655
14214 F20101110_AACIKT freeman_j_Page_50.pro
b38e6d8c3f591ac1021b99c54b082f14
28003fdcfa8626ed3aa3b2d6c9f2e4ffcbfaa91e
F20101110_AACIUO freeman_j_Page_80.tif
0ae8b4f5e8d34ec3f4455a082b7b24ef
4a36ada17b8111ec402edd6ce61eaf2b4bdefd86
6022 F20101110_AACJGJ freeman_j_Page_69.QC.jpg
659ba759d2459c6964d962b3853a003d
0abf1936616a8335d0210bdc6066998245465c69
4091 F20101110_AACJBM freeman_j_Page_04thm.jpg
748ade914b05ab00213d0dec18b415fd
19f41ef99752ab797d7f0ee6051fb594b0cdd953
1051950 F20101110_AACIPR freeman_j_Page_34.jp2
882b200c3675105f3e5e3ced93dfbb3e
21881d4283a023da062fe018d05c0535f449a8fd
743 F20101110_AACIZM freeman_j_Page_41.txt
04f649ff021f240e3e351cc621cc5ffa
f17c0569bb2670bad2ddae9ebc97c431586f8e6d
2180 F20101110_AACIKU freeman_j_Page_52.txt
d54a27a7b2068db7d01796bf39745574
c3e6c3827d69540d4f51a1633ff486845bd46217
F20101110_AACIUP freeman_j_Page_81.tif
d01e1a6e95c58da44d52d634f4f44105
1c91a47020413a1e7e0f9939cc15f0950e0ec143
1945 F20101110_AACJGK freeman_j_Page_69thm.jpg
f7c2016f8fa27ab41c9e9c4448cb05c9
884e6727e8eeb9cbcdb577d903d4bffbe83a2b41
6831 F20101110_AACJBN freeman_j_Page_05.QC.jpg
b1045433133bed2f7066c6afac28f0a5
ce519567c4cce460343cab1a809ea0a49cfc38fe
F20101110_AACIPS freeman_j_Page_35.jp2
9e31225e0f3083a5c04136f843da379a
c33c207a5eed128e9bfbb1ccd43554286b80688a
1308 F20101110_AACIZN freeman_j_Page_42.txt
a7a46a0ac2f8d3784e425a85fe5b1c3d
4a37455d9d5e2bc5b1e441daf737e0fc9110e6e0
20774 F20101110_AACIKV freeman_j_Page_47.pro
f0034c9bd91cb9471cf48e8c8dd85602
d92b209ada5d07158edcd3e44a07bd39aeb37147
F20101110_AACIUQ freeman_j_Page_82.tif
eccdde7c439ec55552b47f3878e64a29
8afe6b7faee257e4feef0688e9484555d384730a
5907 F20101110_AACJGL freeman_j_Page_70.QC.jpg
e5d493f71a973732b5c580ce35436359
f97bbe9b4aa1dc8d97eab6fa4ad9d4b5615bb5b4
1964 F20101110_AACJBO freeman_j_Page_05thm.jpg
9ae80dec2f2055b073908bdb798420b6
e7e10dd253827e3896729ddfb3ea7cd26a3252d4
157661 F20101110_AACIPT freeman_j_Page_36.jp2
19798773d4943a72cd3dca2890daca01
371820bbe15bb39785de91383b7b498b9c076191
684 F20101110_AACIZO freeman_j_Page_43.txt
ba70891b4418f3312062403ae0c4894a
0f0948e49550e29d11cbc235ae2243d3ddce67e2
F20101110_AACIKW freeman_j_Page_39.tif
e3dc8dad7c44c8304253e2603346f32e
beaa641e74da70f9b87c6746549a5136ef8c9d08
1836 F20101110_AACJGM freeman_j_Page_70thm.jpg
ee99ed8fc58676de2b663d7d11de94c0
68c3d65bbbecfb90f8d26b98f200e355753fdd9c
6274 F20101110_AACJBP freeman_j_Page_06.QC.jpg
0483dca2398a61e9248f77580a8a1768
e4d02573f8469884b61d2d017811f74d2519d151
630187 F20101110_AACIPU freeman_j_Page_37.jp2
d3319b987fdad2860e951534d73c44f5
df3242d6d5d46d7767cdfeede6562314cac2411a
1017 F20101110_AACIZP freeman_j_Page_44.txt
5bd6f7f140699857c7784d55302bae7a
cf3c29181b76bdd96e9960100898ee816ee9cc92
13168 F20101110_AACIKX freeman_j_Page_47.QC.jpg
5572fb3d09da9ee616408019aec02b6f
fee9eabbcb5e3ef29b6b14853487fedcdd8e9c51
F20101110_AACIUR freeman_j_Page_83.tif
22c73f0f8baae7c872bcdee830cfbbc0
929c6c0a607b52e19bdc646c09c9e9b454659f7a
9507 F20101110_AACJGN freeman_j_Page_71.QC.jpg
7fbd241c70630e33845adb6540b48cf8
dafc0164f290982173767b003f7616617c86109c
1830 F20101110_AACJBQ freeman_j_Page_06thm.jpg
73d7c1a565829033388f321bf69ae70d
7ad82f180f0752ffb7c7164f142421e45f39daae
302821 F20101110_AACIPV freeman_j_Page_38.jp2
dc9c6257e6f76e72c27f7d8122d6df94
2427ef7f8aa18eab2b7aff20fc98cc704d90ba8f
1129 F20101110_AACIZQ freeman_j_Page_45.txt
24cc76fd937f97ce7c741acffebbc925
38d14e5caed1f1241249182aceba898bd04cca59
21919 F20101110_AACIKY freeman_j_Page_68.jpg
21a80598037b205948e4972c7a9041c8
caa010e0e935fe6121c8855ba7bf8c4ac42cf9ed
F20101110_AACIUS freeman_j_Page_84.tif
e50ab8be978fd01af828c4e13772fe51
32c6196b65a5469929f1a1af38dc5129a83505a2
3100 F20101110_AACJGO freeman_j_Page_71thm.jpg
dd7a091fcdcd9fb3dc89942ba6663aef
ada0b23f16361004525fb42c5d86cea60945508d
12840 F20101110_AACJBR freeman_j_Page_07.QC.jpg
3a22cd3097b81c7cb053700f603cf191
1e578b69277d90a47180971f70b11a2f93c089cd
638487 F20101110_AACIPW freeman_j_Page_39.jp2
36f45ec7e76eca7a4ffc45f714a213c8
cc19930ddc05f89339599c51be35b793e64b5b4e
F20101110_AACIZR freeman_j_Page_46.txt
cb7840a64dc5cc234cf5487af2bcd1f4
641b146c17e62ace97e8549bf71f405a4384bdf8
6102 F20101110_AACIKZ freeman_j_Page_15thm.jpg
67cb39e7061efaf3817d11db93d52861
281c9c9893890176fe759cc1418fca24f40b800c
F20101110_AACIUT freeman_j_Page_85.tif
4ac78003decf6ae1d20494a0fea66d42
a5421331df43c1bdfa6b650590897c83df65d474
27984 F20101110_AACJGP freeman_j_Page_72.QC.jpg
c08bf9e7fab9cf67387a6828a9de69fa
60503e2a85d5e12f3ee77d944cc4d62bb6a57ecd
3636 F20101110_AACJBS freeman_j_Page_07thm.jpg
cb6ccb5211a4251dbf93e45871ac84f9
188f83064a9f8a4bbbf3e79dcc1a797beacde6c0
212942 F20101110_AACIPX freeman_j_Page_40.jp2
3795d751c86afb69b7f27d5f7ac3bcb3
5697ce728bc6e0285bfabd63b5b9d44e35b44c0d
989 F20101110_AACIZS freeman_j_Page_47.txt
a40a3834c787d43eb620ce39fb220c51
98eb8d93f9be30e220c7ddb75fa3e60b9f2157e4
8885 F20101110_AACIUU freeman_j_Page_01.pro
37f978d60b5659ffe3e31f89bc082721
4ff42be4b984f2940617a038f68928c6a4f84920
6347 F20101110_AACJGQ freeman_j_Page_72thm.jpg
2e4f1b89d346ccce363c893f8e9f6697
9052881e6e4a7157af8720207434220c378116bc
24193 F20101110_AACJBT freeman_j_Page_08.QC.jpg
6486cf51305dfe91fffb1142f12944fa
03898160c5b0a83a95408ffe5f5bbd4bb9714c76
290854 F20101110_AACIPY freeman_j_Page_41.jp2
09e4aa03a4c4b110117969c3b38587ac
c59b6fa4c525283700ab287c6b84f394947de27f
980 F20101110_AACIZT freeman_j_Page_48.txt
0f1540138d9015385f33d0c2030a6762
5fd943a64c39d3c3889f69944f29bfec4d560923
861 F20101110_AACIUV freeman_j_Page_02.pro
a5a19f14ac075a42335946a1727f4ef7
637fe2bdb74e844629ba293c94bb976eab2d598c
29995 F20101110_AACINA freeman_j_Page_49.jpg
f30e5cb7d1f42e67b5e3f372f218fa0e
884209e408f02e172d06cc805647f380373c90c6
28341 F20101110_AACJGR freeman_j_Page_73.QC.jpg
5380164d1fe70ea82153a5fc5282c624
0c3cb7e2641635dfe9a77f4bc0e354929db79bf0
5986 F20101110_AACJBU freeman_j_Page_08thm.jpg
7a940998265cb4bb5ad62a4bb610f0c9
9a21f830875090daab99b47960c92e2bd5d86fbd
431251 F20101110_AACIPZ freeman_j_Page_42.jp2
ed7e5928036cee59ec7fb0246d23b9ac
97e53235c70327f76b79b117e612727929db4568
635 F20101110_AACIZU freeman_j_Page_49.txt
72e1a2f8689190536a887006f3b519e6
546513bb93047e81d314960254bf63838ec485e6
34231 F20101110_AACIUW freeman_j_Page_03.pro
2fe60c6246de93b52aad665d4849a053
4b1eeeb35414fcef3a3316e167fb31065c3c786d
30222 F20101110_AACINB freeman_j_Page_50.jpg
164207109b57400b3f01d9adb9779b16
96e0cff434651fcd79eae7853c3de648fa02f972
6850 F20101110_AACJGS freeman_j_Page_73thm.jpg
6119e05468979fb8abc21149ddec1bf9
2f8e00552260927e285683e060b628eb41c8d116
15458 F20101110_AACJBV freeman_j_Page_09.QC.jpg
309b1627cc70a37b3d2115f2c866b814
295bd4bd5245aab624af658de37c9b733c04554c
837 F20101110_AACIZV freeman_j_Page_50.txt
7fdea0b610ef22e2d8b52b559e785544
67b560d9acf0c5d5b492d284935d6e99a6884681
84381 F20101110_AACIUX freeman_j_Page_04.pro
39dc9952dd1ed44104f149b6ea840b34
c07e801df8577032281b0654e532e7d95225a6f2
90737 F20101110_AACINC freeman_j_Page_51.jpg
7f4518df4487bb71dcdeee9dabd83a42
a39eeae089db47b4b368cf57f26a7c5038cadb9f
27446 F20101110_AACJGT freeman_j_Page_74.QC.jpg
8acaac950fb261d91e5020ed3795717a
66e661ded41c20a10489336025a1e98172735ff5
3522 F20101110_AACJBW freeman_j_Page_09thm.jpg
92af9e2c375487e83b9199e5e41b915b
5cd6adddd1f1a1f9c5eddabaa957f7545eb836f3
F20101110_AACISA freeman_j_Page_12.tif
210ca72dd059174212afc3c5e3f00aaf
8f2515bcc991aba85811a748923083f9b1812081
31922 F20101110_AACIUY freeman_j_Page_05.pro
119e551fe68aff3b0ab24bf79a4e0d92
41c4b01b525327b04000cea201a6f99e45304651
90215 F20101110_AACIND freeman_j_Page_52.jpg
2df4772496c4e08847e899ade7f155c8
c3956a96b22ccb3363906153dac99438eb0db61a
6572 F20101110_AACJGU freeman_j_Page_74thm.jpg
01a33e16a5a906f52e248a16b11ae950
d38ebc35144d24a80c849e811259de6d7df98d35
27935 F20101110_AACJBX freeman_j_Page_10.QC.jpg
b07e92d675daa582080c21429e0b9838
014c3b58db634ebe3accea9a94ba2c6c0cffe36f
2207 F20101110_AACIZW freeman_j_Page_51.txt
1dff47209932e01eca43980c6f0a7467
9677e8f161bd8d1c89e5be220648211bcb3cd6bc
F20101110_AACISB freeman_j_Page_13.tif
8a28aa42a77418275bc0e220e869c814
9afec3e0c4b9ac385e23e6553e28eee4d926ffb7
17515 F20101110_AACIUZ freeman_j_Page_06.pro
4404c2b1d35e7110c39bb9b7c901ebbc
1259de62fe6cba6ab3d2230ec160e099f8878fdd
95086 F20101110_AACINE freeman_j_Page_53.jpg
c3c18327dbfb85698594408240c9dc61
94fecf61e6e594940da3cf1ec30da76eb77f30c6
3005 F20101110_AACJGV freeman_j_Page_75.QC.jpg
0840b145095efa337e6c0b6c1782f4a7
ecf1f1991e42f07062fcd2bccb4f10babf2721d5
6399 F20101110_AACJBY freeman_j_Page_10thm.jpg
543f2a58cedc9ba3a6f70a4785788609
7491140986bcf1f476339b5ab04fd0e1856fcb83
2269 F20101110_AACIZX freeman_j_Page_53.txt
1da9cc8a8ed8c2f26d3d81dff84a92e7
e10f5d69ffa1da2f0a13bc245e3fc687baf772b9
F20101110_AACISC freeman_j_Page_14.tif
2807c5dfd8abd898128ed1a681121dff
2cee657122d9973330128d28d41033df27b2dc51
85814 F20101110_AACINF freeman_j_Page_54.jpg
59a9bc78325c5ed4a15a41a3d0bd159a
2709bced09e8bd04625222adbf9527d763bd46bf
839 F20101110_AACJGW freeman_j_Page_75thm.jpg
a34b4b81b15549414c518b5a712d4c6e
abb6cdcd1e35ad89009cf33f38af7dd4ac85e109
27883 F20101110_AACJBZ freeman_j_Page_11.QC.jpg
1d82daa19225ca51b0f95ff280197ea9
a579224c6eea7a00b9614ddc0b39108afed7264b
51741 F20101110_AACIXA freeman_j_Page_62.pro
03912ac6d026694a9c3e9ce5e2ba6233
dfba972d8d9718875999e80dd212b58b1cbb09c4
2072 F20101110_AACIZY freeman_j_Page_54.txt
2d601c70cbf25719e8c3f4d96205f424
42adafa48372459cf3274031601f4617b3ec3e47
F20101110_AACISD freeman_j_Page_15.tif
7b4a2d119388ab2819a47122fc2d321f
4fa8b107099fc60d7bddec023a19397f75dbe9f4
87687 F20101110_AACING freeman_j_Page_55.jpg
b66427627ebcfad054b676bebace1364
965a63e5b35c4c917bea36d37266972a27f17d79
25752 F20101110_AACJGX freeman_j_Page_76.QC.jpg
5071c71626bfa04d66e72eadfe085813
d8f517f22fc5943b8e885b9b68f47b2ba88e6c2a
55987 F20101110_AACIXB freeman_j_Page_63.pro
fb772f4aefe374113b87da33accf0be1
6399fb238d12f0da8f407b33a9fad6276f385ef7
2124 F20101110_AACIZZ freeman_j_Page_55.txt
444073add2d374e23e37b4cfcfed389a
9521fe3dfed10131de3d5f9164546ab7bfb62b4e
F20101110_AACISE freeman_j_Page_16.tif
0dec9698433f9a2c2c4974d52e1ade1f
5cb61454bd73ca33051025da24a3696e309f6617
91573 F20101110_AACINH freeman_j_Page_56.jpg
9584cb65e6392ae8f687c78c7cbc3620
5171bd4c6e2133dc064ec4dc9688e1c784befafe
5980 F20101110_AACJGY freeman_j_Page_76thm.jpg
c82774d26905579bba4699052bd0d3fc
87e7493b89c31d6ad3c3ac54cfa4269d72e6b40d
55639 F20101110_AACIXC freeman_j_Page_64.pro
6233ce9d40259b3cd784e6cad209bc63
97180035b7985bc306a34716e985361474d31827
F20101110_AACISF freeman_j_Page_17.tif
92b81cd6920ca751143494bb730f208f
ccbc414863b54285215e5700aab9f1ffed846827
86047 F20101110_AACINI freeman_j_Page_57.jpg
9ae08d9d55aedb0dbd8a7bda62ee8b13
19b0cb3fd61c7d388e89bb9c1751525984b92bf9
22343 F20101110_AACJGZ freeman_j_Page_77.QC.jpg
f1f00ad974c7676234bbb6218534aefd
d497b1881d00248c06431c73a27d9c479498fe85
8072 F20101110_AACJEA freeman_j_Page_38.QC.jpg
0920c09727b5cca54b26420bcda5b0e1
14bb57e2ccdf6b2937b22bc605acdbaa1b6728a1
49001 F20101110_AACIXD freeman_j_Page_65.pro
4400972e81e5201af400fdcd7863011b
c6f39ab31c541eb6369262c133c65b2d704d4294
F20101110_AACISG freeman_j_Page_18.tif
f506fccc0647104af72e19facb15fb9c
b59eb84346ea100d4dc93269baa89151ea75d5c4
89130 F20101110_AACINJ freeman_j_Page_58.jpg
1badba5dfb1d211df3998eb995e4b40b
86d181e4c31895bf589e8271105f03fc65922c05
2177 F20101110_AACJEB freeman_j_Page_38thm.jpg
e8513400614eb5b1ad94183ce8dd852a
ca41fe59a400266457007c1d2a103239847253d0
24889 F20101110_AACIXE freeman_j_Page_66.pro
3b9060378630005a2204f2017ac3fd1f
8b4fec46891653084526bae00609e05a66f1106e
F20101110_AACISH freeman_j_Page_19.tif
3d65d987f6da94f5c498d4e466984991
14109e805d07740c45567791b30387d865859c68
14604 F20101110_AACJEC freeman_j_Page_39.QC.jpg
edd1bd06c3b918ad977737a3d0781c19
b28550db7ecf1d072c073a8fbfa1f7a38ce4b33f
10693 F20101110_AACIXF freeman_j_Page_67.pro
bae2448814cac8f577314ef4f345cab5
f3c6ac3728ad820b4673cace597696a28f87a50f
F20101110_AACISI freeman_j_Page_20.tif
e012a65db7b870c7e69ac5fc928e55d4
0243a49e6b477e77f55e0c23a205933d231fe2a4
87779 F20101110_AACINK freeman_j_Page_59.jpg
0e8f9e6940db9414a9a353693cc9dca1
a15ae04ac6f830f49a14b53cb1c910b4c0d77d41







FROM PINus ELLIOTTII PLANTATION TO PINUS PALUSTRIS ECOSYSTEM:
THE ROLE OF HERBICIDE IN LONGLEAF PINE RESTORATION



















By

JOHANNA E. FREEMAN


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

2008

































2008 Johanna Freeman









ACKNOWLEDGMENTS


I would like to thank Dr. Shibu Jose, my supervisory committee chair, for the opportunity

to conduct this research, and for his cheerful support and advice. I would also like to thank Drs.

Alan Long and Debbie Miller for serving on my supervisory committee, providing instruction on

the identification and ecology of North Florida vegetation, and assisting me in developing fire

ecology research methods. Thanks also go to Dr. Eric Holzmueller for extensive assistance with

field data collection and analysis throughout the project; to Don Hagan, Michael Morgan, and

Kelly Thayer for assisting me with field data collection; and to Brian Hinton for assistance with

fire methods. I thank the University of Florida herbarium staff, Dr. Walter Judd, and especially

Dr. Susan Carr for invaluable help with the identification of plant specimens. I also thank

Meghan Brennan of the IFAS statistical consulting unit, who assisted me greatly with data

analysis. Thanks go to the Florida Division of Forestry for funding this project and providing the

study site, and especially to site administrator Tom Beitzel and his crew for their assistance in

collecting fire temperature data. Finally, I would like to thank my parents and David Kaplan for

their constant support and encouragement.









TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ......... .................. ............................................................................. 3

LIST OF TABLES .................... ........................ ................ ..............

L IS T O F F IG U R E S ........ ................................................................................... .......... ...... 7

A B STR A C T ......... .. ................................... .......................................................................... 8

CHAPTER

1 IN T R O D U C T IO N ........ ............................................................................... ................ .. 10

2 USING HERBICIDE TO RESTORE THE UNDERSTORY FOLLOWING HARVEST .... 13

In tro d u c tio n ............................................................................................................................. 1 3
Methods .................................................................. 18
Study Area ...................................... .................... 18
Treatm ents ..................................... .............. 19
Data Collection and Analysis....................................................... 21
R e su lts ............................................................................................................... . ........... ..... 2 5
S h ru b C o v e r ........................................................................................................ 2 5
H erbaceous C over ..................................................... 25
W iregrass C over............................. .................... 26
Shrub Stem D density ........................................ 26
S h ru b h e ig h t ......... ... ..................................................................................... 2 7
Understory responses ..................................................................... .... ......... ........ ...... ........ 27
Species D diversity ........................................ 28
Species E venness ........................................ 28
C om m unity O ordination ................................29.............................
Species of Special Interest ............. ............. ................. 29
Fire Tem perature.......................... ........... .............. 29
D iscu ssio n ................................30.............................

3 HERBICIDE AND SUSTAINABLE LONGLEAF PINE SILVICULTURE ........................ 51

In tro d u c tio n ............................................................................................................................. 5 1
Methods .................................................................. 57
Study Area ...................................... .................... 57
T re a tm e n ts .................. ... .................................................................................................... 5 8
Data Collection and Analysis....................................................... 60





4









R e su lts ................... ................... ...................6.........1
S e e d lin g S u rv iv al ................................................................... ................................... 6 1
P percent O ut of G rass Stage ................................................................ ....................... 6 1
Growth: Out of Grass Stage Seedlings ........................................ ........................ 62
G row th: G rass Stage Seedlings................................................... ........... .............. 62
D iscu ssio n ................... ................... ...................6.........2

4 SUMMARY AND CONCLUSIONS ....................... ......................... ..................................... 72

APPENDIX

SPECIES LIST............................................ .............. 76

L IST O F R E F E R E N C E S .................................................................................... .................... 78

B IO G R A PH IC A L SK E T C H ......................................................................... ......................... 85









LIST OF TABLES

Table page

2-1 Shrub cover, stem density, and height. ........................................ ........................ 37

2-2 Indicator species analysis............................................................... .................... ...... 38

2-3 Understory species richness, diversity, and evenness.................................................... 39

2-4 Mean maximum fire temperature (C) by treatment......... .................................... 40

3-1 Pine seedling growth after five seasons ................................................................ 66









LIST OF FIGURES

Figure page

2-1 Ericaceae cover .............................. ........ .................. 41

2-2 O overall shrub cover ..................................... ............................. ............ .. 42

2-3 F ire effects on shrub cover .................................................... ...................................... 43

2-4 H erbaceous plant cover..................... ................................................ .......................... 44

2-5 Changes in cover of plant functional groups before and after fire ................................. 45

2-6 W iregrass cover ......... ................................ ...................... .. ......... .. ............ 46

2-7 Shrub stem density ..................................... ................. ............ .............. .. 47

2-8 Shannon D diversity ............................................................... .. .... ..... .. .............. 48

2-9 O ordination of plant com m unities ..................................................................... ........ 49

2-10 Relationship of fire temperature to dominant shrub species.......................................... 50

3-1 P ine seedling survive al ............. .................................................... ............................. 67

3-2 Post-fire pine seedling survival...................................................... .......................... 68

3-3 Rates of grass stage release by treatm ent................................... ...................... .. ........ 69

3-4 Pine seedling height ................................................................... ........ ..... 70

3-5 Stem volume index of grass stage and out of grass stage seedlings............................. 71









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

FROM PINUS ELLIOTTII PLANTATION TO PINUS PALUSTRIS ECOSYSTEM:
THE ROLE OF HERBICIDE IN LONGLEAF PINE RESTORATION

By

Johanna E. Freeman

May 2008
Chair: Shibu Jose
Major: Interdisciplinary Ecology

Under a natural fire regime, the longleaf pine (Pinuspalustris Mill.) ecosystem is

characterized by an open, longleaf pine-dominated canopy and a diverse, grass-dominated

understory. In the years since European colonization, the vast majority of longleaf pine acreage

has been cut over, fire-suppressed, and fragmented, which has resulted in extensive understory

invasion by shrubs and a concurrent decline in herbaceous species diversity. One of the biggest

challenges to successful restoration of this ecosystem is the persistence of shrubs in the

understory, which suppress longleaf pine seedlings as well as native herbaceous plants.

Herbicide can be used as a supplement to fire in order to enhance shrub control, but must be

studied carefully because of the potential for negative impacts on native plants. Information is

lacking about the effects of herbicide on natural longleaf pine flatwoods communities. We used

a banded application of three herbicides and one tank mix as shrub control treatments following

complete removal of the slash pine canopy and replanting with containerized longleaf pine

seedlings in a mesic-wet flatwoods.

The herbicides tested were Arsenal' (imazapyr), Oust' (sulfometuron methyl), Velpar L

(hexazinone), and an Oust' + Velpar L tank mix. Four years after application, no negative

impacts on understory species richness, diversity, evenness, or community composition were









evident in any of the herbicide treatments. Oaks (Quercus spp.), one of the dominant shrub

genera on the study site, were resistant to sulfometuron methyl, and this herbicide was therefore

ineffective both as a pine release treatment and for enhancing herbaceous species growth.

Imazapyr was the most effective treatment overall, significantly improving longleaf pine

seedling growth and also enhancing herbaceous species cover. Both hexazinone and the

hexazinone + sulfometuron methyl tank mix provided some seedling growth and understory

enhancement as well, though these effects were not as pronounced as those in the imazapyr

treatment. Shrubs resprouted aggressively after a dormant-season prescribed fire in the fifth year

after treatment, indicating that herbicide-related increases in cover of wiregrass and other

herbaceous species may be lost in future fire cycles. Retention of overstory pines for shrub

competition and as a source of fine fuels is recommended in addition to herbicide for sustainable

silviculture in longleaf pine flatwoods.









CHAPTER 1
INTRODUCTION


Two hundred years ago, longleaf pine (Pinuspalustris Mill.) ecosystems covered an

estimated 37 million hectares of the southeastern United States (Frost 2006). Based on fuel

accumulation patterns and plant life histories it is believed that light to moderate intensity surface

fires burned throughout the range of the longleaf pine every two to three years prior to European

colonization (Frost 1996). These frequent fires maintained an open, longleaf pine-dominated

canopy and a grass-dominated understory containing some of the most species-rich plant

communities ever documented outside the tropics (Peet and Allard 1993). However, in the years

following European colonization, fires were suppressed and old-growth longleaf pine systems

were logged almost out of existence. Today, it is estimated that only 841,000 hectares of

naturally regenerating longleaf pine systems still remain, the great majority of which are second

growth and fire-suppressed (Frost 2006). This area contains only about 5,095 hectares of old-

growth longleaf pine (Varner and Kush 2004). Fire suppression has resulted in extensive

understory invasion by hardwood shrubs and a concurrent regional decline in herbaceous species

diversity (Walker and Peet 1983, Mehlman 1992, Gilliam and Platt 1999). In 1995, a U.S.

Biological Survey report listed the longleaf pine ecosystem as the third most endangered

ecosystem in the United States (Noss et al. 1995).

In recent years, recognition of the value of the longleaf pine ecosystem has motivated

widespread restoration efforts throughout the southeast. One of the biggest obstacles to

successful longleaf pine ecosystem restoration is the persistence of hardwoods in the understory

even after the reintroduction of fire (Walker and Siletti 2006). In some cases the reintroduction

of fire alone may restore the desired two-layered savanna structure (Kush et al. 1999), but many

factors can limit the efficacy of fire after a long period of suppression, including insufficient fine









fuels, presence of ladder fuels that may cause damage to crowns, and duff accumulation that can

kill overstory trees when ignited (Walker and Siletti 2006). Growth of shrub rhizomes and root

systems during periods of fire suppression can be extensive, making these shrubs recalcitrant

even when fire is reintroduced (Drewa et al. 2006). Additionally, social factors such as

proximity to residential areas or highways can limit the ability of land managers to use

prescribed fire effectively. Therefore, mechanical and herbicidal hardwood removal techniques

are often considered as supplements to prescribed fire (Walker and Siletti 2006). Hardwood

control is also of key importance for restoration projects that include planting of longleaf pine

seedlings, which are very intolerant of competition and can quickly succumb to competing

vegetation while in the grass stage (Harrington 2006).

Herbicide, used in conjunction with prescribed fire, can be an effective tool for reducing

mid- and understory hardwoods in longleaf pine systems without negatively impacting native

understory species (Brockway et al 1998, Brockway and Outcalt 2000, Provencher et al. 2001).

However, most studies of herbicide as a restoration tool in longleaf pine systems have been

conducted on xeric longleaf pine sites (Litt et al. 2001). Relative to xeric longleaf pine sandhills,

mesic and wet longleaf pine flatwoods are more productive, have different soil characteristics,

and are associated with different understory plant communities. Understory responses to

herbicide in these systems are therefore likely to differ from those documented in sandhills (Litt

et al. 2001). At present, there is only limited information about herbicidal shrub control in mesic

to wet longleaf pine ecosystems.

The purpose of this research was to test the effects of different herbicide treatments, used

in conjunction with prescribed fire, on target and nontarget vegetation in a flatwoods ecosystem

undergoing conversion from a planted slash pine (Pinus elliottii Engelm.) overstory to a longleaf









pine canopy. The study included both overstory and understory components and had the

following objectives:

* To determine whether differences in the level of shrub control provided by sulfometuron
methyl (Oust'), hexazinone (Velpar L), a sulfometuron methyl + hexazinone tank mix, and
imazapyr (Arsenal') were evident in the fourth year after application;

* To determine whether differential treatment effects of sulfometuron, hexazinone, a
sulfometuron + hexazinone mix, and imazapyr on native understory species richness,
diversity, evenness, and community composition were evident in the fourth year after
application;

* To compare the effects of over-the-top applications of sulfometuron methyl, hexazinone, a
sulfometuron methyl + hexazinone mix, and imazapyr on the survival and growth of planted
longleaf pine seedlings;

* To observe how the herbicide treatment plots responded to a prescribed fire administered five
years after herbicide application.


The understory component of the study is described in Chapter 2 of this thesis, the overstory

component is described in Chapter 3, and overall conclusions are discussed in Chapter 4.









CHAPTER 2:
USING HERBICIDE TO RESTORE THE UNDERSTORY FOLLOWING HARVEST

Introduction

Longleaf pine (Pinuspalustris Mill.) ecosystems occupy a wide environmental gradient

in the southeastern United States, from xeric sandhills to mesic and wet flatwoods (Walker and

Peet 1983, Peet 2006). Under a natural fire regime, the understory vegetation associated with the

two dominant species, longleaf pine and wiregrass (Aristida strict Michx.), varies predictably

with site hydrology (Harcombe et al. 1993). Mesic longleaf pine flatwoods, which occur on

Ultisols and Spodosols, fall in the middle of the longleaf pine hydrologic gradient and are highly

productive plant communities (Peet 2006). Whereas the understories of frequently burned xeric

longleaf pine sandhills are truly dominated by grasses and herbaceous plants, in mesic flatwoods

a large shrub component is present even under natural fire conditions (Peet 2006). These shrubs

are kept low to the ground by frequent fire, but in the absence of fire they will quickly succeed

into a dense midstory, eventually suppressing both longleaf pine and herbaceous plant

regeneration (Gilliam and Platt 1999).

Wiregrass is a key structural and functional component of all longleaf pine ecosystems in

the Southern Coastal Plain ecoregion (Peet 2006), and the establishment or enhancement of

wiregrass populations is essential for successful ecological restoration in these systems. Living

and dead wiregrass leaves are highly flammable and act as fire-spreaders, moving surface fires

evenly through the understory of a longleaf pine ecosystem (Clewell 1989). Dead longleaf pine

needles are also very flammable and are a critical component of the system's pyrogenicity

(Williamson and Black 1981, Clewell 1989). Both longleaf pine and wiregrass can therefore be

considered keystone species. Together they perpetuate a frequent fire return interval, which

favors their own reproductive success, acts as a selective force against competing, less fire-









tolerant species, and allows a host of herbaceous understory species to survive and reproduce in

the open, high-light understory (Williamson and Black 1981, Clewell 1989, Rebertus et al. 1993,

Jacqmain et al. 1999).

With high productivity and an abundance of shrubs in the understory, flatwoods are likely

to lose herbaceous species more quickly with fire exclusion than drier longleaf pine sites. Even

slight fire suppression in a flatwoods system leads to increased sprouting and vegetative

reproduction among shrubs, reducing available space for herbaceous plants and lowering species

richness (Glitzenstein et al. 2003). Different plant functional groups experience different rates of

species loss with reduced fire frequency. Mesic sites, with more species to lose and more rare

species, are likely to experience the highest rates of species loss (Leach and Givnish 1996). In

mesic to wet longleaf pine savannas, the species most likely to be reduced or lost are the

dominant bunch grasses, basal rosette species, sedges, other small monocots, and insectivorous

species (Walker and Siletti 2006).

Because the understory species composition and soil properties vary markedly with

hydrology in longleaf pine ecosystems, it is necessary to test restoration methods in a variety of

site conditions in order to develop appropriate restoration protocols. Mesic flatwoods in the

Florida panhandle are generally co-dominated by wiregrass (A. stricta, saw palmetto (Serenoa

repens W. Bartram), runner oak (Quercus elliottii Wilbur), gallberry (Ilex glabra A. Gray), and

hairy wicky (Kalmia hirsuta Walter). The overstory in a mesic longleaf pine flatwoods is

typically an open canopy of longleaf pine (basal area of 2.5 to 12.5m2/ha), with slash pine (Pinus

elliottii Engelm.) and pond pine (Pinus serotina Michx.) codominant in the transitional areas

near wetlands (Peet 2006). Fire is a key management tool for meeting restoration goals in these









systems, but social and environmental factors can often hinder the ability of land managers to use

fire effectively (Chapter 1).

When a restoration project includes conversion of the overstory from a planted slash pine

or loblolly pine (Pinus taeda L.) stand, complete removal of overstory pines across a large area

has the counterproductive effect of releasing shrubs (Brockway et al. 1998, Kush et al. 1999,

McGuire et al. 2001, Kirkman et al. 2007, Pecot et al. 2007). The release of shrubs following

overstory harvest is likely to be more extensive on a site with a past history of fire suppression,

due to the preponderance of underground hardwood rootstocks established during periods

without fire (Olson and Platt 1995; Drewa et al. 2006). Herbicide treatment and mechanical site

preparation can therefore be important components of restoration projects in which a slash or

loblolly pine plantation is being converted to longleaf (Walker and Siletti 2006). A potential

negative impact of herbicide, however, is a reduction in nontarget, desirable understory

vegetation.

Litt et al. (2001) conducted an extensive literature review on the effects of herbicide on

understory vegetation in southern pinelands. They examined 125 studies from which 21 were

used for review. Throughout the herbicide literature, they found a lack of experimental rigor and

inconsistent reporting standards, which made it difficult to reach any general conclusions about

the effects of herbicide on ground-layer vegetation. They were able to conclude only that

widespread use of herbicide to control unwanted vegetation may have undesirable effects on

nontarget plant species, and that additional studies of herbicide impacts are needed before

treating large, diverse landscapes. They recommend that future herbicide studies follow rigorous

experimental designs and make a deliberate effort to track species of special concern.









Litt et al. (2001) also concluded that the effects of herbicide on ground-layer vegetation

in natural flatwoods systems had rarely been measured. They reviewed only two studies

documenting herbicide impacts on groundcover in non-plantation flatwoods ecosystems, both of

which reported decreases in species richness. Decreases in richness ranged from 5.1% for

herbaceous species following hexazinone application (Wilkins et al. 1993), to 71.8% for all

species following treatment with a mix of sulfometuron methyl, glyphosate, and triclopyr (Neary

et al. 1991).

The effects of herbicide on sandhill communities are better documented. Four years after

treatment, hexazinone herbicide was effective in eliminating woody species and increasing

diversity relative to the control, though plots restored with fire alone had even higher species

diversity than those treated with herbicide (Provencher et al. 2001). On another north Florida

sandhill site, an initial decrease in species diversity was documented following hexazinone

application, followed by a recovery and then an increase in the second growing season

(Brockway et al. 1998). A subsequent study on the same site revealed that prescribed fire further

enhanced species diversity in conjunction with hexazinone (Brockway and Outcalt 2000). Three

herbicide release treatments (imazapyr, glyphosphate, and hexazinone) were tested on loblolly

pine plantations at three hilly sites in central Georgia. Seven years after treatment, there were no

significant difference in understory species diversity between any of the treatments including the

control (Boyd et al. 1995). Eleven years after treatment, Miller et al. (1999) arrived at the same

conclusion. Imazapyr, used alone or in conjunction with fire, was effective for restoring grass-

dominated bobwhite quail habitat on a loblolly pine sandhill in the Florida panhandle,

significantly reducing hardwood encroachment while enhancing the growth of native herbaceous

species (Welch et al. 2004). Beneficial results were also reported for imazapyr in a quail habitat









restoration project on hilltops in southern Lousiana (Jones and Chamberlain 2004). However, no

differences in herbaceous species cover and richness relative to the control were found on a site

in the Virginia piedmont treated with three different levels of imazapyr for pine release (Keyser

and Ford 2006). Other pine release studies have also reported beneficial or neutral results for

imazapyr, with wildlife forage vegetation showing recovery and in some cases improvement

after treatment (Hurst 1987, McNease and Hurst 1991).

In order to gain a better understanding of herbicide as a restoration tool for longleaf pine

flatwoods ecosystems, we compared the effects of five herbicide treatments on the understory of

a coastal flatwoods in the Florida panhandle: 1) Sulfometuron methyl (Oust', E.I. du Pont de

Nemours and Company, Wilmington, Del.), 2) Imazapyr (Arsenal', BASF Corporation,

Research Triangle Park, NC), 3) Hexazinone (Velpar L, E.I. du Pont de Nemours and

Company, Wilmington, Del.), 4) A tank mix of Oust' and Velpar L, and 5) a no-herbicide

control. These herbicide treatments were chosen because they are known to be effective

vegetation control treatments for loblolly and slash pine release in the Southeast, but their effects

on flatwoods community composition are not well studied. A recent survey reported that the

forestry herbicide most commonly used in the southeast was Arsenal', followed by tank mixes

of Arsenal' + Oust' and Oust' + Velpar (Shepard et al. 2004).

We examined the longer-term effects of these herbicide treatments on flatwoods

vegetation four years after initial herbicide application (see Ranasinghe 2003 for first year

results), after which we conducted a dormant-season prescribed fire on all treatment plots and

assessed post-fire vegetation composition.









Methods


Study Area

The study was conducted at the Point Washington State Forest, which is located on the

gulf coast of Walton County, Florida (300 20'16" N, 860 04'19" W). The 20-acre study site is

part of a larger mesic flatwoods/wet flatwoods matrix, with an understory dominated by saw

palmetto (S. repens), gallberry (I. glabra), rhizomatous oaks (Quercus minima Small and

Quercus elliottii Wilbur), blueberries (Vaccinium myrsinites Lam. and V darrowii Camp.),

huckleberries (Gaylussacia dumosa Torr. and A. Gray and G. frondosa L. Torr and A. Gray ex.

Torr), bluestem grasses (Andropogon virginicus var. glaucus (L.) Heck and Andropogon

virginicus L.) and wiregrass (Aristida stricta. The majority of the site can be classified as

mesic, but due to subtle topographical variation across the site some areas are at the wetter end of

the mesic flatwoods spectrum and some are at the drier end. For this reason, a randomized

complete block design was used, with three blocks laid out in mesic-wet areas and three blocks

laid out in mesic-dry areas. A pre-treatment vegetation survey identified wetter areas by the

presence of indicator species such as sedges (Cyperaceae). The soils of the study site are sandy,

siliceous, thermic, aeric alaquods of the Leon series (Spodosols consisting of deep, poorly to

very poorly drained soils derived from marine parent material (Ranasinghe, 2003)). Yearly

weather patterns during the study period were variable, with hurricanes in some years and

droughts in others. Total annual precipitation during the study period was 103cm in 2002,

200cm in 2003, 135cm in 2004, 146cm in 2005, and 106cm in 2006 (NOAA 2007). Rainfall in

the months following the prescribed fire was very low, with totals of 0.5cm in March 2007, 4cm

in April 2007, and 0.3cm in May 2007 (NOAA 2007).

Prior to August 2001, the 20-acre study site was a planted slash pine stand with a mean

stand age of about 26 years, which had been burned on a 3-year rotation by the Division of









Forestry since its acquisition of Point Washington State Forest in 1992. A pre-harvest, baseline

data survey was conducted in the areas slated to become study blocks in June 2001. In August

2001, the entire overstory was harvested and the site was prepared for the study with a single,

light roller chopping followed by a prescribed burn in October 2001. In December 2001, six

study blocks, each divided into five 36.6m x 24.4m treatment plots, were established according

to the aforementioned randomized complete block design. Each treatment plot was hand-planted

with 100 containerized longleaf pine seedlings laid out in 10 rows, with 3.1m between rows and

1.8m between the seedlings in each row.

Treatments

Within each block, each 100-seedling plot was randomly selected to receive one of five

herbicide treatments in March of 2002:

1) Sulfometuron methyl (Oius ") at 0.26 kg a.i./ha. This is a moderate application rate

for sulfometuron methyl; other studies have used 0.16 kg a.i./ha (Lauer and Glover, 1998) and

0.21 kg ai/ha (Shiver and Martin 2002, Keyser and Ford 2006), while the manufacturer-

recommended rate for herbaceous weed control in longleaf pine stands is 0.10 to 0.42 kg a.i./ha

(DuPont 2007). Sulfometuron methyl is a selective herbicide primarily effective for controlling

herbaceous species (Lauer and Glover, 1998), and mixing with hexazinone is recommended if

broader spectrum control is desired (DuPont, 2007).

2) Hexazinone (Velpar L) at 0.56 kg a.i./ha. Recommended application rates for

hexazinone range from 2.0 to 6.7 ai kg/ha (Du Pont 2007), while rates successfully used to

control hardwoods in sandhills ecosystems have ranged from 1.1 kg a.i./ha to 2.4 kg a.i/ha

(Brockway et al. 1998, Provencher et al. 2001). The hexazinone application rate in our study

was therefore very low. This rate was chosen because hexazinone had been shown to be









effective at the aforementioned rates, and we were interested in whether an even lower

application rate would be effective.

3) Sulfometuron methyl (0.26 kg a.i./ha) + Hexazinone (0.56 kg a.i./ha) mix. This is a

common tank mix used throughout the Southeast (Shepard et al. 2004) with the potential to

control a broad spectrum of woody and herbaceous plants. Tank mixes are used to achieve more

complete vegetation control than either herbicide alone, but as a result they may also have more

negative implications for native plant communities (Keyser and Ford 2006). Banded herbicide

application, the method used in this study, may help ameliorate these negative impacts (Keyser

and Ford 2006).

4) Imazapyr (Arsenal) at 0.21 kg a.i./ha. For a restoration project, this was a

moderate application rate for imazapyr. Significant hardwood control and habitat improvements

have been obtained using imazapyr at 0.42 kg a.i./ha (Jones and Chamberlain 2004), and at all

rates from 0.08 kg a.i./ha to 0.24 kg a.i..ha (though higher rates did not offer significant

improvements over lower rates) (Keyser and Ford 2006).

5) Control. (No herbicide).

The herbicides were applied in a 1.2m band over the seedling rows using a backpack

sprayer. The reasons for using a banded post-planting herbicide application on unshielded

seedlings were both operational and environmental. Prior to harvest, the study site already

contained many desirable understory species (including wiregrass) and few weedy species

(Ranasinghe 2003). The banded application allowed us to pinpoint the delivery of herbicide to

an area directly around planted seedlings, while native vegetation remained intact in the strips

between rows. This application method also lowered the cost of herbicide treatment relative to a

broadcast application. Within each 10-row treatment plot, three rows received a second









herbicide application in March of 2003, three rows received a second application in April of

2003, and four rows were left with only the first-year treatment. The results from these subplots

were pooled with the first year data in the final analysis because of a lack of significant

differences between one and two-year applications in any of the parameters.

Data Collection and Analysis

The pre-treatment vegetation survey was conducted in June 2001, prior to the overstory

harvest. Post-treatment vegetation surveys were conducted at three and nine months after

treatment, and the results of these first-year surveys have been published elsewhere (Ranasinghe

2003, Ranasinghe et al. 2005). The fourth-year data were collected in June of 2006. Within

each treatment plot, six 1m2 quadrats were randomly selected for understory sampling (two in

each subplot). In all, 36 randomly selected quadrats were surveyed in each herbicide treatment

and 12 were surveyed in the control, for a total of 156 quadrats. The smaller sample size in the

control treatment was due to the fact that a portion of each control plot was sprayed in the second

year, and these rows were not included for analysis. All species were identified and percent

cover of each species was estimated visually using the cover classes defined by the modified

Daubenmire scale (Daubenmire 1959, Peet et al. 1998): 0-1%, 1-2%, 2-5%, 5-10%, 10-25%, 25-

50%, 50-75%, 75-95%, and 95-100%. For each woody species, average stem height was also

estimated and all stems were counted. I. glabra, S. repens, and Quercus spp. (Q. minima and Q.

elliottii), were uniformly distributed across all study plots prior to treatment (Ranasinghe 2003)

and were grouped together for analysis. Woody species belonging to the family Ericaceae

(heaths) were analyzed separately from the other dominant shrubs because they were not evenly

distributed across plots prior to treatment, and also because members of this family are tolerant

to hexazinone and may react differently to treatment than other shrub groups (Wilkins et al 1993,









Outcalt et al 1999, Miller et al 1999). The same procedure was repeated in June of 2007,

following a prescribed fire on February 16, 2007.

Pyrometers were constructed using aluminum forestry tags and Tempilaq heat-sensitive

paints, following a method employed in other fire ecology studies (Hobbs et al. 1984, Iverson et

al. 2004). Thirteen Tempilaq paints (designed to melt at 930 C, 1490C, 2040C, 260C, 3160C,

3710C, 4270C, 4820C, 5380C, 593C, 6490C, 7040C, 760C, and 871C) were applied to each

tag. A second tag was paper-clipped over the front of the first to protect the paints from

charring. Six stakes were randomly placed in each herbicide treatment plot (two in each subplot)

and two in each control, with two pyrometers attached to each stake at 30cm and 80cm, for a

total of 36 pyrometers at each height for each herbicide treatment and 12 at each height for each

control. Each pyrometer gave a measure of the maximum fire temperature at that location: all of

the paint dots below that temperature were melted, while all of the paint dots above it remained

unchanged. When analyzing species cover vs. temperature relationships, the two pyrometer

readings at the 30cm level in each subplot were averaged and plotted against the average of the

two understory quadrats in that subplot. Since the pyrometers were placed randomly, they were

not necessarily in the same locations as the quadrats. Averaging them gave a more accurate

overall picture of the cover and fire intensity in a given plot. Pyrometers on which none of the

paints were burned (i.e. temperatures did not reach 930C) were assumed to be unburned and were

assigned a value of 100C to represent ambient air temperature.

All parameters were analyzed in JMP IN version 5 (SAS Institute, Inc.), using ANOVA

within the framework of a randomized complete block design. The study addressed only the

main effects of herbicide treatment, and tests of these effects were not dependent on the

assumption of no treatment x block interaction. Block effects were therefore treated as random









effects in a univariate ANOVA model with two independent variables: 'treatment' with

'Block&Random' as a covariate. Data were log-transformed where necessary to meet the

assumptions of ANOVA. Significant differences between treatments were separated with

Tukey's HSD or Hsu's MCB. Post-fire treatment differences were analyzed with ANCOVA,

using pre-fire distributions as a covariate. ANCOVA was also used to analyze fourth year

abundance of Ericaceae, using pre-treatment distributions as a covariate. A significance level of

a = 0.05 was used to test all parameters.

Community patterns were analyzed in PC-ORD version 4, using Nonmetric Multi-

Dimensional Scaling (NMDS) and Indicator Species Analysis following the method of Dufrene

and LeGendre (1997). NMDS uses rank-transformed distances to linearize the degree of

difference between plant survey quadrats. Ordination of plant survey quadrats (containing both

species presence and percent cover data) was done using a Sorenson distance measure in PC-

ORD's autopilot mode. The Sorensen coefficient is a measure of percent dissimilarity that can

be applied to either presence-absence data or quantitative data (in this case percent cover values),

and is recommended for analysis of ecological communities (McCune and Grace 2002). In the

autopilot mode, NMDS conducts 40 runs with real data and 50 runs with randomized data.

Species with less than 3 occurrences were dropped from the analysis to reduce the effects of rare

species (McCune and Grace 2002), reducing the total number of species in the analysis from 87

to 55. PC-ORD chooses the solution with the highest dimensionality that also meets the criterion

of having less stress than 95% of the random runs. Each point on the NMDS axes represents a

single plant survey quadrat, and the distance between points on the NMDS axes represents the

Sorenson distance between points in n dimensions (in this case a 3-dimensional solution was

chosen). A Multi-Response Permutation Procedure (MRPP) can then be used to determine the









degree of agreement within groups with regard to an environmental variable, which in this case

was a categorical variable (treatment). The weighting method for the MRPP was n/sum(n), and

we again used a Sorenson distance measure. MRPP yields a test statistic (A) and a p-value (A =

1 when all items are identical within groups (the groups in this case were herbicide treatments),

and A=0 when heterogeneity within groups equals expectation by chance). A therefore expresses

the degree to which variation in position between plant survey quadrats on the ordination axes is

"explained" by the herbicide treatments. Indicator species analysis using the Indicator Value

method defines an indicator species as a species found mostly in a single group and present in

the majority of sites belonging to that group (DuFrene and LeGendre 1997). The indicator value

(IV) is maximum (100%) when the individuals of species I are observed in all sites of only one

group (in this case, the each "group" was an herbicide treatment). IV is obtained by combining

relative abundance and relative frequency values for each species in each group. Relative

abundance of a species in a group is the average abundance of a given species in a given group

of plots over the average abundance of that species in all plots (expressed as a percent), and

relative frequency is the average number of plots in a given group where a given species is

present. Significance of each Indicator Value was established using a 1000-permutation MRPP

which compared the observed maximum IV for each species to the IV obtained from randomized

groups, and generated a p-value for this comparison using the Monte Carlo test of significance.

Species identified as indicators for one of the herbicide treatments are assumed to have been

enhanced by that treatment relative to the others, whereas species identified as indicators for the

control are assumed to have been negatively impacted by the herbicide treatments.









Results


Shrub Cover

Four years after initial herbicide application, percent cover of glabra, S. repens, and

Quercus spp. ranged from 13.5% in the sulfometuron + hexazinone (sulfo + hexa) treatment to

22.9% in the sulfometuron treatment, but did not vary significantly among treatments overall (p

= 0.2154) (Table 2-1). Indicator species analysis revealed a significant association between the

sulfometuron treatment and increased cover of both rhizomatous oak species, Q. elliottii and Q.

minima, four years after treatment (p = 0.040 and p = 0.031) (Table 2-2). Since these species

were not significantly associated with the control plots, this result can be interpreted to mean that

the sulfometuron treatment enhanced oak growth.

Woody species belonging to the family Ericaceae were abundant on the study site, and

included blueberries (V. myrsinites and V. darrowii), huckleberries (G. dumosa and G. frondosa),

hairy wicky (K. hirsuta), and Lyonia lucida Koch fetterbushh). Ericaceae cover four years after

treatment was four to five times higher in the control plots than in any of the herbicide

treatments, a difference which approached significance (p = 0.06), but was moderated by the

uneven pre-treatment distributions and the smaller control sample size (Table 2-1, Figure 2-1).

Overall percent cover of woody species was highest in the control and lowest in the sulfo + hexa

treatment (Figure 2-2). In June 2007, four months after the February 2007 prescribed fire,

percent cover of Quercus spp., I. glabra, and S. repens did not vary significantly among

treatments (p = 0.1740) (Figure 2-3a, Table 1), nor did Ericaceae cover (p = 0.097).

Herbaceous Cover

Herbaceous cover varied significantly among treatments (p = 0.002) (Figure 2-4a). In

both the imazapyr and hexazinone treatments, mean herbaceous cover (46.5% and 46.1%,

respectively) was significantly higher than the control (22.1%). Following fire, however,









herbaceous cover no longer differed significantly among treatments (p = 0.302). Herbaceous

cover decreased significantly following fire in all but the control and sulfometuron plots (Figure

2-4b). Much of this decrease can be attributed to the decrease in wiregrass crown size following

fire (see below). Overall vegetation cover before and after fire is summarized in Figure 2-5.

Wiregrass Cover

Mean wiregrass cover four years after initial herbicide application differed significantly

among treatments (p = 0.04) (Figure 2-6a). Wiregrass cover in the sulfo + hexa treatment was

three times greater than the control (Figure 2-6). In all treatments except the control, wiregrass

cover 4 years after treatment was higher than pre-treatment levels (Figure 2-6c). Following fire,

wiregrass cover decreased in all treatments (Figure 2-6b). The overall decrease in wiregrass

cover following fire is not attributed to wiregrass mortality, but to the decrease in wiregrass

crown size following fire. Wiregrass cover was surveyed only four months after the February

2007 fire, and it was observed that individual wiregrass clumps were alive and regenerating, but

greatly reduced in size from pre-fire levels. A similar effect was documented after the herbicide

treatment in 2002, with an initial decrease in wiregrass cover followed by a rebound over the

next nine months (Figure 2-6c).

Shrub Stem Density

Four years after treatment, mean shrub stem density for Quercus, Ilex, and Serenoa

varied significantly among treatments (p = 0.012), displaying a similar pattern to percent cover

(Table 2-1). Mean stem density in the imazapyr treatment (20 stems/m2) was significantly lower

than sulfometuron (34.7 stems/m2) and hexazinone (32.9 stems/m2). None of the treatments

varied significantly from the control, in which we measured a mean stem density of 31.75

stems/m2. In the first nine months after herbicide application, imazapyr was the only treatment

that did not experience an increase in shrub stem density (Figure 2-7b). In the fourth year after









application, this effect was no longer evident, as shrub stem density no longer varied

significantly among treatments. Following the prescribed fire in February 2007, mean shrub

stem density increased again by around 50% in every treatment (Figure 2-7a). These increases

were significant in all treatments but the control. Variation in Ericaceae stem density among

treatments approached significance in June 2006 (p = .061) (Table 2-1), but this difference was

moderated by the smaller control sample size and uneven pre-treatment distributions. Ericaceae

stem density did not change significantly in any of the treatments following fire (Table 2-1).

Shrub height

Mean shrub stem height for Quercus, Ilex, and Serenoa did not differ significantly among

treatments in June 2006 (p = 0.2341). Stem height ranged from 14.6cm in the imazapyr

treatment to 17.1cm in the hexazinone treatment (Table 2-1). Following the prescribed fire in

February 2007, mean shrub height increased in the imazapyr, sulfo, and sulfo + hexa treatments

(Table 2-1). Differences among treatments following fire were not significant (p = 0.40). Mean

stem height of Ericaceae did not differ significantly among treatments in June 2006 (p = 0.304)

or in June 2007 (p = 0.900), though height increases within some treatments following fire were

significant (Table 2-1).

Understory responses

A total of 93 species were identified during the June 2006 and June 2007 plant surveys

(Appendix A). Weedy species were not prevalent, and we found numerous plant specimens

belonging to functional groups identified by Walker and Peet (1983) and Glitzenstein et al.

(2003) as high-risk for being lost in shrub-invaded flatwoods, such as bunch grasses (Poaceae),

sedges (Cyperaceae), sundews (Droseraceae), and basal rosette species.

Overall species richness four years after herbicide application varied significantly among

treatments (p = 0.01), ranging from 7.4 species/m2 in the control treatment to 9.3 species/m2 in









the sulfometuron treatment (Table 2-3). Herbaceous species richness, however, did not vary

significantly among treatments (p = 0.498) (Table 2-3). Following fire, overall species richness

no longer varied significantly among treatments (p = 0.15). Within treatments, responses to the

fire varied (Table 2-3). Overall species richness increased significantly in the control, but did

not change significantly in the other treatments. Herbaceous species richness did not vary

among treatments before or after fire (p = 0.498 and p = 0.114, respectively).

Species Diversity

Overall Shannon diversity (woody and herbaceous species combined) four years after

herbicide application did not vary significantly among treatments (p = 0.07) (Figure 2-8a, Table

2-3). However, herbaceous species diversity did vary significantly among treatments (p =

0.036), with the highest diversity in the sulfometuron treatment and the lowest in the sulfo +

hexa treatment (Table 2-3). Within treatments, responses to the prescribed fire varied (Figure 2-

8b). Overall diversity increased following fire in the sulfo + hexa and hexazinone treatments,

but did not change significantly in any of the other treatments (Figure 2-8b). Herbaceous

diversity increased in the imazapyr, sulfo + hexa, and hexazinone treatments following fire

(Table 2-3). Initial gains in overall Shannon diversity during the first year after treatment

(Ranasinghe 2003) were no longer evident after four years (Figure 2-8c).

Species Evenness

Mean species evenness (1/D) four years after herbicide application, calculated for woody

and herbaceous species combined, did not vary significantly among treatments overall (p = 0.12)

(Table 2-3). Herbaceous species evenness, however, did vary significantly among treatments (p

= 0.014), with the highest evenness in the sulfometuron treatment and the lowest in the sulfo +

hexa treatment (Table 2-3). Following fire, only the sulfo + hexa treatment experienced a

significant change in overall evenness (from 3.12 to 4.54) (Table 2-3) Herbaceous species









evenness increased in the imazapyr, sulfo + hexa, and hexazinone treatments following fire

(Table 2-3).

Community Ordination

Analysis of the plant community four years after treatment using NMDS and MRPP

revealed only a slight relationship between treatment and species composition (Fig 2-8), with a

small chance-corrected within-group agreement (A) of 0.02 (p = 0.01) for the pre-fire survey

quadrats. Indicator Species Analysis following the method of Dufrene and LeGendre (1997)

revealed significant relationships between some species and treatments (Monte Carlo test of

significance) (Table 2-2).

Species of Special Interest

Curtiss' sandgrass (Calamovilfa curtissii Scribn.), a florida panhandle endemic, was

identified in one of the imazapyr treatment plots during both the June 2006 and June 2007 plant

surveys. Prior to treatment, there was a very low density of legumes on our study site: out of 90

survey quadrats, only two legume specimens were identified, both of which were Mimosa

quadrivalvis var. angustata Barneby (Ranasinghe, unpublished data). After four years, we found

Desmodium lineatum DC, Desmodium strictum DC, and Tephrosia hispidula Pers. in addition to

M. quadrivalvis, again at a low density: a total of eight legume specimens were identified out of

156 survey quadrats. In the post-fire survey, one D. lineatum, four M quadrivalvis, and one

Baptisia lanceolata Elliott specimen were identified. These species were not significantly

associated with any of the treatments.

Fire Temperature

The prescribed fire was cool, fast-moving, and patchy. Mean maximum fire temperature

by treatment ranged from 2520C to 3060C at 30cm and 166C to 2570C at 80cm (Table 2-4), and

did not differ significantly among herbicide treatments at either height (p = 0.42 and p = 0.55,









respectively). A significant negative relationship existed between percent Quercus cover and

maximum fire temperature at 30cm (r2= 0.06, p = 0.032) (Figure 2-9a), whereas a significant

positive relationship existed between percent I. glabra cover and maximum fire temperature at

30cm (r2 = 0.07, p = 0.017) (Figure 2-9b). The negative relationship observed with Quercus is

largely attributable to the fact that areas with high Quercus cover were often only partially

burned or did not ignite at all, whereas the I. glabra tended to ignite readily and burn completely.

Temperature was not significantly related to wiregrass cover (p = 0.73) or overall herbaceous

cover (p = 0.28).

Discussion

Four years after initial herbicide application, differential treatment effects on understory

composition were still evident. Sulfometuron methyl had a releasing effect on the two most

common oak species, Q. minima and Q. elliottii, indicating that this is not an appropriate

herbicide for flatwoods with a high density of oaks. However, in spite of having higher shrub

cover and lower herbaceous cover than the other herbicide treatments, sulfometuron plots also

had the highest fourth year herbaceous species richness, diversity, and evenness. One reason for

this may have been that wiregrass cover in the sulfometuron plots was somewhat lower than the

other treatments, leaving more space for other small herbaceous species. These gains in species

richness, diversity, and evenness are therefore unlikely to persist through future fire cycles

because it will be more difficult to apply prescribed fire to this treatment. Indeed, sulfometuron

was the only treatment in which species richness, diversity, and evenness did not increase

following the prescribed fire in 2007.

Woody shrub cover, height, and stem density were lowest in the imazapyr treatment, but

the differences between imazapyr and the other treatments were not as pronounced as those

observed in the first year after application. These results suggest that the increases in shrub stem









density evident in the first year (Figure 2-7b) were a short-term phenomenon, and subsequent

dieback of sprouts occurred as these shrubs matured and grew taller.

The responses of Ericaceae to our treatments were surprising, given that in other studies

members of this family have been highly hexazinone-resistant (Wilkins et al. 1993, Outcalt et al.

1999, Miller et al. 1999), and hexazinone is commonly used as a release treatment for blueberry

crops. If anything, we expected to see higher Ericaceae cover in the hexazinone treatment, when

in fact this treatment had the lowest Ericaceae cover of all. This result was particularly notable

given that the hexazinone plots had the highest percent cover of Ericaceous shrubs prior to

treatment (Figure 2-1). The sensitivity of Ericaceous shrubs to hexazinone may have been a

function of the timing of application and the poorly drained, sandy soil; Velpar L is labeled for

blueberry release only prior to budbreak in the spring and in conditions where there is no

standing water (DuPont, 2007). Control of these species may be desirable for restoration

scenarios in which the goal is to increase herbaceous cover on a site dominated by Ericaceous

shrubs (Outcalt et al. 1999). However, many of these shrubs are important wildlife food plants

(Hay-Smith and Tanner 1994). If the goal is to avoid negative impacts to these populations, our

results suggest that caution may be in order with regard to soil type, soil water, and timing of

application.

Both imazapyr and hexazinone significantly increased overall herbaceous cover relative

to the control, and the sulfo + hexa treatment significantly increased wiregrass cover relative to

the control. Though the shrub control initially provided by the herbicide treatments was

relatively short-lived, these results suggest that the short-term shrub control allowed herbaceous

species (and wiregrass in particular) to gain more of a foothold in the understory.









The mean fire temperatures we recorded were low for a dormant-season fire in a mesic

flatwoods: mean maximum fire temperatures for several other dormant season fires on

flatwoods sites ranged from 3830C to 5880C (Drewa et al. 2002), whereas our mean maximum

temperatures were 2000C at 80cm and 2740C at 30cm. Across the study site, we made the

general observation that I. glabra burned more readily and thoroughly than Quercus spp., which

often burned only partially, and, in some highly oak-dominated patches, failed to ignite at all.

Since dead pine needles are a critical component of pyrogenicity in longleaf pine systems

(Williamson and Black 1981, Clewell 1989), it is likely that the complete removal of overstory

pines contributed to our inability to apply fire effectively to the site.

In our study, stem counts of both Quercus spp. and I. glabra more than doubled in all

treatments in the four months following fire. Because we did not have an unburned control, we

cannot say to what degree the observed post-fire increases in shrub cover and density were

greater than the increases expected from normal spring growth of these shrubs in the absence of

overstory competition. However, it is safe to say that the prescribed fire was ineffective at

controlling the shrubs, because it did not decrease shrub cover or even maintain it at pre-fire

levels. This result can be attributed to aggressive shrub regrowth due to the lack of competition

from overstory pines, the loss of needlefall as a fine fuel supply, and the length of time since the

previous fire and subsequent herbicide applications (four and five years). Shrub control, and by

extension floral diversity, are maximized by fire return intervals of one to three years (Kirkman

et al. 2001; Glitzenstein et al. 2003). If future fires are conducted on a shorter interval, better

prescribed fire results may be obtained.

The season of burn may also have contributed to the difficulty we had in applying

prescribed fire effectively. It is generally believed that summer fires were probably the most









prevalent selective force prior to European colonization of the Southeast, and are more effective

at controlling shrubs and maintaining herbaceous plant diversity in the understory (Glitzenstein

et al 1995, Streng et al. 1993). Others have reported shrub resprouting responses to dormant-

season fire in flatwoods similar to those we observed. On mesic and wet sites in Florida and

Lousiana, shrub stem densities were greater than pre-treatment levels after dormant-season fire,

and continued to increase with repeated dormant season fires, whereas long-term growing season

fires maintained stem densities at the same level (Drewa et al. 2002). In another recent study, it

was observed that stem densities of root-crown bearing shrubs (Quercus and Ilex spp.) after

repeated dormant-season fires were seven times greater than those subjected to growing-season

fires (Drewa et al. 2006). In one case, stem densities of root-crown bearing shrubs were 50 times

greater in South Carolina savannas managed with annual dormant season fires than in plots

managed with annual growing-season fires over a 30 year period (Waldrop et al 1992). Some of

the resprouting ability of Quercus can be attributed to a past history of fire suppression, during

which underground stems are likely to have grown extensively (Drewa et al. 2006). Given that

our site was previously a slash pine plantation, periods of fire suppression are likely to have

occurred in the past during seedling establishment and may have contributed to the recalcitrance

of the dominant woody shrubs.

However, results indicating that dormant-season fire increases shrub cover in flatwoods

are not universal. In one case, a long-term dormant season prescribed fire regime significantly

reduced shrub cover on an I. glabra-dominated flatwoods while increasing herbaceous species

cover and diversity (Brockway and Lewis 1997), suggesting that dormant-season fire is a viable

option for flatwoods management when growing-season fire cannot be used. Hiers et al. (2000)

offer experimental evidence, based on patterns of legume reproduction, that variable fire regimes









will maintain a broader suite of native species than either growing-season or dormant-season fire

alone, and cite studies suggesting that significant variation in the frequency and season of

lightning may have occurred during the 7000 years BP that longleaf pine communities have been

established in the southern coastal plain.

Though dormant season fire may be effective and even desirable in some cases, our

results suggest that dormant season fire should be avoided when species of low flammability are

abundant, there is a lack of fine fuels, and there are no overstory pines to compete with shrubs

for light and soil resources. Ilex glabra is known to be one of the most flammable flatwoods

species, while oaks are among the least flammable (Behm et al. 2004), and this may be another

part of the reason for the success of dormant-season fire on an I. glabra-dominated site reported

by Brockway and Lewis (1997). In our study, though stem density of both Quercus and Ilex

increased following fire, only Quercus increased in mean percent cover across treatments; mean

I. glabra cover remained roughly the same or decreased. It may be that in the coming growing

seasons, some of the initial flush of basal I. glabra sprouts will die back, and this species will not

show an overall cover increase. Since Quercus increased in cover as well as stem density, it

seems more likely that the fire-related increase will have long-term effects.

In summary, only positive herbicide impacts on wiregrass and herbaceous cover were

observed in this study, and understory species diversity, richness, evenness, and community

composition were largely unaffected. The imazapyr treatment, which had the most dramatic

shrub control effects in the first year after application, still had the lowest shrub levels and the

highest herbaceous cover after four years, indicating that this was the most successful treatment

overall. The poor vegetation control offered by sulfometuron, which was largely attributable to

the resistance of Q. elliottii and Q. minima, indicates that it is not an appropriate herbicide for









use in an oak-dominated flatwoods community. However, the sulfometuron + hexazinone tank

mix showed promise as a flatwoods restoration treatment, offering better control over Quercus

spp. and significantly increasing wiregrass cover relative to the control. Though hexazinone

plots had significantly higher overall herbaceous cover than the control, our results suggest that

this herbicide, at the rate applied, did not provide much control over the dominant shrubs in the

long term. Because we applied a very low rate of hexazinone, it is difficult to make any absolute

comparisons between hexazinone and the other herbicides based on the results of this study.

However, given the success others have had with hexazinone (Brockway et al 1998, Brockway

and Outcalt 2000, Provencher et al 2001), it seems likely that we would have seen greater shrub

control and attendant increases in herbaceous cover had we used a higher rate.

As of June 2007, it was too early to tell whether the herbicide-related increases in

wiregrass and overall herbaceous cover will be maintained through future fire cycles. The

Florida Division of Forestry normally burns this site on a three-year fire interval, and results

from more frequent burning may be more successful than those we observed five years after the

initial herbicide application and four years after the second application. Because the entire

overstory was harvested, it may be several fire cycles before the new longleaf pine canopy

provides sufficient needlefall to sustain the level of prescribed fire necessary for shrub control.

In the meantime, this limitation is likely to be exacerbated by the Florida Division of Forestry's

inability to use growing-season fire on this site due to its proximity to a highway and several

beach housing communities. If the trends of increasing shrub cover, height, and density

continue, the herbicide-related gains in wiregrass and overall herbacous cover may be lost. The

aggressive post-fire resprouting of shrubs observed in this study underscores the utility of

herbicide as a supplemental restoration tool, especially in flatwoods ecosystems where undesired









changes in understory community composition can be rapid and difficult to reverse. It also

underscores the need for effective prescribed fire in order to maintain the early gains in wiregrass

and herbaceous cover conferred by herbicide application.










Table 2-1. Mean percent cover, stem density, and height of shrubs 4 years after treatment
(YAT) (June 2006) and following prescribed fire (June 2007).
% Cover Stem Density (stems/m2) Stem Height (cm)
Q, S, I Ericaceae Q, S, I Ericaceae Q, S, I Ericaceae
2006 Post- 2006 Post- 2006 Post- 2006 Post- 2006 Post- 2006 Post-


Treatment 4YAT Fire 4YAT Fire 4YAT Fire 4YAT Fire 4YAT Fire 4YAT Fire
Control 18.7 22.5 21.3 2.8 31.8ab 56.3 42.9 10.3 16.1 21.1 11.6 12.5

Imazapyr 13.6 24.3* 5.1 2.3 20.8b 47.8* 8.4 8.9 14.6 19.7* 12.3 12.1

Sulfometuron 22.9 34.0* 4.4 5.6 34.7a 64.9* 17.2 25.0 16.9 22.3* 12.2 15.2*

Sulfo+Hexa 13.5 21.5 5.8 3.6 25.1ab 50.9* 16.5 12.8 15.9 21.6* 12.0 17.0*

Hexazinone 19.9 26.6 4.0 3.3 32.9a 59.1* 13.0 9.6 17.1 19.7 11.9 16.1*

ANOVA/ P= P= P= P= P= P= P= P= P= P= P= P=
ANCOVA 0.215 0.174 0.060 0.097 0.012 0.331 0.061 0.108 0.234 0.400 0.304 0.900


*Letters denote significant differences between means in the same column at a = 0.05
based on Tukey-Kramer HSD test. (*) denotes significant post-fire change within
treatment (Tukey-Kramer HSD). P-values are for randomized block ANOVA / ANCOVA
within columns (significant at a = 0.05). Q, S, I stands for Quercus spp., S. repens and .
glabra.











Table 2-2. Indicator species analysis following the method of Dufrene and LeGendre (1997).
IV from randomized groups
Observed
Species Treatment Indicator Value (IV) Mean S. Dev P*
Asclepias cineria Sulfo + Hexa 15.5 5.4 2.91 0.010
Pteridium aquilinum Sulfometuron 20.3 11.7 3.4 0.020

Lachnocaulon auceps Imazapyr 12.2 5 2.86 0.025

Lachnanthes caroliniana Hexazinone 16.7 4.1 2.5 0.010

Quercus elliottii Sulfometuron 19.1 12.6 3.27 0.040

Quercus minima Sulfometuron 29.7 21.7 3.56 0.031

Smilaxpumila Control 12.5 3.7 2.39 0.029

* Monte Carlo test of significance, a =0.05










Table 2-3. Species Richness (species/m2), Diversity (Shannon Index), and Evenness
(Simpson's Index, 1/D) 4 years after treatment (YAT) (June 2006) and following fire
(June 2007).
Species Richness (spp/m2) Shannon Diversity Evenness (1/D)
Overall Herbaceous Overall Herbaceous Overall Herbaceous
2006 Post- 2006 Post- 2006 Post- 2006 Post- 2006 Post- 2006 Post-
Treatment 4YAT Fire 4YAT Fire 4YAT Fire 4YAT Fire 4YAT Fire 4YAT Fire
Control 7.36c 9.75* 4.58 6.83 1.34 1.67 1.10ab 1.59 3.54ab 4.24 3.02ab 4.36

Imazapyr 8.97ab 9.0 5.75 5.97 1.48 1.70 1.01ab 1.40* 3.51ab 4.43 2.45ab 3.67*

Sulfometuron 9.25a 8.19 5.41 5.05 1.57 1.51 1.17a 1.30 4.15a 3.80 3.03a 3.53

Sulfo +Hexa 8.03bc 9.0 5.13 6.14 1.34 1.69* 0.85b 1.40* 3.12b 4.54* 2.11b 3.79*

Hexazinone 8.22bc 8.86 5.33 5.94 1.35 1.65* 0.88b 1.45* 3.01b 4.07 2.13b 3.95*

ANOVA/ P= P= P= P P= P= P= P= P= P P P
ANCOVA 0.014 0.151 0.498 0.114 0.066 0.131 0.036 0.293 0.116 0.249 0.014 0.533
*Letters denote significant differences among means in the same column at a = 0.05 based
on Tukey-Kramer HSD test. (*) denotes significant post-fire change within treatment
(Tukey-Kramer HSD). P-values are for randomized block ANOVA / ANCOVA within
columns (significant at a = 0.05).










Table 2-4. Mean maximum fire
temperature (C) by
treatment.
Mean Fire Temp. (oC)
Treatment 30cm 80cm
Control 276 (57) 181(36)

Imazapyr 275 (32) 190 (27)

Sulfometuron 306 (29) 197 (21)

Sulfo + Hexa 260(26) 166 (22)

Hexazinone 252 (30) 257 (30)

ANOVA p = 0.420 p =0.546
* P-values for randomized block
ANOVA are significant at a = 0.05.
Standard error in parentheses.
















35 0 OPre-Harvest
S30 4 yrs post-treatment
0 25
20
15
10



Control Imazapyr Sulfo Sulfo + Hexa Hexazinone
Treatment

Figure 2-1. Pre-harvest (June 2001) and June 2006 percent cover of Ericaceae by treatment.
Pre-harvest Ericaceae cover was not uniform among treatment plots (ANOVA, p =
0.009). 4 years post-treatment, variance in Ericaceae cover by treatment approached
significance (ANCOVA, p = 0.060).














June 2006 (4 years after treatment)


3 40 -
[ Ericaceae
a) 30
SSerenoa
20- Quercus

10 Ilex


Control Imazapyr Sulfo Sulfo+Hexa Hexa
Treatment

45
June 2007 (post-fire)
40
35
S30 -O Ericaceae

b) 25 Serenoa
I QIerCuI
15 -
S10 E Ilex
15

0-
Control Imazapyr Sulfo Silfo+Hexa Hexa
'Ieatment
Figure 2-2. a) Mean percent shrub cover four years after treatment (June 2006) for Quercus spp.,
L glabra, S. repens and Ericaceous shrubs (Vaccinium spp., K. hirsuta, Gaylussacia
spp., and L. lucida). b) Mean percent shrub cover in June 2007, four months after fire.






















Control Imazapyr Sulfo Sulfo+Hexa
Treatment


Control


Imazapyr


Sulfo


Sulfo + Hexa


Non-Ericaceae
5 Pre-Fire
a Post-Fire


Hexa


Ericaceae
O Pre-Fire
* Post-Fire


Hexa


Treatment


Figure 2-3. Mean percent shrub cover before (June 2006) and after fire (June 2007) for a)
Quercus spp., I. glabra, and S. repens, and b) Ericaceae. Differences between and
within treatments are not significant on either graph.


0

a)


30
25
20
b) S 15
10


L Ai I FT
















a abc ab
S50 abc

S40 bc
O Pre-Fire
U Post-Fire
30

~ 20

2010 i
0 -


Control Imazapyr Sulfo Slfo+Hexa Hexazinone
'I-eatment

Figure 2-4. Mean % herbaceous cover (including both forbs and graminoids) 4 years after
treatment (Pre-fire) and 5 years after treatment (Post-fire). Pre-fire means not sharing
the same letter are significantly different (Tukey-Kramer HSD, a = .05). (*) indicates
significant pre-fire / post-fire differences within treatments (Tukey-Kramer HSD,
a=.05).






















80 June 2006 (4 years after treatment)
70
60
50

a) b40
30
20
10 5- Herbaceous
0O Sh ubs
Control Imazapyr Sulfo Sulfo +Hex Hexzinone
80 h Ericaceous
Shrubs
70 June 2007 (Following Feb. 2007 fire)

60

50

b) 40
2 30

20

10


Control Imazapyr Sulfo Sulfo+Hexa Hexazinone
Treatment


Figure 2-5. Overall vegetation cover in a) June 2006 (4 years after treatment) and b) June 2007
(4 months after fire). Cover is separated into herbaceous, shrub, and Ericaceous
shrub components.

























a

ab










Sulfo Hexa Hexa


l Pre-Fire
* Post-Fire


--- Control
--- Imazapyr
- Sulfo
-*- Sulfo+Hexa
-X Hexa


Pre-harvest 3 Mos 9 Mos 4yrs Post-fire
Time


Figure 2-6. a) Mean % cover of wiregrass four years after initial treatment (pre-fire) and five
years after treatment (post-fire). Pre-fire means not sharing the same letter are
significantly different (Hsu's MCB, a = 0.05). (*) indicates a significant pre-
fire/post-fire difference (Tukey-Kramer HSD, a = 0.05). b) Mean % wiregrass cover
within treatments over the course of the 5-year experiment. First-year results from
Ranasinghe (2003 and unpublished data), and Ranasinghe et al. (2005).


40
35
o 30
I 25


15
10
5
0


b





Control


b








Sulfo
Treatment


a

Imazapyr


35
S30
S25

b) 20
S15

10
a










*


a a
ab

^rlb


D Pre-Fire
* Post-Fire


Hexa


-o- Control
-I- Imazapyr
- Sulfo
-- Sulfo+Hexa
-- Hexa


Post-fire


Figure 2-7. a) Mean shrub stem density (stems/m2) before (June 2006) and after (June 2007) fire
for Quercus spp., I glabra, and S. repens. Pre-fire means not sharing the same letter
are significantly different. (*) indicates a significant change following fire (Tukey-
Kramer HSD, a = 0.05). b) Changes in shrub stem density within each treatment over
the course of the 5-year experiment. First-year results from Ranasinghe (2003 and
unpublished data), and Ranasinghe et al. (2005).


4f1
a


a) I
E
=
VI

I


*l

IIl

ffll


Control


Imazapyr


Sulfo
Treatment


Sulfo+Hexa


4 100
80

b) 60
40
c 20


Pre-harvest


3 Mos


9 Mos
Treatment




















r rlr*


Sulfo+Hexa


--- Control
- Imazapyr
- Sulfo
-- Sulfo+Hexa
-- Hexa


Pre-treat 3 Mos 9 Mos 4yrs Post-Fire
Time
Figure 2-8. a) Mean species diversity (Shannon Index) by treatment four years after treatment
(Pre-fire) and five years after treatment (Post-fire). Pre-fire differences among
treatments are not significant (ANOVA, p = 0.13). (*) indicates a significant pre-
fire/post-fire difference (Tukey-Kramer HSD, a = 0.05). b) Changes in species
richness within each treatment over the course of the 5-year experiment. First-year
results from Ranasinghe (2003 and unpublished data), and Ranasinghe et al. (2005).


a)


Control


O Pre-Fire
* Post-Fire


Imazapyr


Sulfo


Treatment


2.6
2.4
2 2.2
a 2
b) 1.8
S1.6
1.4
















AL
0


A m, A
0 AI





A A



x U
C;


t
U


A 0


*L,
A


* Control
* Imazapyr
Sulfo
+ Sulfo + Hexa
A Hexazinone


-1.5. *
-1 5 -05 05 15
Axis 1

Figure 2-9. NMDS ordination of plant communities. Each point represents one plant survey
quadrat. Chance-corrected within-group agreement (A) = 0.02 (a = 0.05, p = 0.01).



mm



















E 500-
o
C-

1 400-
0)

1 300-
0)
S -
200-
--

0)
L100-


0-






600-


E 500-
0)
o
16 400-

C -



200-


1-00-


0 10 20 30

Quercus % Cover


I I5
40 50


U -I* I I T I i- "1 i -I I
0 5 10 15 20 25 30
Ilex % Cover



Figure 2-10. a) relationship between mean Quercus cover and mean fire temperature (C) (r2

0.06, p = 0.03) in treatment subplots, and b) relationship between mean Ilex cover and

mean fire temperature (C) (r2 = 0.13, p = 0.002) in treatment subplots.


I


rr\n


I







~



I









CHAPTER 3
HERBICIDE AND SUSTAINABLE LONGLEAF PINE SILVICULTURE

Introduction

Perhaps more than any other forested ecosystem in the United States, the longleaf pine

(Pinus palustris Mill.) savanna lends itself to simultaneous management for timber and

biodiversity. The understory communities associated with longleaf pine are dominated by

herbaceous plants adapted to frequent ground layer disturbance in the form of fire, and regular

overstory disturbance by lightning, windthrow, and hurricanes (Platt and Rathbun 1993, Platt et

al. 2002). In longleaf pine systems with a frequent fire regime and an intact seedbank, most

native herbaceous species respond positively to the increased light levels in newly-opened

canopy gaps (Keddy et al. 2006, Platt et al. 2006), and it has been hypothesized that many of

these species are such weak competitors for light that they could be considered fugitive or

ephemeral (Keddy et al. 2006). In one case, removal of the overstory resulted in a 20% increase

in the number of understory species in gaps within five years, and surveys of older gaps indicate

that similar increases have been maintained over several decades (Platt et al. 2006). Wiregrass

(Aristida strict Michx.), the dominant bunchgrass in most longleaf pine systems and a keystone

species with regard to pyrogenicity, is also released in response to overstory removal (McGuire

2001). Since the renowned plant biodiversity of the longleaf pine ecosystem resides largely in

the herbaceous component (Walker and Peet 1983, Kirkman et al. 2001), the removal of large

overstory trees for timber harvest has the potential to benefit the ecosystem.

However, there are several important caveats to the analogy between silviculture and

natural disturbance in longleaf pine systems (Mitchell et al. 2006). One such caveat is that the

removal of overstory pines across a large area also releases shrubs (Brockway et al. 1998, Kush

et al. 1999, McGuire et al. 2001, Kirkman et al. 2007, Pecot et al. 2007), which, in the absence of









adequate control, eventually leads to the exclusion of herbaceous plants (Gilliam and Platt 1999,

Glitzenstein et al. 2003). The release of shrubs following overstory harvest is likely to be more

extensive on a site with a past history of fire suppression, due to the preponderance of

underground hardwood rootstocks established during periods without fire (Olson and Platt 1995;

Drewa et al. 2006). Shrub invasion excludes herbaceous species not only through competition

for light and resources, but through the alteration of fire regimes away from the high-frequency

fire intervals perpetuated by wiregrass and pine needles. Hardwoods produce litter of lower

flammability than pines or wiregrass, which can lead to longer fire return intervals in gaps,

ultimately exerting a negative feedback on the system's pyrogenicity and perpetuating a

hardwood midstory (Williamson and Black 1981, Rebertus et al. 1993, Jacqmain et al. 1999).

When a large portion of the overstory is removed, this negative feedback cycle is likely to be

exacerbated by the subsequent lack of needlefall and the potential loss of wiregrass due to

harvest-related soil disturbance. These are the two fine fuel inputs essential to maintaining a

frequent fire return interval in a longleaf pine ecosystem, and without them the keystone species

in the system will not regenerate (Williamson and Black 1981, Clewell 1989).

The ability of longleaf pine seedlings to survive and emerge from the grass stage will also

be impacted if there are not adequate fine fuels to carry fire through the system and keep

hardwoods suppressed. Longleaf seedlings are highly intolerant of competition and will not

initiate height growth if an excessive amount of competing vegetation is present (Boyer 1963,

Haywood 2000, Ramsey et al. 2003).

One approach to these problems, which has received increasing attention in recent years,

is to develop an uneven-aged silvicultural system that replicates natural smaller-scale patterns of

disturbance and succession so as not to disrupt the overall continuity of fine fuels and understory









dynamics (Palik et al. 2002, Mitchell et al. 2006, Brockway et al. 2006). It has long been

observed that longleaf pine seedlings regenerate only where disturbances have created canopy

openings (Boyer 1963). Several recent studies have looked at patterns of regeneration in natural

gaps as well as gaps created by group- or single-tree selection. Naturally-regenerating longleaf

pine seedlings on a xeric sandhill were clustered near the center of gaps but were absent around

the outer edge of each gap, an observation which led Brockway and Outcalt (1998) to propose

the existence of a 12-16m "seedling exclusion zone" resulting from 1) higher fire intensity due to

high litter accumulation at the bases of trees around the gap edge, and 2) competition with the

root systems of adult pines around the gap edge. Some studies of longleaf pine regeneration

have supported this hypothesis. As basal area in a mature longleaf pine stand increased, fire-

related seedling mortality also increased and seedling size decreased (Grace and Platt 1995),

while higher light intensity and N availability in gap centers were correlated with seedling

survival and growth (Palik et al. 1997). Though subsequent research has not found evidence for

a seedling exclusion zone per se, suppressed growth with increasing proximity to adults has been

shown in all studies. Container-grown seedlings planted in gaps exhibited maximum growth in

response to high light intensity at the center of gaps as small as .1-ha, but no strong response to

nutrient availability (McGuire et al. 2001). Others have reported that seedling survival was

actually higher toward the outer edges of gaps, though the seedlings able to survive in the center

were significantly larger than those closer to the edges (Gagnon et al. 2003, Pecot et al. 2007).

Results from a trenching experiment suggest that this tradeoff is due to underground facilitation

by adult longleaf pines at gap edges vs. greater light availability at the gap center (Pecot et al.

2007). Trenching also revealed that hardwoods were immediately released when belowground

competition from mature canopy trees was removed (Pecot et al. 2007). The upshot of these









recent studies is that small group selection and single-tree selection are recommended as

silvicultural treatments for longleaf pine systems in which both biodiversity conservation and

timber yield are desired. These methods enable new seedlings to establish at acceptably high

survival rates without a major disruption in the supply of fine fuels or excessive hardwood

release.

But what of a system in which there are no longleaf pines in the overstory, and the

desired outcome is a complete conversion of the overstory from slash pine (Pinus elliottii

Engelm.) to longleaf pine? The same issues with hardwood control and fine fuel supply must be

considered. It has recently been shown that a gradual replacement approach is also an effective

method for making the conversion from slash to longleaf. The single-tree selection method was

used to create plots with variable basal area densities, into which longleaf pine seedlings were

introduced in conjunction with various mechanical and herbicidal hardwood control treatments.

High rates of seedling survival were achieved at all densities, though rates of grass stage

emergence were low and were not expected to increase until the next harvest cycle (Kirkman et

al. 2007).

The second approach to slash-longleaf conversion, and the one employed in this study, is

to remove the entire slash pine canopy, replant with longleaf pine seedlings, and use mechanical

and herbicidal treatments to control competing vegetation. The natural analog of this type of

clearcut is a very large-scale disturbance, such as a hurricane (Mitchell et al. 2006). Since both

longleaf pine seedlings and native herbaceous species respond positively to full sunlight, this

method, too, has the potential to restore a longleaf pine canopy and a diverse, herbaceous

understory. Some of the potential benefits of this approach, as compared to single-tree and

group-selection methods, are: 1) less logging damage to the site over time because only one









harvest is required, 2) lower management costs, 3) lower personnel skill level and inventory

information needs, 4) higher short-term timber output, and 5) better equipment access and

maneuverability for effective site preparation in an area with severe understory competition, such

as a saw palmetto-dominated flatwoods (Palik et al. 2002, Brockway et al. 2006). If desired,

subsequent management of the new longleaf pine stand can follow a single-tree or group-

selection prescription in order to perpetuate the structural and functional diversity of the system

over the long-term.

The biggest challenges facing this method will be 1) to achieve control of competing

shrubs without negatively impacting native understory vegetation, 2) to maintain or enhance

wiregrass cover, and 3) to successfully apply fire to the system in spite of the loss of overstory

fine fuel inputs. The studies described above indicate that supplemental hardwood control in the

form of herbicide or mechanical treatment will almost always be necessary after clearcutting a

pine overstory, because of the release of hardwoods and the lack of fine fuels for prescribed fire.

Even in a group- or single-tree selection system, additional hardwood control may be necessary

if pyrogenicity is already compromised due to a past history of fire suppression and an

entrenched shrub component. In the single-tree selection study described above, the herbicide

treatment was the only one in which hardwood stem density did not increase over the course of

the study (Kirkman et al. 2007).

Following overstory harvest, longleaf pine plantings must be carefully managed in order

to achieve survival and growth rates adequate for stand replacement. A reasonable range of

survival rates for containerized plantings is 60-75%, but in order to achieve this range of survival

rates, site preparation is necessary before seedlings are planted (Brockway et al. 2006). Grasses

and other herbaceous species are the most serious short-term competitors for newly established









longleaf seedlings, while woody species control is important for the development of fine fuels to

perpetuate the ecosystem in the longer-term (Haywood 2005). On a frequently-burned site

dominated by herbaceous species, prescribed fire may provide adequate site preparation, but on

sites with heavy woody plant competition, mechanical or herbicide treatments are necessary

(Brockway et al. 2006). Intensive mechanical site preparation techniques such as bedding,

disking, and harrowing are effective for longleaf pine establishment (Johnson and Gjerstad

2006), but are very detrimental to wiregrass (Clewell 1989) and are therefore not options on sites

where understory restoration is a goal. Therefore, herbicide, which removes competing

vegetation without disturbing the soil, has potential utility as a component of a sustainable forest

management strategy.

In order to gain a better understanding of the role of herbicide in sustainable longleaf pine

silviculture, we compared the effects of five post-planting, banded herbicide treatments on the

growth and survival of containerized longleaf pine seedlings in a coastal flatwoods in the Florida

panhandle: 1) Sulfometuron methyl (Oust', E.I. du Pont de Nemours and Company,

Wilmington, Del.), 2) Imazapyr (Arsenal', BASF Corporation, Research Triangle Park, NC), 3)

Hexazinone (Velpar L, E.I. du Pont de Nemours and Company, Wilmington, Del.), 4) A tank

mix of Oust' and Velpar L, and 5) a no-herbicide control. These herbicide treatments were

chosen because they are commonly used for loblolly (Pinus taeda L.) and slash pine release in

the Southeast, but their effects on longleaf pine growth and flatwoods community composition

are not well studied. A recent survey by Shepard et al. (2004) reported that the forestry herbicide

most commonly used in the Southeast was Arsenal', followed by tank mixes of Arsenal +

Oust' and Oust' + Velpar. These tank mixes are used to achieve broader spectrum control, but

as a result they may also have more negative implications for native plant communities (Keyser









and Ford 2006). Sulfometuron methyl and hexazinone are effective as over-the-top pine release

applications on longleaf pine seedlings (Haywood 2000, Ramsey and Jose 2004, Haywood

2005), but effects of a mixed application on longleaf seedlings or on the understory of a natural

flatwoods system have not been studied. Imazapyr mixed with triclopyr was effective as a pre-

planting application to release longleaf pine seedlings on a poorly-drained site (Knapp et al.

2006), but information on imazapyr as an over-the-top release treatment for longleaf is also

limited.

We examined the longer-term effects of these herbicide treatments on longleaf pine

seedlings five growing seasons after initial application (see Ranasinghe 2003 for first year

results). We then conducted a dormant-season prescribed fire on all treatment plots and assessed

post-fire survival at the beginning of the sixth growing season.

Methods

Study Area

The study was conducted at the Point Washington State Forest, which is located on the

gulf coast of Walton County, Florida (300 20'16" N, 860 04'19" W). The 20-acre study site is

part of a larger mesic flatwoods/wet flatwoods matrix, dominated by saw palmetto (Serenoa

repens W. Bartram), gallberry (Ilex glabra A. Gray), rhizomatous oaks (Quercus minima Small

and Q. elliottii Wilbur), blueberries (Vaccinium myrsinites Lam. and V. darrowii Camp.),

huckleberries (Gaylussacia dumosa Torr. and A. Gray and G. frondosa L. Torr and A. Gray ex.

Torr), bluestem grasses (Andropogon virginicus var. glaucus (L.) Heck and Andropogon

virginicus L.) and wiregrass (Aristida stricta). The majority of the site can be classified as

mesic, but due to subtle topographical variation across the site some areas are at the wetter end of

the mesic flatwoods spectrum and some are at the drier end. For this reason, we used a

randomized complete block design, with three blocks laid out in mesic-wet areas and three









blocks laid out in mesic-dry areas. A pre-treatment vegetation survey identified wetter areas by

the presence of indicator species such as sedges (Cyperaceae). The soils of the study site are

sandy, siliceous, thermic, aeric alaquods of the Leon series (Spodosols consisting of deep, poorly

to very poorly drained soils derived from marine parent material (Ranasinghe 2003)). Yearly

weather patterns during the study period were variable. Total annual precipitation during the

study period was 103cm in 2002, 200cm in 2003, 135cm in 2004, 146cm in 2005, and 106cm in

2006 (NOAA 2007). Rainfall in the months following the prescribed fire was very low, with

totals of .5cm in March 2007, 4cm in April 2007, and .3cm in May 2007 (NOAA 2007).

Prior to August 2001, the 20-acre study site was a planted slash pine stand with a mean

stand age of about 26 years, which had been burned on a 3-year rotation by the Division of

Forestry since its acquisition of Point Washington State Forest in 1992. A pre-harvest, baseline

data survey was conducted in the areas slated to become study blocks in June 2001. In August

2001, the entire overstory was harvested and the site was prepared for the study with a single,

light roller chopping followed by a prescribed burn in October 2001. In December 2001, six

study blocks, each divided into five 36.6m x 24.4m treatment plots, were established according

to the aforementioned randomized complete block design. Each treatment plot was hand-planted

with 100 containerized longleaf pine seedlings laid out in 10 rows, with 3.1m between rows and

1.8m between the seedlings in each row.

Treatments

Within each block, each 100-seedling plot was randomly selected to receive one of five

herbicide treatments in March of 2002:

1) Sulfometuron methyl (Oius ") at 0.26 kg a.i./ha. This is a moderate application rate

for sulfometuron methyl; other studies have used 0.16 kg a.i./ha (Lauer and Glover, 1998) and

0.21 kg ai/ha (Shiver and Martin 2002, Keyser and Ford 2006), while the manufacturer-









recommended rate for herbaceous weed control in longleaf pine stands is 0.10 to 0.42 kg a.i./ha.

Sulfometuron methyl is a selective herbicide primarily effective for controlling herbaceous

species (Lauer and Glover 1998), and mixing with hexazinone is recommended if broader

spectrum control is desired (DuPont 2007).

2) Hexazinone: (Velpar L) at 0.56 kg a.i./ha. Recommended application rates for

hexazinone range from 2.0 to 6.7 kg a.i./ha (E.I. Du Pont De Nemours), while rates successfully

used to control hardwoods in sandhills ecosystems have ranged from 1.1 kg a.i./ha to 2.4 kg a.i

/ha (Brockway et al. 1998, Provencher et al. 2001). The hexazinone application rate in our study

was therefore very low. This rate was chosen because hexazinone had been deemed effective at

the aforementioned rates, and we were interested in whether an even lower application rate

would provide vegetation control.

3) Sulfometuron + Hexazinone mix (0.26 kg a.i./ha Sulfometuron methyl + 0.56 kg

a.i./ha Hexazinone). This is a common tank mix used throughout the Southeast (Shepard et al.

2004) with the potential to control a broad spectrum of woody and herbaceous plants.

4) Imazapyr (Arsenal) at 0.21 kg a.i./ha. For a restoration project, this was a

moderate application rate for imazapyr. Significant hardwood control and habitat improvements

have been obtained using imazapyr at 0.42 kg a.i./ha (Jones and Chamberlain 2004), and at all

rates from 0.08 kg a.i./ha to 0.24 kg a.i..ha (though higher rates did not offer significant

improvements over lower rates) (Keyser and Ford 2006). Arsenal' is not labeled for use over

longleaf pine at the recommended rates of 0.32 to 0.42 kg a.i./ha until after the end of the second

growing season (BASF 2007), and has caused stunted growth in loblolly pine seedlings (Barber

1991). We chose the rate of 0.21 kg a.i./ha in order to determine whether a lower-than-

recommended application rate would prove less injurious to pine seedlings.









The reasons for using a banded post-planting herbicide application on unshielded

seedlings were both operational and environmental. Prior to harvest, the study site already

contained many desirable understory species (including wiregrass) and only a few weedy

species, none of which were exotic (Ranasinghe 2003). The banded application allowed us to

pinpoint the delivery of herbicide to an area directly around planted seedlings, while native

vegetation remained intact in the strips between rows. This application method also lowered the

cost of herbicide treatment relative to a broadcast application.

The herbicides were applied in a 1.2m band over the seedling rows using a backpack

sprayer. Within each 10-row treatment plot, three rows received a second herbicide application

in March of 2003, three rows received a second application in April of 2003, and four rows were

left with only the first-year treatment. The results from these subplots were pooled with the first

year data for most parameters because of a lack of significant differences between one and two-

year applications.

Data Collection and Analysis

In November 2006, the root collar diameter (RCD) and height to the top of bud were

measured on every seedling. Following a prescribed fire in February of 2007 (Chapter 2), post-

fire seedling mortality was assessed in June 2006. Survival rate was calculated as a percentage

of each 10-tree, 56m2 treatment row. Rate of grass-stage release was calculated as a percentage

of surviving seedlings in each treatment row with a bud height greater than 12cm, following

Haywood (2000). Height and RCD comparisons were made separately for seedlings in the grass

stage (GS) and out of the grass stage (OOGS). The study addressed only the main effects of

herbicide treatment, and tests of these effects were not dependent on the assumption of no

treatment x block interaction. Block effects were therefore treated as random effects in a

univariate ANOVA model with two independent variables: 'treatment' with 'Block&Random' as









a covariate. Data were log-transformed where necessary to meet the assumptions of ANOVA.

Significant differences between treatments were separated with the Tukey-Kramer HSD test.

Post-fire survival was analyzed with ANCOVA, using pre-fire survival as a covariate. All

parameters were analyzed in JMP IN version 5 (SAS Institute, Inc.).

Results

Seedling Survival

Five growing seasons after treatment (November 2006), survival rate varied significantly

among herbicide treatments (a = 0.05, p < 0.0001) (Figure 3-1). The highest survival rates

occurred in the hexazinone and control treatments (80% and 78.5%, respectively) (Figure 3-1).

Survival rates were significantly lower in the sulfometuron + hexazinone (sulfo + hexa) and

imazapyr treatments (64.2% and 63.2%, respectively). It should be noted that the mortality in

the imazapyr and sulfo + hexa treatments occurred shortly after herbicide application

(Ranasinghe 2003), and there was little subsequent mortality in any of the treatments.

None of the treatment plots experienced significant mortality following fire (Figure 3-2).

However, there were some slight differences in fire-related mortality among the different

treatments, and as a result, survival rate no longer varied significantly by treatment following fire

(p = 0.238). In the imazapyr treatment, the survival rate dropped by only .7% following fire, as

compared to decreases of 3% to 3.5% in the other treatments.

Percent Out of Grass Stage

Five growing seasons after herbicide application (November 2006), the mean grass-stage

release rate varied significantly among treatments (p < 0.0001). The highest level of grass-stage

release after four years was observed in the imazapyr treatment, though both sulfo + hexa and

hexazinone also significantly raised the release rate relative to the control (Figure 3-3).









Growth: Out of Grass Stage Seedlings

The mean RCD, height, and SVI of seedlings out of the grass stage (OOGS) varied

significantly among treatments (p = 0.0013, p < 0.0001, and p < 0.0001, respectively). In all

parameters, imazapyr showed the greatest improvement relative to the control, though sulfo +

hexa also significantly improved height relative to the control (Table 3-1, Figure 3-4, Figure 3-

5). RCD of OOGS seedlings varied less among treatments than height or SVI, which is to be

expected due to the preferential investment in height growth characteristic of longleaf pine

seedlings in the "bolting stage."

Growth: Grass Stage Seedlings

Root collar diameter (RCD), height, and SVI of grass-stage seedlings varied significantly

among treatments (p < 0.0001, p = 0.003, and p = 0.0015, respectively) (Table 3-1, Figure 3-5),

with the smallest seedlings in the sulfometuron treatment. Among seedlings that were still in the

grass stage after five growing seasons, none of the herbicide treatments showed significant

improvements relative to the control. This pattern has been observed since the first growing

season after treatment (Ranasinghe 2003), though it was unclear whether the lower seedling size

was due to seedling injury from sulfometuron or the poor shrub control evident in the

sulfometuron treatment since the first growing season (Chapter 2).

Discussion

Five growing seasons after treatment, imazapyr plots had a grass stage release rate more

than twice that of the control, and the mean stem volume index of OOGS seedlings in the

imazapyr treatment was also more than double that in the control. Although direct application of

imazapyr to recently-planted seedlings caused significant mortality in the first year after

planting, the overall post-fire survival rate of 62.5% in this treatment is still within the acceptable

range of 60-75% recommended for adequate overstory restocking (Brockway et al. 2006).









Seedlings in this treatment experienced a fire-related mortality rate of only .06%, whereas fire

mortality was around 3% in all of the other treatments. As a result, the differences in survival

rate between imazapyr and the other treatments were no longer significant following fire.

Though the imazapyr treatment provided the best overall pine release results, both

hexazinone and the sulfometuron + hexazinone tank mix also showed promise as over-the-top

longleaf pine release treatments for flatwoods ecosystems. Significant improvements in the

grass-stage release rate were achieved with both of these treatments, and the OOGS seedlings in

plots treated with the tank mix were significantly taller than those in the control. An absolute

comparison between the effectiveness of these two formulations versus imazapyr is not

warranted because we used a very low rate of hexazinone. However, based on success others

have had using hexazinone for longleaf pine release (Haywood 2000, Ramsey and Jose 2004,

Haywood 2005), we can expect that at a higher rate-both alone and in the tank mix-this

herbicide would have compared more favorably with imazapyr. Contrary to the results

previously observed with longleaf pine seedlings planted in an old field (Ramsey and Jose 2004),

sulfometuron alone did not provide any pine release benefits over the control. This was most

likely due to the higher degree of woody competition on this site, against which sulfometuron

was significantly less effective than the other treatments (Chapter 2).

For grass stage seedlings, variation among treatments was much less pronounced. Grass

stage seedlings in the sulfometuron treatment were significantly smaller than those in the control,

but none of the other treatments differed significantly from the control in any of the grass stage

parameters. A possible explanation for this pattern is that as vegetation regrew in the years

following application, any seedlings not released relatively early in the process were suppressed

and remained in the grass stage. As differences in understory vegetation between treatments









became less pronounced over the years, differences in biomass accumulation rates between

suppressed seedlings probably also decreased. In contrast, the earlier in the process a pine

seedling was released, the better its chances may have been of gaining a competitive advantage

for light, water, and soil resources. The early gains made by seedlings in the imazapyr plots

(Ranasinghe 2003) apparently translated into an increasing competitive advantage over the years,

leading to much larger seedlings in these plots even as vegetation differences between plots

became less pronounced. In a scenario such as this, where the dual goals of silviculture and

ecosystem restoration dictate light site preparation and banded herbicide application in spite of

the heavy flatwoods competition, early release from the grass stage may be critical to the

ultimate success of the planted pines.

The site preparation burn conducted following overstory harvest in 2001 consumed the

last of the accumulated slash pine needles, and, though none of the herbicide treatments

decreased wiregrass cover, distribution of wiregrass across the site was patchy (Chapter 2). Five

years after initial herbicide treatment, we had difficulty applying a prescribed fire of sufficient

intensity to control shrubs-a phenomenon which was undoubtedly exacerbated by length of

time since the application of herbicide and the need to apply fire during the dormant season

(Chapter 2). The lack of fine fuels was also probably a major factor limiting our ability to apply

fire effectively, given the key role played by dead pine needles in the pyrogenicity of a longleaf

pine system (Williamson and Black 1981, Noss 1989). Our results support the theory proposed

by Kirkman et al. (2007) that clearcutting, even when the goal is a total canopy conversion from

slash pine to longleaf pine, may ultimately hinder the success of a restoration project.

The overstory and understory effects of imazapyr, hexazinone, and the sulfometuron +

hexazinone tank mix lined up well with the restoration goals for this site. These treatments,









especially imazapyr, released pines while also maintaining or improving wiregrass cover and

overall cover of native herbaceous species. While tank mixes have been widely studied in

silvicultural applications, they have understandably been avoided as habitat treatments thus far,

out of concern that a broad spectrum mix will negatively impact native species (Keyser and Ford

2006). However, our results indicate that this tank mix not only released pines, but had positive

impacts on the understory, which might have been even more significant if a higher level of

hexazinone had been included in the mix (Chapter 2). An important caveat to this conclusion is

that the use of a banded application method may have been a key to the positive result. On pine

plantations in the Virginia piedmont, broadcast application of an imazapyr + sulfometuron

methyl tank mix significantly lowered herbaceous species diversity relative to the control and to

imazapyr-only treatments, and banded application was suggested as an alternative for sites with

wildlife habitat concerns (Keyser and Ford 2006).

Recent studies of sustainable silviculture in longleaf pine systems have cited the potential

negative impacts of herbicide as a justification for the use of group or single-tree selection, since

these harvesting methods should enhance the effectiveness of prescribed fire as a management

alternative to herbicide (Mitchell et al. 2006, Pecot et al. 2007, Kirkman et al. 2007). However,

we observed only beneficial impacts on understory species following chemical treatment,

indicating that herbicide can in fact play a positive supporting role in conservation-oriented

longleaf pine silviculture. Based on the results of this study, we recommend banded application

of imazapyr as an ecological restoration tool for flatwoods sites with aggressive understory

shrubs.










Table 3-1. Root collar diameter (mm), height (mm), and stem volume index
(basal area2 x height) of grass stage and OOGS seedlings five years after
initial herbicide treatment.
Root Collar
Diameter (mm) Height (mm) Stem Volume Index
Grass Out of Grass Out of Grass Out of
Treatment Stage Grass Stage Stage Grass Stage Stage Grass Stage
Control 25.4a 39.5abc 63.1ab 389.1c 492ab 6332b

Imazapyr 23.6ab 41.8a 70.2a 663.6a 467ab 14,542a

Sulfometuron 22.3b 38.6c 62.2b 490.8bc 388b 9437b

Sulfo + Hexa 25.0a 39.4bc 71.0a 513.1b 519a 10,042b

Hexazinone 25.1a 40.6ab 65.4ab 484.1bc 499a 6332b
ANOVA p< p= p= p< p= P<
0.0001 0.0013 0.003 0.0001 0.0015 0.0001
*Letters denote significant differences between means in the same column at
a = 0.05 based on Tukey-Kramer HSD test. P-values are for randomized
block ANOVA within columns (significant at a = 0.05).
















90%
a a
5 80% a

S70% b
60%

50%
40%

30% -
20% .
Control Imazapyr Sulfo Sulfo + Hexa Hexa
Treatment
Figure 3-1. Mean survival rate by treatment after five growing seasons, calculated as # surving /
# planted in each treatment row. Means not sharing the same letter are significantly
different (Tukey-Kramer HSD).













90.0%
80.0%
70.0%
2 60.0%
50.0% -O Pre-Fire
S40.0% Post-Fire
S30.0%
t 20.0%
10.0%
0.0%
Control Imazapyr Sulfo Sulfo+Hexa Hexa
Treatment

Figure 3-2. Percent seedling survival (trees/row), before and after fire. None of the treatments
experienced significant mortality due to fire (Tukey-Kramer HSD, a = 0.05).














80%
70% a
7 ab
60% b
50% -
40% C
30% -
o 20% -
10% -
0%
Control Imazapyr Sulfo Sulfo + Hexa Hexa
Treatment

Figure 3-3. Mean grass-stage release rate by treatment after five growing seasons, calculated as #
released / # surviving in each treatment row. Means not sharing the same letter are
significantly different (Tukey-Kramer HSD).
















800 -

700 a

S600 -

500 be

400 -

300 I

200

100
Control Imazapyr Sulfo Sulfo + Hexa Hexazinone
Treatment

Figure 3-4. Mean height (mm) of out of grass stage seedlings by treatment. Means not sharing
the same letter are significantly different (Tukey-Kramer HSD, a = 0.05)























a)


Grass Stage


Control Imazapyr


Sulfo Sulfo + Hexa Hexazinone


Treatment


Control Imazapyr Sulfo Sulfo + Hexa Hexazinone
Treatment

Figure 3-5. Mean stem-volume index (SVI) of a) grass stage seedlings and b) OOGS seedlings
by treatment. Means not sharing the same letter are significantly different (Tukey-
Kramer HSD, a = 0.05).


18000
16000
14000
12000
10000
b) so000
6000
4000
2000
0


Out of
Grass Stage









CHAPTER 4
SUMMARY AND CONCLUSIONS


One of the biggest challenges to successful restoration of the longleaf pine (Pinus

palustris Mill.) ecosystem is the persistence of shrubs in the understory, which suppress longleaf

pine seedlings as well as native herbaceous plants. Herbicide can be used as a supplement to fire

in order to enhance shrub control, but must be studied carefully because of the potential for

negative impacts on native plants. Information is lacking about the effects of herbicide on

natural longleaf pine flatwoods communities. We used a banded application of three herbicides

and one tank mix as shrub control treatments following harvest of a slash pine stand and

replanting with containerized longleaf pine seedlings in a mesic-wet flatwoods. The herbicides

tested were Arsenal' (imazapyr), Oust' (sulfometuron methyl), Velpar L (hexazinone), and an

Oust' + Velpar L tank mix.

Imazapyr herbicide significantly improved longleaf pine seedling growth, due to its

effectiveness at controlling competing shrubs during the first growing seasons after planting.

Though these shrub control effects were short-lived, the short term shrub control allowed

herbaceous species to gain more of a foothold in the understory, and significant increases in

herbaceous cover relative to the control were recorded in imazapyr-treated plots four years after

initial treatment. In a longleaf pine silviculture project with both timber and conservation goals,

this was a very desirable result, because of the high level of biodiversity in the herbacous

component of longleaf pine ecosystems (Walker and Peet 1983, Kirkman et al. 2001). No

negative impacts on understory species richness, diversity, evenness, or community composition

resulted from the application of this herbicide. Though over-the-top application of imazapyr

caused significant longleaf pine seedling mortality, the mean survival rate of seedlings in

imazapyr-treated plots was 62.5% after five growing seasons and 61.8% after a fifth-year









prescribed fire, which is still within the acceptable range of 60-75% recommended for adequate

overstory restocking (Brockway et al. 2006). We therefore recommend imazapyr for use in

sustainable longleaf pine silviculture, as it enhanced both longleaf pine seedling growth and

herbaceous cover in the understory.

Hexazinone, both alone and in a tank mix with sulfometuron, also showed promise as

both a longleaf pine release treatment and as an understory restoration tool. Four years after

treatment, hexazinone plots had significantly higher herbaceous cover than the control, and plots

treated with a sulfometuron + hexazinone tank mix had significantly higher wiregrass cover than

the control. Significant improvements in the grass stage release rate were achieved with both of

these treatments, and released seedlings in plots treated with the tank mix were significantly

taller than those in the control. An absolute comparison between these two formulations versus

imazapyr is not warranted because we used a very low rate of hexazinone. However, based on

success others have had using hexazinone for longleaf pine release (Haywood 2000, Ramsey and

Jose 2004, Haywood 2005) and restoration (Brockway et al. 1998, Brockway and Outcalt 2000,

Provencher et al. 2001) we can expect that at a higher rate-both alone and in the tank mix-this

herbicide would have compared more favorably with imazapyr.

Sulfometuron methyl did not improve pine release relative to the control, and instead had

a releasing effect on the two most common oak species (Quercus minima and Quercus elliottii),

indicating that this is not an appropriate herbicide for flatwoods with a high density of oaks.

However, in spite of having higher shrub cover and lower herbaceous cover than the other

herbicide treatments, sulfometuron plots also had the highest herbaceous species richness,

diversity, and evenness four years after initial treatment. One reason for this may have been that

wiregrass cover in the sulfometuron plots was somewhat lower than the other herbicide









treatments, leaving more space for other small herbaceous species. These gains in species

richness, diversity, and evenness are therefore unlikely to persist through future fire cycles

because it will be more difficult to apply prescribed fire to this treatment. In support of this

prediction, we observed that sulfometuron was the only treatment in which species richness,

diversity, and evenness did not increase following the prescribed fire in 2007.

The long-term sustainability of any longleaf pine ecosystem restoration project depends

on the manager's ability to apply prescribed fire with enough frequency and intensity to control

shrubs. The herbicides applied in this study provided short-term shrub control and had positive

impacts on longleaf pine growth, wiregrass cover, and overall herbaceous cover. However,

complete removal of the slash pine overstory at the start of the project resulted in a lack of fine

fuels and presumably also encouraged aggressive regrowth of shrubs due to the loss of

competition from mature pines (Brockway et al. 1998, Kush et al. 1999, McGuire et al. 2001,

Kirkman et al. 2007, Pecot et al. 2007). These hindrances were most likely exacerbated by the

length of time since herbicide application (five years since first application and four years since

second application), as well as the use of dormant-season fire due to the site's proximity to a

highway and residential areas (Waldrop et al. 1992, Drewa et al. 2002, Drewa et al. 2006).

Though three of the four herbicides we tested were successful at releasing longleaf pine

seedlings, the herbicide-related gains in wiregrass and overall herbaceous cover may be lost in

the coming years due to the aggressive resprouting of shrubs. We therefore make the following

recommendation to those managing flatwoods sites for both timber and biodiversity: when

possible, some overstory pines should be retained following harvest to serve as a source of fine

fuels. Banded application of imazapyr is recommended in order to improve longleaf pine









seedling growth and enhance herbaceous cover in the understory, especially on sites where

growing-season fire is not an option.











APPENDIX
SPECIES LIST


Table A-1. Species list.
Species name
Andropogon virginicus var. glaucus
Andropogon virginicus var. virginicus
Aristida spiciformis
Aristida stricta
Asclepias cinerea
Balduina uniflora
Baptisia lanceolata
Calamovilfa curtissii
Carphephorus odoratissimus
C( I, % gossypina subsp. cruseana
(C ,Ii. -'.' mariana
Cladoniaceae
Ctenium aromaticum
Cynanchum i,, .I'4, h ,iii. tl
Cyrilla racemiflora
Desmodium lineatum
Desmodium strictum
Dichanthelium aciculare
Dichanthelium ensifolium
Dichanthelium ovale
Dichanthelium erectifolium
Dichanthelium strigosum
Drosera capillaris
Eryngium yuccifolium
Eupatorium capillifolium
Eupatorium compostifolium
Eupatorium mohrii
Eupatorium pilosum
Eupatorium rotundifolium
Euphorbia inundata
Eurybia eryngiifolia
Euthamia graminifolia
Gaylussacia dumosa
Gaylussaciafrondosa
Gelsemium sempervirens
Gratiola hispida
Houstonia procumbens
Hypericum brachyphyllum
Hypericum hypericoides
Hvpoxis sessilus


Common name
Chalky Bluestem
Broomsedge Bluestem
Bottlebrush Threeawn
Wiregrass
Carolina Milkweed
Oneflower Honeycombhead
Gopherweed
Curtiss's Sandgrass
Deer's Tongue
Cruise's Goldenaster
Maryland Goldenaster
Deer Moss
Toothache Grass
Gulf Coast Swallowwort
Swamp Titi
Sand Ticktrefoil
Pinebarren Ticktrefoil
Needleleaf Witchgrass
Cypress Witchgrass
Eggleaf Witchgrass
Erectleaf Witchgrass
Roughhair Witchgrass
Pink Sundew
Button Rattlesnakemaster
Dog Fennel
Yankeeweed
Mohr's Thoroughwort
Rough Boneset
Roundleaf Thoroughwort
Florida Pineland Spurge
Thistleleaf Aster
Flattop Goldenrod
Dwarf Huckleberry
Dangleberry
Yellow Jessamine
Rough Hedgehyssop
Roundleaf Bluet
Coastalplain St. John's Wort
St. Andrew's Cross
Yellow Star Grass


Family
Poaceae
Poaceae
Poaceae
Poaceae
Apocynaceae
Asteraceae
Fabaceae
Poaceae
Asteraceae
Asteraceae
Asteraceae
Cladoniaceae
Poaceae
Apocynaceae
Cyrillaceae
Fabaceae
Fabaceae
Poaceae
Poaceae
Poaceae
Poaceae
Poaceae
Droseraceae
Apiaceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Euphorbiaceae
Asteraceae
Asteraceae
Ericaceae
Ericaceae
Gelsemiaceae
Veronicaceae
Rubiaceae
Clusiaceae
Clusiaceae
Hypoxidaceae


Nomenclature follows Wunderlin (2004). Species are divided into the following life forms: Grass =
member of the Poaceae family, Sedge = member of the Cyperaceae family, Monocot = other monocot,
Legume = member of the Fabaceae family, Carnivorous = carnivorous forb, Forb = other forb, Fern =
member of the order Filicophyta, Shrub = woody understory species, Tree = woody overstory species.


Life Form
Grass
Grass
Grass
Grass
Forb
Forb
Legume
Grass
Forb
Forb
Forb
Moss
Grass
Forb
Shrub
Legume
Legume
Grass
Grass
Grass
Grass
Grass
Carnivorous
Forb
Forb
Forb
Forb
Forb
Forb
Forb
Forb
Forb
Shrub
Shrub
Vine
Forb
Forb
Shrub
Shrub
Monocot












Table A-1. Continued.
Species name
Ilex coriacea
Ilex glabra
Ilex opaca
Kalmia hirsuta
Lachnanthes caroliniana
Lachnocaulon auceps
Liatris gracilis
Liatris tenuifolia
Lycania michauxii
Lyonia lucida
Mimosa quadrivalvis
Mitreola sessilifolia
Muhlenbergia capillaris
Opuntia humifusa
Orchidaceae
Osmunda cinnamomea
Photinia pyrifolia
Pinus elliottii
Piteopsis graminifolia
Polygala lutea
Polygala nana
Polygonella polygama
Pteridium aquilinum
Pterocaulon pychnostachium
Quercus elliottii
Quercus incana
Quercus minima
Rhexia alifanus
Rhexia petiolata
Rhynchospora Jlitlia,
Rhynchospora plumosa
Sabatia brevifolia
Scleria ciliata
Serenoa repens
Seriocarpus tortifolius
Smilax auriculata
Smilax glauca
Smilax laurifolia
Smilax pumila
Solidago odora
Sporobolusjunceus
\ihll,,, 1 sylvatica
Stylisma patens
Stylisma villosa
Symphiotrichum adnatum
Tephrosia hispidula
Tragia urens
Vaccinium darrowii
Vaccinium elliottii
Vaccinium myrsinites
Vernonia angustifolia
Vitis rotundifolia
Xyris caroliniana


Common name
Giant Gallberry
Gallberry
American Holly
Hairy Wicky
Carolina Redroot
Whitehead Bog Button
Slender Blazing Star
Shortleaf Blazing Star
Gopher Apple
Fetterbush
Sensitive Vine
Swamp Hompod
Cutover Muhly
Prickly Pear
Orchid sp.
Cinnamon Fern
Red Chokeberry
Slash Pine
Grassleaf Goldenaster
Orange Milkwort
Candyroot
October Weed
Bracken Fern
Blackroot
Runner Oak
Bluejack Oak
Dwarf Live Oak
Savannah Meadow-Beauty
Fringed Meadow-Beauty
Threadleaf Beaksedge
Plumed Beaksedge
Shortleaf Rosegentian
Fringed Nutrush
Saw Palmetto
Whitetop Aster
Earleaf Greenbrier
Wild Sarsparilla
Laurel-leaf Greenbrier
Sarsparilla Vine
Sweet Goldenrod
Pineywoods Dropseed
Queen's Pleasure
Coastalplain Dawnflower
Hairy Dawnflower
Scaleleaf Aster
Sprawling Hoarypea
Wavyleaf Noseburn
Darrow's Blueberry
Elliott's Blueberry
Shiny Blueberry
Tall Ironweed
Muscadine Grape
Carolina Yelloweyed Grass


Family
Aquifoliaceae
Aquifoliaceae
Aquifoliaceae
Ericaceae
Haemodoraceae
Eriocaulaceae
Asteraceae
Asteraceae
Chrysobalanaceae
Ericaceae
Fabaceae
Loganiaceae
Poaceae
Cactaceae
Orchidaceae
Osmundaceae
Rosaceae
Pinaceae
Asteraceae
Polygalaceae
Polygalaceae
Polygonaceae
Dennstaedtiaceae
Asteraceae
Fagaceae
Fagaceae
Fagaceae
Melastomataceae
Melastomataceae
Cyperaceae
Cyperaceae
Gentianaceae
Cyperaceae
Aracaceae
Asteraceae
Smilacaceae
Smilacaceae
Smilacaceae
Smilacaceae
Asteraceae
Poaceae
Euphorbiaceae
Convolvulaceae
Convolvulaceae
Asteraceae
Fabaceae
Euphorbiaceae
Ericaceae
Ericaceae
Ericaceae
Asteraceae
Vitaceae
Xyridaceae


Life Form
Shrub
Shrub
Shrub
Shrub
Monocot
Monocot
Forb
Forb
Forb
Shrub
Legume
Forb
Grass
Forb
Monocot
Fern
Shrub
Tree
Forb
Forb
Forb
Forb
Fern
Forb
Shrub
Tree
Shrub
Forb
Forb
Sedge
Sedge
Forb
Sedge
Shrub
Forb
Monocot
Monocot
Monocot
Monocot
Forb
Grass
Forb
Forb
Forb
Forb
Legume
Forb
Shrub
Shrub
Shrub
Forb
Vine
Monocot









LIST OF REFERENCES


BASF Corporation. 2007. Arsenal' AC label. http://www.cdms.net/LDat/ld542007.pdf. BASF
Agricultural Products, Research Triangle Park, NC.

Behm, A.L., Duryea, M.L., Long, A.J., and W.C. Zipperer. 2004. Flammability of native
understory species in pine flatwood and hardwood hammock ecosystems and implications for
the wildland-urban interface. International Journal of Wildland Fire 13:355-365.

Boyd, R.S., Freeman, J.D., Miller, J.H., and M.B. Edwards. 1995. Forest herbicide influences
on floristic diversity seven years after broadcast pine release treatments in central Georgia,
U.S.A. New Forests 10:17-37.

Boyer, W.D. 1963. Development of longleaf pine seedlings under parent trees. USDA Forest
Service, Southern Forest Experiment Station, Research Paper SO-4, New Orleans, LA.

Brockway, D.G. and C.E. Lewis. 1997. Long-term effects of dormant season prescribed fire on
plant community diversity, structure, and productivity in a longleaf pine-wiregrass
ecosystem. Forest Ecology and Management 96:167-183.

Brockway, D.G. and K.W. Outcalt. 1998. Gap-phase regeneration in longleaf pine wiregrass
ecosystems. Forest Ecology and Management 106:125-139.

Brockway, D.G., Outcalt, K.W., and R.N. Wilkins. 1998. Restoring longleaf pine wiregrass
ecosystems: plant cover, diversity, and biomass following low-rate hexazinone application on
Florida sandhills. Forest Ecology and Management 103:159-175.

Brockway, D.G. and K.W. Outcalt. 2000. Restoring longleaf pine wiregrass ecosystems:
hexazinone application enhances effects of prescribed fire. Forest Ecology and Management
137:121-138

Brockway, D.G., Outcalt, K.W., and W.D. Boyer. 2006. Pages 95-133 in Jose, S., Jokela, E.,
and Miller, D.L., editors. The Longleaf Pine Ecosystem: Ecology, Silviculture, and
Restoration. Springer Science + Business Media, LLC., New York.

Clewell, A.F. 1989. Natural history of wiregrass (Aristida strict Michx., Graminae). Natural
Areas Journal 9:234-245.

Croker, T.C. Jr. 1975. Seedbed preparation aids natural regeneration of longleaf pine. USDA
Forest Service Research Paper. Southern Forest Experiment Station No. SO-112.

Daubenmire, R.F. 1959. Canopy coverage method of vegetation analysis. Northwest Scientist
33:43-64.

Drewa, P.B., Platt, W.J., and E.B. Moser. 2002. Fire effects on resprouting of shrubs in
headwaters of southeastern longleaf pine savannas. Ecology 83(3):755-767.









Drewa, P.B., Thaxton, J.M., and W.J. Platt. 2006. Responses of root-crown bearing shrubs to
differences in fire regimes in Pinuspalustris (longleaf pine) savannas: Exploring old- growth
questions in second growth systems. Applied Vegetation Science 9:27-36.

Dufrene, M. & P. Legendre. 1997. Species assemblages and indicator species: the need for a
flexible asymmetrical approach. Ecological Monographs 67:345-366.

DuPontTM. 2007. Velpar L Label. http://www.dupont.com/ag/vm/products/velparl.html.
E. I. du Pont de Nemours and Company,Wilmington, DE.

DuPontTM. 2007. Oust' XP Label.
http://www.dupont.com/ag/us/prodinfo/prodsearch/information/H65144.pdf
E. I. du Pont de Nemours and Company,Wilmington, DE.

Frost, C.C. 2006. History and Future of the Longleaf Pine Ecosystem. Pages 9-47 in Jose, S.,
Jokela, E., and Miller, D.L., editors. The Longleaf Pine Ecosystem: Ecology, Silviculture,
and Restoration. Springer Science + Business Media, LLC., New York.

Gagnon, J.L., Jokela, E.J., Moser, W.K., and D.A. Huber. 2003 Dynamics of artificial
regeneration in gaps within a longleaf pine flatwoods ecosystem. Forest Ecology and
Management 172:133-144.

Gilliam, F.S. and W.J. Platt. 1999. Effects of long-term fire exclusion on tree species
composition and stand structure in an old-growth Pinuspalustris forest. Plant Ecology
140:15-26

Glitzenstein, J.S., Streng, D.R., and D.D. Wade. 2003. Fire frequency effects on longleaf pine
(Pinuspalustris Mill.) vegetation in South Carolina and northeast Florida, USA. Natural
Areas Journal 23(1):22-37.

Grace, S.L., and W.J. Platt. 1995. Effects of adult tree density and fire on the demography off
pregrass stage juvenile longleaf pine (Pinuspalustris Mill.) Journal of Ecology 83:75-86.

Hay-Smith, L. and G.W. Tanner. 1994. Responses of wildlife food plants to a forest herbicide.
University of Florida Cooperative Extension Fact Sheet WRS-9.

Haywood, J.D. 2000. Mulch and hexazinone herbicide shorten the time longleaf pine seedlings
are in the grass stage and increase height growth. New Forests 19:279-290.

Haywood, J.D. 2005. Effects of herbaceous and woody plant control on Pinuspalustris growth
and foliar nutrients through six growing seasons. Forest Ecology and Management 214:384-
397.

Harcombe, P.A., Glitzenstein, J.S., Knox, R.G., Orzell, S.L., E.L. Bridges. 1993. Longleafpine
vegetation of the Southern Atlantic and Eastern Gulf Coast regions: a preliminary
classification. Proceedings of the Tall Timbers Fire Ecology Conference 18:83-104









Hiers, J.K., Wyatt, R., and R.J. Mitchell. 2000. The effects of fire regime on legume
reproduction in longleaf pine savannas: is a season selective? Oecologia 125:521-530.

Hurst, G.A. and W.E. Palmer. 1987. Vegetation responses to imazapyr for pine release.
Proceedings of the Southern Weed Science Society 40:247.

Hobbs, R.J., Currall, J.E.P., and C.H. Gimingham. 1984. The use of"thermocolor" pyrometers
in the study of heath fire behavior. Journal of Ecology 72:241-250.

Iverson, L.R., Yaussy, D.A., Rebbeck, J., Hutchinson, T.F., Long, R.P., and A.M. Prasad. 2004.
A comparison of thermocouples and temperature paints to monitor spatial and temporal
characteristics of landscape-scale prescribed fires. International Journal of Wildland Fire.
13(3):311-322.

Jacqmain, E.I., Jones, R.H., and R.J. Mitchell. 1999. Influences of frequent cool-season burning
across a soil moisture gradient on oak community structure in longleaf pine ecosystems.
American Midland Naturalist 141:85-100.

Jones, J.D., and M.J. Chamberlain. 2004. Efficacy of herbicides and fire to improve vegetative
conditions for northern bobwhites in mature pine forests. Wildlife Society Bulletin
32(4):1077-1084.

Johnson, R. and D.J. Gjerstad. 2006. Restoring the overstory of longleaf pine ecosystems.
Pages 271-295 in Jose, S., Jokela, E., and Miller, D.L., editors. The LongleafPine
Ecosystem: Ecology, Silviculture, and Restoration. Springer Science + Business Media,
LLC., New York.

Keyser, P.D., and V.L. Ford. 2006. Wildlife habitat and herbicides: an evaluation of a widely
applied tank mix. Southern Journal of Applied Forestry 30(1):46-51.

Keddy, P.A., Smith, L., Campbell, D.R., Clark, M., and G. Montz. 2006. Patterns of herbaceous
plant diversity in a Louisiana pine savanna. Applied Vegetation Science 9:17-26.

Kirkman, L.K., Mitchell, R.J., Helton, R.C., and M.B. Drew. 2001. Productivity and species
richness across and environmental gradient in a fire-dependent ecosystem. American Journal
of Botany 88(11):2119-2128.

Kirkman, L.K., Mitchell, R.J., Kaiser, M.J., Pecot, S.D., and K.L. Coffey. 2007. The perpetual
forest: using undesirable species to bridge restoration. Journal of Applied Ecology 44:604-
614.

Knapp, B.O., Wang, G.G., Walker, J.L., and S. Cohen. 2006. Effects of site preparation
treatments on early growth and survival of planted longleaf pine (Pinuspalustris Mill.)
seedlings in North Carolina. Forest Ecology and Management 226:122-128.









Kush, J.S., R.S. Meldahl, and W.D. Boyer. 1999. Understory plant community response after
23 years of hardwood control treatments in natural longleaf pine (Pinuspalustris) forests.
Canadian Journal of Forest Research 29:1047-1054.

Lauer, D.K. and G.R. Glover. 1998. Early pine response to control of herbaceous and shrub
vegetation in flatwoods. Southern Journal of Applied Forestry 15(4):201-208.

Leach, M.K. and T.J. Givnish. 1996. Ecological determinants of species loss in remnant
prairies. Science 273:1555-1558.

Litt, A.R., Herring, B.J. and L. Provencher. 2001. Herbicide effects on ground-layer vegetation
in southern pinelands, USA: a review. Natural Areas Journal 21(2):177-188.

McNease. L.L. and G.A. Hurst. 1991. Preharvest treatment with imazapyr for seed tree
regeneration and habitat improvement. Proceedings of the Southern Weed Science Society
44:265.

Mehlman, D.W. 1992. Effects of fire on plant community composition of North Florida second
growth pineland. Bulletin of the Torrey Botanical Club 119:376-383.

McCune, B. and Grace, J. 2002. Analysis of Ecological Communities. MJM Software Design.
Gleneden Beach, OR.

McGuire, J.P., Mitchell, R.J., Moser, E.B., Pecot, S.D., Gjerstad, D.H., and C.W. Hedman.
2001. Gaps in a gappy forest: plant resources, longleaf pine regeneration, and understory
response to tree removal in longleaf pine savannas. Canadian Journal of Forest Research
31:765-778.

Miller, J.H., R.S. Boyd, and M.B. Edwards. 1999. Floristic diversity, stand structure, and
composition 11 years after herbicide site preparation. Canadian Journal of Forest Research
29:1073-1083.

Mitchell, R.J., Hiers, J.K., O'Brien, J.J., Jack, S.B., and R.T. Engstrom. 2006. Silviculture that
sustains: the nexus between silviculture, frequent prescribed fire, and conservation of
biodiversity in longleaf pine forests of the southeastern United States. Canadian Journal of
Forest Research 36:2724-2736.

Neary, D.G., J.E. Smith, B.F. Swindel, and K.V. Miller. 1991. Effects of forestry herbicides on
plant species diversity. Proceedings of the Southern Weed Science Society 44:266-272.

NOAA Satellite and Information Service: National Climatic Data Center. 2007. Panama City,
FL Station 5N.
http://www4.ncdc.noaa.gov/cgiwin/wwcgi.dll?wwDI~StnSrch-StnID~20004593.

Noss, R.F. 1989. Longleaf pine and wiregrass: keystone components of an endangered
ecosystem. Natural Areas Journal 9:211-213.









Noss, R.F., LaRoc, III, E.T., and J.M. Scott. 1995. Endangered ecosystems of the United States:
a preliminary assessment of loss and degradation. USDI National Biological Service,
Washington, D.C., Biological Report No. 28.

Olson, M.S. and W.J. Platt. 1995. Effects of habitat and growing season fires on resprouting of
shrubs in longleaf pine savannas. Vegetatio 119:101-118.

Outcalt, K.W., Williams, M.E., and 0. Onokpise. 1999. Restoring Aristida strict to Pinus
palustris ecosystems on the Atlantic Coastal Plain, U.S.A. Restoration Ecology 7(3):262-
270.

Peet, R.K., Wentworth, T.R., and P.S. White. 1998. Method for recording vegetation
composition and structure. Castanea 63:262-274.

Palik, B.J., Mitchell, R.J., Houseal, G., and N. Pederson. 1997. Effects of canopy structure on
resource availability and seedling responses in a longleaf pine ecosystem. Canadian Journal
of Forest Research 27:1458-1464.

Palik, B.J., Mitchell, R.J., J.K. Hiers. 2002. Modeling silviculture after natural disturbance to
sustain biodiversity in the longleaf pine (Pinuspalustris) ecosystem: balancing complexity
and implementation. Forest Ecology and Management 155:347-356.

Pecot, S.D., Mitchell, R.J., Palik, B.J., Moser, E.B., and J.K. Hiers. 2007. Competitive
responses of seedlings and understory plants in longleaf pine woodlands: separating canopy
influences above and below ground. Canadian Journal of Forest Research 37:634-648.

Peet, R.K. 2006. Ecological classification of longleaf pine woodlands. Pages 51-93 in Jose, S.,
Jokela, E., and Miller, D.L., editors. The LongleafPine Ecosystem: Ecology, Silviculture,
and Restoration. Springer Science + Business Media, LLC., New York.

Peet, R.K. and D.J. Allard. 1993. Longleaf pine-dominated vegetation of the southern Atlantic
and eastern Gulf Coast region, USA. Proceedings of the Tall Timbers Fire Ecology
Conference 18:45-81.

Platt, W.J. and S.L. Rathbun. 1993. Dynamics of an old-growth longleaf pine population.
Proceedings of the Tall Timbers Fire Ecology Conference 18:275-297.

Platt, W.J., Beckage, B., Doren, R.F. & H.H. Slater. 2002. Interactions of large-scale
disturbances: prior fire regimes and hurricane mortality of savanna pines. Ecology 83:1566-
1572.

Platt, W.J., Carr, S.M., Reilly, M., and J. Fahr. 2006. Pine savanna overstorey influences on
ground-cover biodiversity. Applied Vegetation Science 9:37-50.









Provencher, L., Herring, B.J., Gordon, D.R., Rogers, H.L., Galley, K.E.M., Tanner, G.W.,
Hardesty, J.L., and L.A. Brennan. 2001. Effects of hardwood reduction techniques on
longleaf pine sandhill vegetation in northwest Florida. Restoration Ecology 9:13-27.

Ramsey, C.L., S. Jose, B.J. Brecke, and S. Merritt. 2003. Growth response of longleaf pine
(Pinuspalustris Mill.) seedlings to fertilization and herbaceous weed control in an old field
in southern USA. Forest Ecology and Management 172(2-3):281-289.

Ramsey, C.L., and S. Jose. 2004. Growth, survival, and physiological effects of hexazinone and
sulfometuron methyl applied overtop of longleaf pine seedlings. Southern Journal of
Applied Forestry 28(1):48-54.

Ranasinghe, S. 2003. Role of herbicides in longleaf pine flatwoods restoration: pine growth,
understory vegetation response, and fate of applied herbicides. Master's thesis. University
of Florida, Gainesville.

Ranasinghe, S., Jose, S., and C. Ramsey. 2005. Understory species dynamics following
herbicidal release treatments on a longleaf pine site in the Southern U.S.A. In New Research
on Forest Ecosystems, A.R. Burk, ed. Nova Science Publishers, Inc.

Rebertus, A.J., Williamson, G.B., and W.J. Platt. 1993. Impacts of temporal variation in fire
regime on savanna oaks and pines. Proceedings of the Tall Timbers Fire Ecology
Conference 18:215-225.

Shepard, J.P., Creighton, J., and H. Duzan. 2004. Forestry herbicides in the United States: an
overview. Wildlife Society Bulletin 32(4):1020-1027.

Varner III, J.M., and J.S. Kush. 2004. Remnant old-growth longleaf pine (pinus palustris Mill.)
savannas and forests of the Southeastern USA; status and threats. Natural Areas Journal
24:141-149.

Waldrop, T.A., White, D.L., and S.M. Jones. 1992. Fire regimes for pine-grassland
communities in the southeastern United States. Forest Ecology and Management 47:195-
210.

Walker, J., and R.K. Peet. 1983. Composition and species diversity of pine-wiregrass savannas
of the Green Swamp, North Carolina. Vegetatio 55:163-179.

Walker, J. and A.M. Silletti. 2006. Restoring the Ground Layer of LongleafPine Ecosystems.
Pages 297-333 in Jose, S., Jokela, E., and Miller, D.L., editors. The LongleafPine Ecosystem:
Ecology, Silviculture, and Restoration. Springer Science + Business Media, LLC., New York.

Welch, J.R., Miller, K.V., Palmer, W.E., and T.B. Harrington. 2004. Response of understory
vegetation important to the northern bobwhite following imazapyr and mechanical
treatments. Wildlife Society Bulletin 32(4): 1071-1076.









Wilkins, R.N., W.R. Marion, D.G. Neary, and G.W. Tanner. 1993. Vascular plant community
dynamics following hexazinone site preparation in the lower Coastal Plain. Canadian
Journal of Forest Research 23:2216-2229

Williamson, G.B. and E.M. Black. 1981. High temperature of forest fires under pines as a
selective advantage over oaks. Nature 293:643-644.









BIOGRAPHICAL SKETCH

Johanna Freeman is a native of western New York state. She received a B.S. in natural

resources from Cornell University in 2001, where her studies focused on animal behavior and

environmental policy. She worked as a land-use planner at the New York City Department of

Parks and Recreation for three years prior to entering the masters program in interdisciplinary

ecology at the University of Florida.





PAGE 1

FROM PINUS ELLIOTTII PLANTATION TO PINUS PALUSTRIS ECOSYSTEM: THE ROLE OF HERBICIDE IN LONGLEAF PINE RESTORATION By JOHANNA E. FREEMAN 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 2008 1

PAGE 2

2008 Johanna Freeman 2

PAGE 3

ACKNOWLEDGMENTS I would like to thank Dr. Shibu Jose, my s upervisory committee chair, for the opportunity to conduct this research, and for hi s cheerful support and advice. I would also like to thank Drs. Alan Long and Debbie Miller for serving on my supervisory committee, providing instruction on the identification and ec ology of North Florida vegetation, and assisting me in developing fire ecology research methods. Thanks also go to Dr. Eric Holzmueller for extensive assistance with field data collection and anal ysis throughout the pr oject; to Don Hagan, Michael Morgan, and Kelly Thayer for assisting me with field data co llection; and to Brian Hinton for assistance with fire methods. I thank the University of Florid a herbarium staff, Dr. Walter Judd, and especially Dr. Susan Carr for invaluable help with the id entification of plant sp ecimens. I also thank Meghan Brennan of the IFAS statistical consulti ng unit, who assisted me greatly with data analysis. Thanks go to the Florida Division of Forestry for funding this project and providing the study site, and especially to site administrator Tom Beitzel and hi s crew for their assistance in collecting fire temperature data. Finally, I woul d like to thank my parents and David Kaplan for their constant support and encouragement. 3

PAGE 4

TABLE OF CONTENTS page ACKNOWLEDGMENTS .............................................................................................................. 3 LIST OF TABLES .......................................................................................................................... 6 LIST OF FIGURES ........................................................................................................................ 7 ABSTRACT .................................................................................................................................... 8 CHAPTER 1 INTRODUCTION .................................................................................................................. 10 2 USING HERBICIDE TO RESTORE THE UNDERSTORY FOLLOWING HARVEST .... 13 Introduction ............................................................................................................................. 13 Methods ................................................................................................................................... 18 Study Area ........................................................................................................................ 18 Treatments ......................................................................................................................... 19 Data Collection and Analysis ............................................................................................ 21 Results ..................................................................................................................................... 25 Shrub Cover ...................................................................................................................... 25 Herbaceous Cover ............................................................................................................. 25 Wiregrass Cover ................................................................................................................ 26 Shrub Stem Density .......................................................................................................... 26 Shrub height ...................................................................................................................... 27 Understory responses ........................................................................................................ 27 Species Diversity .............................................................................................................. 28 Species Evenness .............................................................................................................. 28 Community Ordination ..................................................................................................... 29 Species of Special Interest ................................................................................................ 29 Fire Temperature ............................................................................................................... 29 Discussion ............................................................................................................................... 30 3 HERBICIDE AND SUSTAINABLE L ONGLEAF PINE SILVICULTURE ........................ 51 Introduction ............................................................................................................................. 51 Methods ................................................................................................................................... 57 Study Area ........................................................................................................................ 57 Treatments ......................................................................................................................... 58 Data Collection and Analysis ............................................................................................ 60 4

PAGE 5

Results ..................................................................................................................................... 61 Seedling Survival .............................................................................................................. 61 Percent Out of Grass Stage ............................................................................................... 61 Growth: Out of Grass Stage Seedlings ............................................................................. 62 Growth: Grass Stage Seedlings ........................................................................................ 62 Discussion ............................................................................................................................... 62 4 SUMMARY AND CONCLUSIONS ..................................................................................... 72 APPENDIX SPECIES LIST........................................................................................................................ 76 LIST OF REFERENCES .............................................................................................................. 78 BIOGRAPHICAL SKETCH ........................................................................................................ 85 5

PAGE 6

LIST OF TABLES Table page 2-1 Shrub cover, stem density, and height. ............................................................................. 37 2-2 Indicator species analysis .................................................................................................. 38 2-3 Understory species rich ness, diversity, and evenness ....................................................... 39 2-4 Mean maximum fire temperature (C) by treatment ......................................................... 40 3-1 Pine seedling growth after five seasons ............................................................................ 66 6

PAGE 7

LIST OF FIGURES Figure page 2-1 Ericaceae cover ........................................................................................................ ....... 41 2-2 Overall shrub cover. .......................................................................................................... 42 2-3 Fire effects on shrub cover. ............................................................................................... 43 2-4 Herbaceous plant cover ................................................................................................... .. 44 2-5 Changes in cover of plant func tional groups before and after fire ................................... 45 2-6 Wiregrass cover ................................................................................................................ 46 2-7 Shrub stem density ....................................................................................................... ..... 47 2-8 Shannon Diversity ............................................................................................................. 48 2-9 Ordination of plant communities ...................................................................................... 49 2-10 Relationship of fire temperat ure to dominant shrub species. ............................................ 50 3-1 Pine seedling survival. ...................................................................................................... 67 3-2 Post-fire pine seedling survival. ........................................................................................ 68 3-3 Rates of grass stage release by treatment. ......................................................................... 69 3-4 Pine seedling height .......................................................................................................... 70 3-5 Stem volume index of grass st age and out of grass stage seedlings ................................. 71 7

PAGE 8

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 FROM PINUS ELLIOTTII PLANTATION TO PINUS PALUSTRIS ECOSYSTEM: THE ROLE OF HERBICIDE IN LONGLEAF PINE RESTORATION By Johanna E. Freeman May 2008 Chair: Shibu Jose Major: Interdisciplinary Ecology Under a natural fire regime, the longleaf pine ( Pinus palustris Mill.) ecosystem is characterized by an open, longleaf pine-dominated canopy and a diverse, grass-dominated understory. In the years since Eu ropean colonization, the vast majo rity of longleaf pine acreage has been cut over, fire-suppressed, and fragmented, which has resulted in extensive understory invasion by shrubs and a concurrent decline in herbaceous species diversity. One of the biggest challenges to successful restoration of this eco system is the persistence of shrubs in the understory, which suppress longleaf pine seedlin gs as well as native herbaceous plants. Herbicide can be used as a supplement to fire in order to enhance shrub control, but must be studied carefully because of the potential for ne gative impacts on native plants. Information is lacking about the effects of herbicide on natural longleaf pine flatwoods communities. We used a banded application of three he rbicides and one tank mix as sh rub control treatments following complete removal of the slash pine canopy and replanting with contai nerized longleaf pine seedlings in a mesic-wet flatwoods. The herbicides tested were Arsenal (imazapyr), Oust (sulfometuron methyl), Velpar L (hexazinone), and an Oust + Velpar L tank mix. Four years after application, no negative impacts on understory species richness, diversity, evenness, or community composition were 8

PAGE 9

evident in any of the herbicide treatments. Oaks ( Quercus spp.), one of the dominant shrub genera on the study site, were resi stant to sulfometuron methyl, a nd this herbicide was therefore ineffective both as a pine release treatment and for enhancing herbaceous species growth. Imazapyr was the most effective treatment ove rall, significantly improving longleaf pine seedling growth and also enhancing herbaceous species cover. Both hexazinone and the hexazinone + sulfometuron methyl tank mix pr ovided some seedling growth and understory enhancement as well, though these effects were not as pronounced as those in the imazapyr treatment. Shrubs resprouted a ggressively after a dormant-season pr escribed fire in the fifth year after treatment, indicating that herbicide-relate d increases in cover of wiregrass and other herbaceous species may be lost in future fire cycles. Retention of overstory pines for shrub competition and as a source of fine fuels is reco mmended in addition to herbicide for sustainable silviculture in longleaf pine flatwoods. 9

PAGE 10

CHAPTER 1 INTRODUCTION Two hundred years ago, longleaf pine ( Pinus palustris Mill.) ecosystems covered an estimated 37 million hectares of the southeastern United States (Frost 2006). Based on fuel accumulation patterns and plant life hi stories it is believed that light to moderate intensity surface fires burned throughout the range of the longleaf pine every two to three years prior to European colonization (Frost 1996). These frequent fire s maintained an open, longleaf pine-dominated canopy and a grass-dominated understory contai ning some of the most species-rich plant communities ever documented outside the tropics (P eet and Allard 1993). However, in the years following European colonization, fires were supp ressed and old-growth longleaf pine systems were logged almost out of existence. Toda y, it is estimated that only 841,000 hectares of naturally regenerating longleaf pine systems still remain, the great majority of which are second growth and fire-suppressed (Frost 2006). This area contains only about 5,095 hectares of oldgrowth longleaf pine (Varner and Kush 2004). Fire suppression has resulted in extensive understory invasion by hardwood shrubs and a conc urrent regional decline in herbaceous species diversity (Walker and Peet 1983, Mehlman 1992, Gilliam and Platt 1999). In 1995, a U.S. Biological Survey report liste d the longleaf pine ecosystem as the third most endangered ecosystem in the United States (Noss et al. 1995). In recent years, recognition of the value of the longleaf pine ec osystem has motivated widespread restoration efforts throughout the southeast. One of the biggest obstacles to successful longleaf pine ecosystem restoration is the pe rsistence of hardwoods in the understory even after the rein troduction of fire (Walker and Siletti 2006 ). In some cases the reintroduction of fire alone may restore the de sired two-layered savanna struct ure (Kush et al. 1999), but many factors can limit the efficacy of fire after a long period of suppr ession, including insufficient fine 10

PAGE 11

fuels, presence of ladder fuels that may cause damage to crowns, and duff accumulation that can kill overstory trees when ignited (Walker and Siletti 2006). Growth of shrub rhizomes and root systems during periods of fire suppression can be extensive, making these shrubs recalcitrant even when fire is reintroduced (Drewa et al 2006). Additionally, social factors such as proximity to residential areas or highways can limit the ability of land managers to use prescribed fire effectively. Therefore, mechan ical and herbicidal hardwood removal techniques are often considered as supplements to prescr ibed fire (Walker and Siletti 2006). Hardwood control is also of key importance for restorati on projects that include planting of longleaf pine seedlings, which are very intolerant of competition and can quickly succumb to competing vegetation while in the gr ass stage (Harrington 2006). Herbicide, used in conjunction with prescrib ed fire, can be an eff ective tool for reducing midand understory hardwoods in longleaf pine systems without negatively impacting native understory species (Brockway et al 1998, Broc kway and Outcalt 2000, Provencher et al. 2001). However, most studies of herbicide as a restora tion tool in longleaf pine systems have been conducted on xeric longleaf pine si tes (Litt et al. 2001). Relative to xeric longleaf pine sandhills, mesic and wet longleaf pine flatwoods are more productive, have different soil characteristics, and are associated with differe nt understory plant communities Understory responses to herbicide in these systems are therefore likely to differ from those documented in sandhills (Litt et al. 2001). At present, there is only limited information about herbicidal shrub control in mesic to wet longleaf pine ecosystems. The purpose of this research was to test the effects of different herbicide treatments, used in conjunction with prescribed fire, on target and nontarget vege tation in a flatwoods ecosystem undergoing conversion from a planted slash pine ( Pinus elliottii Engelm.) overstory to a longleaf 11

PAGE 12

pine canopy. The study included both overstory and understory components and had the following objectives: To determine whether differences in the leve l of shrub control provided by sulfometuron methyl (Oust), hexazinone (Velpar L), a sulfometuron methyl + hexazinone tank mix, and imazapyr (Arsenal) were evident in the fourth year after application; To determine whether differential treatm ent effects of sulfometuron, hexazinone, a sulfometuron + hexazinone mix, and imazapyr on native understory species richness, diversity, evenness, and community composition were evident in the fourth year after application; To compare the effects of over-the-top applic ations of sulfometuron methyl, hexazinone, a sulfometuron methyl + hexazinone mix, and imaza pyr on the survival and growth of planted longleaf pine seedlings; To observe how the herbicide treatment plots res ponded to a prescribed fi re administered five years after herbicide application. The understory component of the study is describe d in Chapter 2 of this thesis, the overstory component is described in Chapter 3, and overa ll conclusions are discussed in Chapter 4. 12

PAGE 13

CHAPTER 2: USING HERBICIDE TO RESTORE THE UNDERSTORY FOLLOWING HARVEST Introduction Longleaf pine ( Pinus palustris Mill.) ecosystems occupy a wide environmental gradient in the southeastern United States, from xeric sandhills to mesic and wet flatwoods (Walker and Peet 1983, Peet 2006). Under a natural fire regime the understory vegetati on associated with the two dominant species, longl eaf pine and wiregrass (Aristida stricta Michx.), varies predictably with site hydrology (Harcombe et al. 1993). Me sic longleaf pine flatwoods, which occur on Ultisols and Spodosols, fall in th e middle of the longleaf pine hydr ologic gradient and are highly productive plant communities (Peet 2006). Whereas the understories of fre quently burned xeric longleaf pine sandhills are truly dominated by gra sses and herbaceous plants, in mesic flatwoods a large shrub component is present even under na tural fire conditions (Peet 2006). These shrubs are kept low to the ground by frequent fire, but in the absence of fire th ey will quickly succeed into a dense midstory, eventually suppressing both longleaf pine and herbaceous plant regeneration (Gilliam and Platt 1999). Wiregrass is a key structural and functional component of all longleaf pine ecosystems in the Southern Coastal Plain eco region (Peet 2006), and the estab lishment or enhancement of wiregrass populations is essentia l for successful ecological restora tion in these systems. Living and dead wiregrass leaves are highly flammable and act as fire-spreaders, moving surface fires evenly through the understory of a longleaf pine ecosystem (Cle well 1989). Dead longleaf pine needles are also very flammable and are a cri tical component of the systems pyrogenicity (Williamson and Black 1981, Clewell 1989). Both l ongleaf pine and wiregrass can therefore be considered keystone species. Together they perpetuate a frequent fire return interval, which favors their own reproductive success, acts as a selective force against competing, less fire13

PAGE 14

tolerant species, and allows a host of herbaceous understory species to survive and reproduce in the open, high-light understory (Williamson and Black 1981, Clewell 1989, Rebertus et al. 1993, Jacqmain et al. 1999). With high productivity and an abundance of shru bs in the understory, flatwoods are likely to lose herbaceous species more quickly with fire exclusion than drier longl eaf pine sites. Even slight fire suppression in a fl atwoods system leads to incr eased sprouting and vegetative reproduction among shrubs, reducing available space for herbaceous plants and lowering species richness (Glitzenstein et al. 2003) Different plant functional groups experience different rates of species loss with reduced fire fre quency. Mesic sites, with more species to lose and more rare species, are likely to experience the highest rates of species loss (Leach and Givnish 1996). In mesic to wet longleaf pine savannas, the species most likely to be re duced or lost are the dominant bunch grasses, basal rosette species, sedges, other sm all monocots, and insectivorous species (Walker and Siletti 2006). Because the understory species composition and soil properties vary markedly with hydrology in longleaf pine ecosystems, it is necessary to test restoration me thods in a variety of site conditions in order to deve lop appropriate restor ation protocols. Mesic flatwoods in the Florida panhandle are generally co-dominated by wiregrass ( A. stricta ), saw palmetto ( Serenoa repens W. Bartram), runner oak (Quercus elliottii Wilbur), gallberry ( Ilex glabra A. Gray), and hairy wicky ( Kalmia hirsuta Walter). The overstory in a mesic longleaf pine flatwoods is typically an open canopy of longleaf pine (basal area of 2.5 to 12.5m2/ha), with slash pine ( Pinus elliottii Engelm.) and pond pine ( Pinus serotina Michx.) codominant in the transitional areas near wetlands (Peet 2006). Fire is a key management tool for m eeting restoration goals in these 14

PAGE 15

systems, but social and environmental factors can often hinder the ability of land managers to use fire effectively (Chapter 1). When a restoration project includes conversion of the overstory from a planted slash pine or loblolly pine ( Pinus taeda L.) stand, complete removal of overstory pines across a large area has the counterproductiv e effect of releasing shrubs (Bro ckway et al. 1998, Kush et al. 1999, McGuire et al. 2001, Kirkman et al. 2007, Pecot et al. 2007). Th e release of shrubs following overstory harvest is likely to be more extensive on a site with a past history of fire suppression, due to the preponderance of underground hardw ood rootstocks estab lished during periods without fire (Olson and Platt 1995 ; Drewa et al. 2006). Herbicide treatment and mechanical site preparation can therefore be important components of restoration projects in which a slash or loblolly pine plantation is being converted to longleaf (Wal ker and Siletti 2006). A potential negative impact of herbicide, however, is a reduction in nontarget desirable understory vegetation. Litt et al. (2001) conducted an extensive litera ture review on the effects of herbicide on understory vegetation in southern pinelands. They examined 125 studies from which 21 were used for review. Throughout the herbicide literature they found a lack of experimental rigor and inconsistent reporting standards, which made it di fficult to reach any ge neral conclusions about the effects of herbicide on ground-layer vegetati on. They were able to conclude only that widespread use of herbicide to control unwan ted vegetation may have undesirable effects on nontarget plant species, and that additional studies of herbicid e impacts are needed before treating large, diverse la ndscapes. They recommend that future herbicide studies follow rigorous experimental designs and make a deliberate effort to track sp ecies of special concern. 15

PAGE 16

Litt et al. (2001) also concluded that the effects of herbicide on ground-layer vegetation in natural flatwoods systems had rarely been measured. They reviewed only two studies documenting herbicide impacts on groundcover in nonplantation flatwoods ecosystems, both of which reported decreases in species richness. Decreases in richness ranged from 5.1% for herbaceous species following hexazinone applic ation (Wilkins et al. 1993), to 71.8% for all species following treatment with a mix of sulfometuron methyl, glyphosate, and triclopyr (Neary et al. 1991). The effects of herbicide on sandhill communitie s are better documented. Four years after treatment, hexazinone herbicide was effectiv e in eliminating woody species and increasing diversity relative to the control, though plots restored with fire alone had even higher species diversity than those treated with herbicide (Provencher et al. 2001). On another north Florida sandhill site, an initial decrease in species diversity was documented following hexazinone application, followed by a recovery and then an increase in the second growing season (Brockway et al. 1998). A subseque nt study on the same site revealed that prescribed fire further enhanced species diversity in conjunction with hexazinone (Brockway a nd Outcalt 2000). Three herbicide release treatments (imazapyr, glyphosphat e, and hexazinone) were tested on loblolly pine plantations at three hilly sites in central Ge orgia. Seven years after treatment, there were no significant difference in understory species diversity between any of the treatments including the control (Boyd et al. 1995). Eleven years after tr eatment, Miller et al. (1999) arrived at the same conclusion. Imazapyr, used alone or in conjuncti on with fire, was effective for restoring grassdominated bobwhite quail hab itat on a loblolly pine sandhi ll in the Florida panhandle, significantly reducing hardwood encr oachment while enhancing the growth of native herbaceous species (Welch et al. 2004). Beneficial results were also reported for imazapyr in a quail habitat 16

PAGE 17

restoration project on hilltops in southern Lousiana (Jones and Chamberlain 2004). However, no differences in herbaceous species cover and richness relative to the control were found on a site in the Virginia piedmont treated with three diffe rent levels of imazapyr for pine release (Keyser and Ford 2006). Other pine releas e studies have also reported be neficial or neutral results for imazapyr, with wildlife forage vegetation showing recovery a nd in some cases improvement after treatment (Hurst 19 87, McNease and Hurst 1991). In order to gain a better unders tanding of herbicide as a rest oration tool for longleaf pine flatwoods ecosystems, we compared the effects of five herbicide treatments on the understory of a coastal flatwoods in the Florida panhandle: 1) Sulfometuron methyl (Oust, E.I. du Pont de Nemours and Company, Wilmington Del.), 2) Imazapyr (Arsenal, BASF Corporation, Research Triangle Park, NC), 3) Hexazinone (Velpar L, E.I. du Pont de Nemours and Company, Wilmington, Del.), 4) A tank mix of Oust and Velpar L, and 5) a no-herbicide control. These herbicide treatments were chosen because they are known to be effective vegetation control treatments for l oblolly and slash pine release in the Southeast, but their effects on flatwoods community composition are not well studied. A recent survey reported that the forestry herbicide most commonly used in the southeast was Arsenal, followed by tank mixes of Arsenal + Oust and Oust + Velpar (Shepard et al. 2004). We examined the longer-term effects of these herbicide treatments on flatwoods vegetation four years after in itial herbicide application (s ee Ranasinghe 2003 for first year results), after which we conducted a dormant-seas on prescribed fire on all treatment plots and assessed post-fire vegetation composition. 17

PAGE 18

Methods Study Area The study was conducted at the Point Washi ngton State Forest, which is located on the gulf coast of Walton County, Florida (30 20 N, 86 04 W). The 20-acre study site is part of a larger mesic flatwoods/wet flatwoods matrix, with an understory dominated by saw palmetto (S. repens ), gallberry ( I. glabra ), rhizomatous oaks ( Quercus minima Small and Quercus elliottii Wilbur), blueberries ( Vaccinium myrsinites Lam. and V. darrowii Camp.), huckleberries ( Gaylussacia dumosa Torr. and A. Gray and G. frondosa L. Torr and A. Gray ex. Torr), bluestem grasses ( Andropogon virginicus var. glaucus (L.) Heck and Andropogon virginicus L.) and wiregrass ( Aristida stricta ). The majority of the s ite can be classified as mesic, but due to subtle topographi cal variation across the site some areas are at the wetter end of the mesic flatwoods spectrum and some are at the drier end. For th is reason, a randomized complete block design was used, with three blocks laid out in mesic-wet areas and three blocks laid out in mesic-dry areas. A pre-treatment vegetation survey identified wetter areas by the presence of indicator species su ch as sedges (Cyperaceae). The soils of the study site are sandy, siliceous, thermic, aeric alaquods of the Leon se ries (Spodosols consisting of deep, poorly to very poorly drained soils derive d from marine parent material (Ranasinghe, 2003)). Yearly weather patterns during the study period were va riable, with hurricanes in some years and droughts in others. Total annual precipitat ion during the study period was 103cm in 2002, 200cm in 2003, 135cm in 2004, 146cm in 2005, and 106cm in 2006 (NOAA 2007). Rainfall in the months following the prescribed fire was ve ry low, with totals of 0.5cm in March 2007, 4cm in April 2007, and 0.3cm in May 2007 (NOAA 2007). Prior to August 2001, the 20-acre study site was a planted slash pine stand with a mean stand age of about 26 years, which had been burned on a 3-year rotation by the Division of 18

PAGE 19

Forestry since its acquisition of Point Washington State Forest in 1992. A pre-harvest, baseline data survey was conducted in the areas slated to become study blocks in June 2001. In August 2001, the entire overstory was harves ted and the site was prepared for the study with a single, light roller chopping followed by a prescribed burn in October 2001. In December 2001, six study blocks, each divided into five 36.6m x 24.4m treatment plots, were established according to the aforementioned randomized complete block design. Each treatment plot was hand-planted with 100 containerized longleaf pine seedlings laid out in 10 rows, with 3.1m between rows and 1.8m between the seedlings in each row. Treatments Within each block, each 100-seedling plot was randomly selected to receive one of five herbicide treatments in March of 2002: 1) Sulfometuron methyl (Oust) at 0.26 kg a.i./ha. This is a moderate application rate for sulfometuron methyl; other studies have us ed 0.16 kg a.i./ha (Lauer and Glover, 1998) and 0.21 kg ai/ha (Shiver and Mart in 2002, Keyser and Ford 2006), while the manufacturerrecommended rate for herbaceous weed control in longleaf pine stands is 0.10 to 0.42 kg a.i./ha (DuPont 2007). Sulfometuron methyl is a selectiv e herbicide primarily effective for controlling herbaceous species (Lauer and Glover, 1998), an d mixing with hexazinone is recommended if broader spectrum control is desired (DuPont, 2007). 2) Hexazinone (Velpar L) at 0.56 kg a.i./ha. Recommended application rates for hexazinone range from 2.0 to 6.7 ai kg/ha (Du Pont 2007), while rates successfully used to control hardwoods in sandhills ecosystems have ranged from 1.1 kg a.i./ha to 2.4 kg a.i/ha (Brockway et al. 1998, Provencher et al. 2001). The hexazinone application rate in our study was therefore very low. This rate was chosen because hexazinone had been shown to be 19

PAGE 20

effective at the aforementioned rates, and we were interested in whether an even lower application rate woul d be effective. 3) Sulfometuron methyl (0.26 kg a.i./ha) + Hexazinone (0.56 kg a.i./ha) mix. This is a common tank mix used throughout the Southeast (S hepard et al. 2004) with the potential to control a broad spectrum of woody and herbaceous pl ants. Tank mixes are used to achieve more complete vegetation control than either herbicide alone, but as a result they may also have more negative implications for native plant commun ities (Keyser and Ford 2006). Banded herbicide application, the method used in this study, may he lp ameliorate these ne gative impacts (Keyser and Ford 2006). 4) Imazapyr (Arsenal) at 0.21 kg a.i./ha. For a restoration project, this was a moderate application rate for imazapyr. Signifi cant hardwood control and habitat improvements have been obtained using imazapyr at 0.42 kg a. i./ha (Jones and Chamberlain 2004), and at all rates from 0.08 kg a.i./ha to 0.24 kg a.i..ha (t hough higher rates did not offer significant improvements over lower rates) (Keyser and Ford 2006). 5) Control. (No herbicide). The herbicides were applied in a 1.2m ba nd over the seedling rows using a backpack sprayer. The reasons for using a banded post-planting herbicide application on unshielded seedlings were both operational and environmenta l. Prior to harvest, the study site already contained many desirable understory species (i ncluding wiregrass) a nd few weedy species (Ranasinghe 2003). The banded application allowed us to pinpoint the delivery of herbicide to an area directly around planted seedlings, while na tive vegetation remained intact in the strips between rows. This application method also lowered the cost of herbicide treatment relative to a broadcast application. Within each 10-row treatment plot, three rows received a second 20

PAGE 21

herbicide application in March of 2003, three rows received a se cond application in April of 2003, and four rows were left with only the firstyear treatment. The results from these subplots were pooled with the first year data in the final analysis because of a lack of significant differences between one and two-year appl ications in any of the parameters. Data Collection and Analysis The pre-treatment vegetation survey was conducted in June 2001, prior to the overstory harvest. Post-treatment vegetation surveys we re conducted at three and nine months after treatment, and the results of these first-year surveys have been published elsewhere (Ranasinghe 2003, Ranasinghe et al. 2005). The fourth-year data were collect ed in June of 2006. Within each treatment plot, six 1m2 quadrats were randomly selected for understory sampling (two in each subplot). In all, 36 randomly selected quadr ats were surveyed in each herbicide treatment and 12 were surveyed in the cont rol, for a total of 156 quadrats. The smaller sample size in the control treatment was due to the fact that a port ion of each control plot was sprayed in the second year, and these rows were not included for anal ysis. All species were identified and percent cover of each species was estimat ed visually using the cover cl asses defined by the modified Daubenmire scale (Daubenmire 1959, Peet et al 1998): 0-1%, 1-2%, 2-5%, 5-10%, 10-25%, 2550%, 50-75%, 75-95%, and 95-100%. For each woody species, average stem height was also estimated and all stems were counted. I. glabra, S. repens and Quercus spp. ( Q. minima and Q. elliottii ), were uniformly distributed across all study plots prior to treatment (Ranasinghe 2003) and were grouped together for analysis. Woody species belonging to the family Ericaceae (heaths) were analyzed separately from the other dominant shrubs because they were not evenly distributed across plots prior to treatment, and also because members of this family are tolerant to hexazinone and may react differently to trea tment than other shrub groups (Wilkins et al 1993, 21

PAGE 22

Outcalt et al 1999, Miller et al 1999). The same procedure was repeated in June of 2007, following a prescribed fire on February 16, 2007. Pyrometers were constructed using aluminum forestry tags and Tempilaq heat-sensitive paints, following a method employed in other fire ecology studies (Hobbs et al. 1984, Iverson et al. 2004). Thirteen Tempilaq paints (designed to melt at 93 C, 149C, 204C, 260C 316C 371C, 427C, 482C, 538C, 593C, 649C, 704C, 760C, and 871C) were applied to each tag. A second tag was paper-clipped over the fron t of the first to protect the paints from charring. Six stakes were randomly placed in each herbicide treatment plot (two in each subplot) and two in each control, with two pyrometers attached to each stake at 30cm and 80cm, for a total of 36 pyrometers at each height for each herbic ide treatment and 12 at each height for each control. Each pyrometer gave a measure of the ma ximum fire temperature at that location: all of the paint dots below that temperature were melted, while all of the paint dots above it remained unchanged. When analyzing species cover vs. temperature relationships, the two pyrometer readings at the 30cm level in e ach subplot were averaged and pl otted against the average of the two understory quadrats in that subplot. Since the pyrometers were placed randomly, they were not necessarily in the same locations as the qua drats. Averaging them gave a more accurate overall picture of the cover and fire intensity in a gi ven plot. Pyrometers on which none of the paints were burned (i.e. temperat ures did not reach 93C) were assumed to be unburned and were assigned a value of 10C to represent ambient air temperature. All parameters were analyzed in JMP IN version 5 (SAS Institute, Inc.), using ANOVA within the framework of a randomized complete block design. The study addressed only the main effects of herbicide treatment, and test s of these effects were not dependent on the assumption of no treatment x block interaction. Block effects were therefore treated as random 22

PAGE 23

effects in a univariate ANOVA model with tw o independent variable s: treatment with Block&Random as a covariate. Data were log-transformed where necessary to meet the assumptions of ANOVA. Significant differences between treatments were separated with Tukeys HSD or Hsus MCB. Post-fire treatme nt differences were analyzed with ANCOVA, using pre-fire distributions as a covariate. ANCOVA was also used to analyze fourth year abundance of Ericaceae, using pre-treatment distribu tions as a covariate. A significance level of = 0.05 was used to test all parameters. Community patterns were analyzed in PC-ORD version 4, using Nonmetric MultiDimensional Scaling (NMDS) a nd Indicator Species Analysis following the method of Dufrene and LeGendre (1997). NMDS uses rank-transformed distances to linearize the degree of difference between plant survey quadrats. Ordi nation of plant survey quadrats (containing both species presence and percent cover data) was don e using a Srenson distance measure in PCORDs autopilot mode. The Srensen coefficient is a measure of percent dissimilarity that can be applied to either presence-absence data or quantitative data (in this case percent cover values), and is recommended for analysis of ecological communities (McCune and Grace 2002). In the autopilot mode, NMDS conducts 40 runs with real data and 50 runs with randomized data. Species with less than 3 occurrences were dropped from the analysis to reduce the effects of rare species (McCune and Grace 2002), reducing the total number of species in the analysis from 87 to 55. PC-ORD chooses the solution with the highes t dimensionality that also meets the criterion of having less stress than 95% of the random runs. Each point on the NMDS axes represents a single plant survey quadrat, and the distance be tween points on the NMDS axes represents the Srenson distance between points in n dimensions (in this case a 3-dimensional solution was chosen). A Multi-Response Permutation Procedur e (MRPP) can then be used to determine the 23

PAGE 24

degree of agreement within groups with regard to an environmental variable, which in this case was a categorical variable (treatment). The weighting method for the MRPP was n/sum(n), and we again used a Srenson distance measure. MRPP yields a test statistic (A) and a p-value (A = 1 when all items are identical within groups (the groups in this case were herbicide treatments), and A=0 when heterogeneity within gr oups equals expectation by chance). A therefore expresses the degree to which variation in position between plant survey quadrats on the ordination axes is explained by the herbicide treatments. Indicat or species analysis using the Indicator Value method defines an indicator species as a specie s found mostly in a sing le group and present in the majority of sites belonging to that group (DuFrene and LeGendre 1997). The indicator value (IV) is maximum (100%) when the individuals of species I are observed in a ll sites of only one group (in this case, the each group was an herb icide treatment). IV is obtained by combining relative abundance and relative frequency va lues for each species in each group. Relative abundance of a species in a group is the averag e abundance of a given species in a given group of plots over the average abundance of that spec ies in all plots (expres sed as a percent), and relative frequency is the average number of plots in a given group where a given species is present. Significance of each Indicator Valu e was established using a 1000-permutation MRPP which compared the observed maximum IV for each species to the IV obtained from randomized groups, and generated a p-value for this comparison using the Monte Carlo test of significance. Species identified as indicators fo r one of the herbicide treatmen ts are assumed to have been enhanced by that treatment relati ve to the others, whereas species identified as indicators for the control are assumed to have been negatively impacted by the herbicide treatments. 24

PAGE 25

Results Shrub Cover Four years after initial herbic ide application, pe rcent cover of I. glabra, S. repens and Quercus spp. ranged from 13.5% in the sulfometuron + hexazinone (sulfo + hexa) treatment to 22.9% in the sulfometuron treatment, but did not vary significantly among treatments overall (p = 0.2154) (Table 2-1). Indicator species analysis revealed a significant association between the sulfometuron treatment and increased c over of both rhizomatous oak species, Q. elliottii and Q. minima four years after treatment (p = 0.040 and p = 0.031) (Table 2-2). Since these species were not significantly associated with the control pl ots, this result can be in terpreted to mean that the sulfometuron treatment enhanced oak growth. Woody species belonging to the family Er icaceae were abundant on the study site, and included blueberries ( V. myrsinites and V. darrowii ), huckleberries (G. dumosa and G. frondosa ), hairy wicky ( K. hirsuta ), and Lyonia lucida Koch (fetterbush). Ericac eae cover four years after treatment was four to five times higher in th e control plots than in any of the herbicide treatments, a difference which approached si gnificance (p = 0.06), but was moderated by the uneven pre-treatment distributions and the smaller control sample size (Table 2-1, Figure 2-1). Overall percent cover of woody speci es was highest in the control a nd lowest in the sulfo + hexa treatment (Figure 2-2). In June 2007, four m onths after the February 2007 prescribed fire, percent cover of Quercus spp., I. glabra, and S. repens did not vary significantly among treatments (p = 0.1740) (Figure 2-3a, Tabl e 1), nor did Ericaceae cover (p = 0.097). Herbaceous Cover Herbaceous cover varied signi ficantly among treatments (p = 0.002) (Figure 2-4a). In both the imazapyr and hexazinone treatments, mean herbaceous cover (46.5% and 46.1%, respectively) was significantly higher than th e control (22.1%). Following fire, however, 25

PAGE 26

herbaceous cover no longer differed significan tly among treatments (p = 0.302). Herbaceous cover decreased significantly following fire in all but the control and sulfometuron plots (Figure 2-4b). Much of this decrease can be attributed to the decrease in wiregrass crown size following fire (see below). Overall vegeta tion cover before and after fire is summarized in Figure 2-5. Wiregrass Cover Mean wiregrass cover four y ears after initial herbicide application differed significantly among treatments (p = 0.04) (Figure 2-6a). Wiregr ass cover in the sulfo + hexa treatment was three times greater than the cont rol (Figure 2-6). In all treatme nts except the control, wiregrass cover 4 years after treatment was higher than pre-tr eatment levels (Figure 2-6c). Following fire, wiregrass cover decreased in all treatments (F igure 2-6b). The overall decrease in wiregrass cover following fire is not attributed to wiregr ass mortality, but to the decrease in wiregrass crown size following fire. Wiregrass cover was su rveyed only four mont hs after the February 2007 fire, and it was observed that individual wiregrass clumps were alive and regenerating, but greatly reduced in size from pre-fire levels. A similar effect was documented after the herbicide treatment in 2002, with an init ial decrease in wiregrass cove r followed by a rebound over the next nine months (Figure 2-6c). Shrub Stem Density Four years after treatment, mean shrub stem density for Quercus Ilex and Serenoa varied significantly among treatments (p = 0.012), displaying a similar pa ttern to percent cover (Table 2-1). Mean stem density in the imazapyr treatment (20 stems/m2) was significantly lower than sulfometuron (34.7 stems/m2) and hexazinone (32.9 stems/m2). None of the treatments varied significantly from the control, in whic h we measured a mean stem density of 31.75 stems/m2. In the first nine months after herbicid e application, imazapyr was the only treatment that did not experience an increase in shrub stem density (Figure 2-7b). In the fourth year after 26

PAGE 27

application, this effect was no longer evident, as shrub stem density no longer varied significantly among treatments. Following the pr escribed fire in February 2007, mean shrub stem density increased again by around 50% in ev ery treatment (Figure 2-7a). These increases were significant in all treatment s but the control. Variation in Ericaceae stem density among treatments approached significance in June 2006 (p = .061) (Table 2-1), but this difference was moderated by the smaller control sample size an d uneven pre-treatment distributions. Ericaceae stem density did not change significantly in a ny of the treaments following fire (Table 2-1). Shrub height Mean shrub stem height for Quercus Ilex and Serenoa did not differ significantly among treatments in June 2006 (p = 0.2341). Stem height ranged from 14.6cm in the imazapyr treatment to 17.1cm in the hexazinone treatment (T able 2-1). Following the prescribed fire in February 2007, mean shrub height increased in the imazapyr, sulfo, and sulfo + hexa treatments (Table 2-1). Differences among treatments following fire were not significant (p = 0.40). Mean stem height of Ericaceae did not differ signif icantly among treatments in June 2006 (p = 0.304) or in June 2007 (p = 0.900), though height increases within some treatments following fire were significant (Table 2-1). Understory responses A total of 93 species were identified dur ing the June 2006 and June 2007 plant surveys (Appendix A). Weedy species were not prevalent, and we found numerous plant specimens belonging to functional groups identified by Walk er and Peet (1983) and Glitzenstein et al. (2003) as high-risk for being lost in shrub-invaded flatwoods, such as bunch grasses (Poaceae), sedges (Cyperaceae), sundews (Droser aceae), and basal rosette species. Overall species richness four years after he rbicide application varied significantly among treatments (p = 0.01), ranging from 7.4 species/m2 in the control treatment to 9.3 species/m2 in 27

PAGE 28

the sulfometuron treatment (Table 2-3). Herbace ous species richness, however, did not vary significantly among treatments (p = 0.498) (Table 23). Following fire, overall species richness no longer varied significantly among treatments (p = 0.15). Within treatments, responses to the fire varied (Table 2-3). Over all species richness increased signi ficantly in the control, but did not change significantly in the other treatments. Herbaceous species richness did not vary among treatments before or after fire (p = 0.498 and p = 0.114, respectively). Species Diversity Overall Shannon diversity (woody and herbaceous species combined) four years after herbicide application did not vary significantl y among treatments (p = 0.07) (Figure 2-8a, Table 2-3). However, herbaceous species diversity did vary signifi cantly among treatments (p = 0.036), with the highest diversity in the sulfometuron treatment a nd the lowest in the sulfo + hexa treatment (Table 2-3). Within treatments, re sponses to the prescribed fire varied (Figure 28b). Overall diversity increased following fire in the sulfo + hexa and hexazinone treatments, but did not change significantly in any of the other treatments (Figure 2-8b). Herbaceous diversity increased in the imazapyr, sulfo + hexa and hexazinone treatments following fire (Table 2-3). Initial gains in overall Shannon diversity during the first year after treatment (Ranasinghe 2003) were no longer evident after four years (Figure 2-8c). Species Evenness Mean species evenness (1/D) f our years after herbicide appl ication, calculated for woody and herbaceous species combined, did not vary significantly among treatments overall (p = 0.12) (Table 2-3). Herbaceous species evenness, how ever, did vary significantly among treatments (p = 0.014), with the highest evenness in the sulfom eturon treament and the lo west in the sulfo + hexa treatment (Table 2-3). Following fire, only the sulfo + hexa tr eatment experienced a significant change in overall evenness (from 3.12 to 4.54) (Table 2-3) Herbaceous species 28

PAGE 29

evenness increased in the imazapyr, sulfo + hexa and hexazinone treatments following fire (Table 2-3). Community Ordination Analysis of the plant community four years after treatment using NMDS and MRPP revealed only a slight relationship between treat ment and species composition (Fig 2-8), with a small chance-corrected within-gro up agreement (A) of 0.02 (p = 0.01) for the pre-fire survey quadrats. Indicator Species Analysis following the method of Dufrene and LeGendre (1997) revealed significant relationships between some species and treatments (Monte Carlo test of significance) (Table 2-2). Species of Special Interest Curtiss sandgrass ( Calamovilfa curtissii Scribn.), a florida panhandle endemic, was identified in one of the imazapyr treatment pl ots during both the June 2006 and June 2007 plant surveys. Prior to treatment, there was a very low density of legumes on our study site: out of 90 survey quadrats, only two legume specimens were identified, both of which were Mimosa quadrivalvis var. angustata Barneby (Ranasinghe, unpublished data). After four years, we found Desmodium lineatum DC, Desmodium strictum DC, and Tephrosia hispidula Pers. in addition to M. quadrivalvis, again at a low density: a total of eight legume specimens were identified out of 156 survey quadrats. In the post-fire survey, one D. lineatum four M. quadrivalvis, and one Baptisia lanceolata Elliott specimen were identified. These species were not significantly associated with any of the treatments. Fire Temperature The prescribed fire was cool, fast-moving, and patchy. Mean maximum fire temperature by treatment ranged from 252C to 306C at 30cm and 166C to 257C at 80cm (Table 2-4), and did not differ significantly among herbicide treatm ents at either height (p = 0.42 and p = 0.55, 29

PAGE 30

respectively). A significant negative relationship existed between percent Quercus cover and maximum fire temperature at 30cm (r2 = 0.06, p = 0.032) (Figure 2-9a), whereas a significant positive relationship existed between percent I. glabra cover and maximum fire temperature at 30cm (r2 = 0.07, p = 0.017) (Figure 2-9b). The ne gative relationship observed with Quercus is largely attributable to the fact that areas with high Quercus cover were often only partially burned or did not ignite at all, whereas the I. glabra tended to ignite readily and burn completely. Temperature was not significantly related to wiregrass cover (p = 0.73) or overall herbaceous cover (p = 0.28). Discussion Four years after initial herbic ide application, differential treatment effects on understory composition were still evident. Sulfometuron me thyl had a releasing e ffect on the two most common oak species, Q. minima and Q. elliottii indicating that this is not an appropriate herbicide for flatwoods with a high density of oa ks. However, in spite of having higher shrub cover and lower herbaceous cover than the othe r herbicide treatments, sulfometuron plots also had the highest fourth year herbaceous species ri chness, diversity, and evenness. One reason for this may have been that wiregrass cover in th e sulfometuron plots was so mewhat lower than the other treatments, leaving more space for other small herbaceous sp ecies. These gains in species richness, diversity, and evenness are therefore un likely to persist through future fire cycles because it will be more difficult to apply prescribed fire to this treatment. Indeed, sulfometuron was the only treatment in which species richne ss, diversity, and eve nness did not increase following the prescribed fire in 2007. Woody shrub cover, height, and stem density we re lowest in the imazapyr treatment, but the differences between imazapyr and the other treatments were not as pronounced as those observed in the first year after a pplication. These results suggest that the increases in shrub stem 30

PAGE 31

density evident in the first year (Figure 2-7b) were a short-term phenomenon, and subsequent dieback of sprouts occurred as these shrubs matured and grew taller. The responses of Ericaceae to our treatments were surprising, given that in other studies members of this family have been highly hexa zinone-resistant (Wilkins et al. 1993, Outcalt et al. 1999, Miller et al. 1999), and hexazinone is comm only used as a release treatment for blueberry crops. If anything, we expected to see higher Ericaceae cover in the hexazinone treatment, when in fact this treatment had the lowest Ericaceae cove r of all. This result was particularly notable given that the hexazinone plots had the highest percent cover of Ericaceous shrubs prior to treatment (Figure 2-1). The sensitivity of Eric aceous shrubs to hexazinone may have been a function of the timing of application and the poor ly drained, sandy soil; Velpar L is labeled for blueberry release only prior to budbreak in the spring and in conditions where there is no standing water (DuPont, 2007). Control of thes e species may be desirable for restoration scenarios in which the goal is to increase he rbaceous cover on a site dominated by Ericaceous shrubs (Outcalt et al. 1999). However, many of these shrubs are important wildlife food plants (Hay-Smith and Tanner 1994). If the goal is to avoid negative impacts to these populations, our results suggest that cauti on may be in order with regard to soil type, soil water, and timing of application. Both imazapyr and hexazinone significantly increased overall herbaceous cover relative to the control, and the sulfo + hexa treatment significantly increased wire grass cover relative to the control. Though the shrub control initially provided by the herbicide treatments was relatively short-live d, these results suggest th at the short-term shrub control allowed herbaceous species (and wiregrass in particular) to gain more of a foothold in the understory. 31

PAGE 32

The mean fire temperatures we recorded were low for a dormant-season fire in a mesic flatwoods: mean maximum fire temperatures for several other dormant season fires on flatwoods sites ranged from 383C to 588C (Dre wa et al. 2002), whereas our mean maximum temperatures were 200C at 80cm and 274C at 30cm. Across the study site, we made the general observation that I. glabra burned more readily and thoroughly than Quercus spp., which often burned only partially, and, in some highly oakdominated patches, failed to ignite at all. Since dead pine needles are a critical component of pyrogenic ity in longleaf pine systems (Williamson and Black 1981, Clewell 1989), it is like ly that the complete removal of overstory pines contributed to our inability to apply fire effectively to the site. In our study, stem counts of both Quercus spp. and I. glabra more than doubled in all treatments in the four months following fire. Because we did not have an unburned control, we cannot say to what degree the observed post-fire increases in shrub c over and density were greater than the increases expected from normal sp ring growth of these shrubs in the absence of overstory competition. However, it is safe to sa y that the prescribed fire was ineffective at controlling the shrubs, because it did not decrease shrub cover or even maintain it at pre-fire levels. This result can be attri buted to aggressive shrub regrowth due to the lack of competition from overstory pines, the loss of needlefall as a fine fuel supply, and the length of time since the previous fire and subsequent herb icide applications (four and five years). Shrub control, and by extension floral diversity, are maximized by fire re turn intervals of one to three years (Kirkman et al. 2001; Glitzenstein et al. 2003 ). If future fires are conduct ed on a shorter interval, better prescribed fire results may be obtained. The season of burn may also have contribu ted to the difficulty we had in applying prescribed fire effectively. It is generally be lieved that summer fires were probably the most 32

PAGE 33

prevalent selective force prior to European colonization of the S outheast, and are more effective at controlling shrubs and maintaining herbaceous plant diversity in the understory (Glitzenstein et al 1995, Streng et al. 1993). Others have reported shrub resprouti ng responses to dormantseason fire in flatwoods similar to those we obs erved. On mesic and wet sites in Florida and Lousiana, shrub stem densities we re greater than pre-treatment levels after dormant-season fire, and continued to increase with repeated dormant season fires, whereas long-term growing season fires maintained stem densities at the same le vel (Drewa et al. 2002). In another recent study, it was observed that stem densities of root-crown bearing shrubs ( Quercus and Ilex spp.) after repeated dormant-season fires were seven times gr eater than those subjec ted to growing-season fires (Drewa et al. 2006). In one case, stem densities of root-cro wn bearing shrubs were 50 times greater in South Carolina savannas managed w ith annual dormant season fires than in plots managed with annual growing-season fires over a 30 year period (Waldrop et al 1992). Some of the resprouting ability of Quercus can be attributed to a past hi story of fire suppression, during which underground stems are likely to have grown extensively (Drewa et al. 2006). Given that our site was previously a slash pine plantation, periods of fire suppression are likely to have occurred in the past during seed ling establishment and may have c ontributed to the recalcitrance of the dominant woody shrubs. However, results indicating that dormant-se ason fire increases shr ub cover in flatwoods are not universal. In one case, a long-term dorma nt season prescribed fire regime significantly reduced shrub cover on an I. glabra -dominated flatwoods while increasing herbaceous species cover and diversity (Brockway and Lewis 1997), sugge sting that dormant-season fire is a viable option for flatwoods management when growing-s eason fire cannot be use d. Hiers et al. (2000) offer experimental evidence, based on patterns of legume reproduction, that variable fire regimes 33

PAGE 34

will maintain a broader suite of native species than either growing-season or dormant-season fire alone, and cite studies suggesti ng that significant variation in the frequency and season of lightning may have occurred during the 7000 years BP that longleaf pine communities have been established in the sout hern coastal plain. Though dormant season fire may be effectiv e and even desirable in some cases, our results suggest that dormant season fire should be avoided when species of low flammability are abundant, there is a lack of fine fuels, and there are no overstory pines to compete with shrubs for light and soil resources. Ilex glabra is known to be one of the most flammable flatwoods species, while oaks are among the least flammable (Behm et al. 2004), and this may be another part of the reason for the succe ss of dormant-season fire on an I. glabra-dominated site reported by Brockway and Lewis (1997). In our study, though stem density of both Quercus and Ilex increased following fire, only Quercus increased in mean percent cover across treatments; mean I. glabra cover remained roughly the same or decrease d. It may be that in the coming growing seasons, some of the in itial flush of basal I. glabra sprouts will die back, and this species will not show an overall cove r increase. Since Quercus increased in cover as well as stem density, it seems more likely that the fire-related increase will have long-term effects. In summary, only positive herbicide impact s on wiregrass and herbaceous cover were observed in this study, and understory species diversity, richness, evenness, and community composition were largely unaffected. The imazapyr treatment, which had the most dramatic shrub control effects in the first year after application, still had the lowest shrub levels and the highest herbaceous cover after four years, indica ting that this was the mo st successful treatment overall. The poor vegetation control offered by sulfometuron, which was largely attributable to the resistance of Q. elliottii and Q. minima indicates that it is not an appropriate herbicide for 34

PAGE 35

use in an oak-dominated flatwoods community. However, the sulfometuron + hexazinone tank mix showed promise as a flatwoods restor ation treatment, offeri ng better control over Quercus spp. and significantly increasing wiregrass cove r relative to the control. Though hexazinone plots had significantly higher overall herbaceous cove r than the control, our results suggest that this herbicide, at the rate app lied, did not provide much control over the dominant shrubs in the long term. Because we applied a ve ry low rate of hexazinone, it is difficult to make any absolute comparisons between hexazinone and the other he rbicides based on the re sults of this study. However, given the success others have had w ith hexazinone (Brockway et al 1998, Brockway and Outcalt 2000, Provencher et al 2001), it seems lik ely that we would have seen greater shrub control and attendant increases in herb aceous cover had we used a higher rate. As of June 2007, it was too early to tell wh ether the herbicide-related increases in wiregrass and overall herbaceous cover will be maintained through future fire cycles. The Florida Division of Forestry normally burns this site on a three-year fire interval, and results from more frequent burning may be more successful than those we observed five years after the inital herbicide application a nd four years after the second a pplication. Because the entire overstory was harvested, it may be several fire cycles before the new longleaf pine canopy provides sufficient needlefall to sustain the level of prescribed fire necessary for shrub control. In the meantime, this limitation is likely to be exacerbated by the Florida Division of Forestrys inability to use growing-season fire on this site due to its proximity to a highway and several beach housing communities. If the trends of increasing shrub cover, height, and density continue, the herbicide-related gains in wiregra ss and overall herbacous cover may be lost. The aggressive post-fire resprouting of shrubs observed in this study underscores the utility of herbicide as a supplemental rest oration tool, especially in fl atwoods ecosystems where undesired 35

PAGE 36

changes in understory community composition can be rapid and diffi cult to reverse. It also underscores the need for effective prescribed fire in order to maintain the early gains in wiregrass and herbaceous cover conferre d by herbicide application. 36

PAGE 37

Table 2-1. Mean percent cover, st em density, and height of shrubs 4 years after treatment (YAT) (June 2006) and following prescribed fire (June 2007). Control Imazapyr Sulfometuron Sulfo + Hexa Hexazinone Control Imazapyr Sulfometuron Sulfo + Hexa Hexazinone Treatment ANOVA/ ANCOVA p = 0.215 18.7 13.6 22.9 13.5 19.9 p = 0.215 18.7 13.6 22.9 13.5 19.9 18.7 13.6 22.9 13.5 19.9 22.5 24.3* 34.0* 21.5 26.6 22.5 24.3* 34.0* 21.5 26.6 21.3 5.1 4.4 5.8 4.0 p = 0.060 21.3 5.1 4.4 5.8 4.0 21.3 5.1 4.4 5.8 4.0 p = 0.060 2.8 2.3 5.6 3.6 3.3 2.8 2.3 5.6 3.6 3.3 p = 0.174 31.8ab20.8b34.7a25.1ab32.9a31.8ab20.8b34.7a25.1ab32.9ap = 0.012 56.3 47.8* 64.9* 50.9* 59.1* 56.3 47.8* 64.9* 50.9* 59.1* p = 0.331 p = 0.097 % Cover Q, S, I 2006 4YAT PostFire Ericaceae 2006 4YAT PostFire Stem Density (stems/m2) Q, S, I 2006 4YAT PostFire Ericaceae 2006 4YAT PostFire Stem Height (cm) Q, S, I 2006 4YAT PostFire Ericaceae 2006 4YAT PostFire % Cover Q, S, I 2006 4YAT PostFire Ericaceae 2006 4YAT PostFire % Cover Q, S, I 2006 4YAT PostFire Q, S, I 2006 4YAT PostFire Ericaceae 2006 4YAT PostFire Ericaceae 2006 4YAT PostFire Stem Density (stems/m2) Q, S, I 2006 4YAT PostFire Ericaceae 2006 4YAT PostFire Stem Density (stems/m2) Q, S, I 2006 4YAT PostFire Q, S, I 2006 4YAT PostFire Ericaceae 2006 4YAT PostFire Ericaceae 2006 4YAT PostFire Stem Height (cm) Q, S, I 2006 4YAT PostFire Ericaceae 2006 4YAT PostFire Stem Height (cm) Q, S, I 2006 4YAT PostFire Q, S, I 2006 4YAT PostFire Ericaceae 2006 4YAT PostFire Ericaceae 2006 4YAT PostFire 16.1 14.6 16.9 15.9 17.1 p = 0.234 16.1 14.6 16.9 15.9 17.1 16.1 14.6 16.9 15.9 17.1 p = 0.234 21.1 19.7* 22.3* 21.6* 19.7 21.1 19.7* 22.3* 21.6* 19.7 21.1 19.7* 22.3* 21.6* 19.7 42.9 8.4 17.2 16.5 13.0 p = 0.061 42.9 8.4 17.2 16.5 13.0 42.9 8.4 17.2 16.5 13.0 p = 0.061 10.3 8.9 25.0 12.8 9.6 p = 0.108 10.3 8.9 25.0 12.8 9.6 10.3 8.9 25.0 12.8 9.6 p = 0.108 p = 0.400 11.6 12.3 12.2 12.0 11.9 p = 0.304 11.6 12.3 12.2 12.0 11.9 11.6 12.3 12.2 12.0 11.9 p = 0.304 12.5 12.1 15.2* 17.0* 16.1* p = 0.900 12.5 12.1 15.2* 17.0* 16.1* 12.5 12.1 15.2* 17.0* 16.1* p = 0.900 *Letters denote significant differences between means in the same column at = 0.05 based on Tukey-Kramer HSD test. (*) denot es significant post-fire change within treatment (Tukey-Kramer HSD). P-values are for randomized block ANOVA / ANCOVA within columns (significant at = 0.05). Q, S, I stands for Quercus spp., S. repens and I. glabra 37

PAGE 38

Table 2-2. Indicator species analysis following the method of Dufrene and LeGendre (1997). IV from randomized groups Asclepias cineria Pteridium aquilinum Lachnocaulon auceps Lachnanthes caroliniana Quercus elliottii Species Quercus minima Smilax pumila Treatment Sulfo + Hexa Sulfometuron Imazapyr Hexazinone Sulfometuron Sulfometuron Control Observed Indicator Value (IV) 15.5 20.3 12.2 16.7 19.1 29.7 12.5 Mean 5.4 11.7 5 4.1 12.6 21.7 3.7 S. Dev 2.91 3.4 2.86 2.5 3.27 3.56 2.39 p* 0.010 0.020 0.025 0.010 0.040 0.031 0.029 IV from randomized groups IV from randomized groups Asclepias cineria Pteridium aquilinum Lachnocaulon auceps Lachnanthes caroliniana Quercus elliottii Species Quercus minima Smilax pumila Treatment Sulfo + Hexa Sulfometuron Imazapyr Hexazinone Sulfometuron Sulfometuron Control Observed Indicator Value (IV) 15.5 20.3 12.2 16.7 19.1 29.7 12.5 Mean 5.4 11.7 5 4.1 12.6 21.7 3.7 S. Dev 2.91 3.4 2.86 2.5 3.27 3.56 2.39 p* 0.010 0.020 0.025 0.010 0.040 0.031 0.029 Asclepias cineria Pteridium aquilinum Lachnocaulon auceps Lachnanthes caroliniana Quercus elliottii Species Quercus minima Smilax pumila Asclepias cineria Pteridium aquilinum Lachnocaulon auceps Lachnanthes caroliniana Quercus elliottii Species Quercus minima Smilax pumila Treatment Sulfo + Hexa Sulfometuron Imazapyr Hexazinone Sulfometuron Sulfometuron Control Treatment Sulfo + Hexa Sulfometuron Imazapyr Hexazinone Sulfometuron Sulfometuron Control Sulfo + Hexa Sulfometuron Imazapyr Hexazinone Sulfometuron Sulfometuron Sulfo + Hexa Sulfometuron Imazapyr Hexazinone Sulfometuron Sulfometuron Control Observed Indicator Value (IV) 15.5 20.3 12.2 16.7 19.1 29.7 12.5 Observed Indicator Value (IV) 15.5 20.3 12.2 16.7 19.1 29.7 12.5 15.5 20.3 12.2 16.7 19.1 29.7 15.5 20.3 12.2 16.7 19.1 29.7 12.5 Mean 5.4 11.7 5 4.1 12.6 21.7 3.7 Mean 5.4 11.7 5 4.1 12.6 21.7 3.7 5.4 11.7 5 4.1 12.6 21.7 5.4 11.7 5 4.1 12.6 21.7 3.7 S. Dev 2.91 3.4 2.86 2.5 3.27 3.56 2.39 S. Dev 2.91 3.4 2.86 2.5 3.27 3.56 2.39 2.91 3.4 2.86 2.5 3.27 3.56 2.91 3.4 2.86 2.5 3.27 3.56 2.39 p* 0.010 0.020 0.025 0.010 0.040 0.031 0.029 p* 0.010 0.020 0.025 0.010 0.040 0.031 0.029 0.010 0.020 0.025 0.010 0.040 0.031 0.010 0.020 0.025 0.010 0.040 0.031 0.029 Monte Carlo test of significance, =0.05 38

PAGE 39

Table 2-3. Species Richness (species/m2), Diversity (Shannon Index), and Evenness (Simpsons Index, 1/D) 4 years after treatment (YAT) (June 2006) and following fire (June 2007). Control Imazapyr Sulfometuron Sulfo + Hexa Hexazinone Control Imazapyr Sulfometuron Sulfo + Hexa Hexazinone Treatment ANOVA/ ANCOVA p = 0.014 Species Richness (spp/m2) Overall 2006 4YAT PostFire Herbaceous 2006 4YAT PostFire Species Richness (spp/m2) Overall 2006 4YAT PostFire Overall 2006 4YAT PostFire Herbaceous 2006 4YAT PostFire Herbaceous 2006 4YAT PostFire p = 0.498 p = 0.151 PostFire Shannon Diversity Overall 2006 4YAT PostFire Herbaceous 2006 4YAT PostFire Shannon Diversity Overall 2006 4YAT PostFire Herbaceous 2006 4YAT Shannon Diversity Overall 2006 4YAT PostFire Overall 2006 4YAT PostFire Herbaceous 2006 4YAT Herbaceous 2006 4YAT p = 0.066 p = 0.131 p = 0.114 Evenness (1/D) Overall 2006 4YAT PostFire Herbaceous 2006 4YAT PostFire Evenness (1/D) Overall 2006 4YAT PostFire Overall 2006 4YAT PostFire Herbaceous 2006 4YAT PostFire Herbaceous 2006 4YAT PostFire p = 0.116 1.34 1.48 1.57 1.34 1.35 1.34 1.48 1.57 1.34 1.35 1.10ab1.01ab1.17a0.85b0.88b1.10ab1.01ab1.17a0.85b0.88b7.36c8.97ab9.25a8.03bc8.22bc7.36c8.97ab9.25a8.03bc8.22bc4.58 5.75 5.41 5.13 5.33 4.58 5.75 5.41 5.13 5.33 3.54ab3.51ab4.15a3.12b3.01b3.54ab3.51ab4.15a3.12b3.01b3.02ab2.45ab3.03a2.11b2.13b3.02ab2.45ab3.03a2.11b2.13b9.75* 9.0 8.19 9.0 8.86 9.75* 9.0 8.19 9.0 8.86 6.83 5.97 5.05 6.14 5.94 6.83 5.97 5.05 6.14 5.94 1.67 1.70 1.51 1.69* 1.65* 1.67 1.70 1.51 1.69* 1.65* 1.59 1.40* 1.30 1.40* 1.45* 1.59 1.40* 1.30 1.40* 1.45* 4.24 4.43 3.80 4.54* 4.07 4.24 4.43 3.80 4.54* 4.07 p = 0.249 p = 0.036 p = 0.293 p = 0.014 p = 0.533 4.36 3.67* 3.53 3.79* 3.95* 4.36 3.67* 3.53 3.79* 3.95* *Letters denote significant differences among means in the same column at = 0.05 based on Tukey-Kramer HSD test. (*) denotes sign ificant post-fire change within treatment (Tukey-Kramer HSD). P-values are for randomized block ANOVA / ANCOVA within columns (significant at = 0.05). 39

PAGE 40

Table 2-4. Mean maximum fire temperature (C) by treatment. Treatment 30cm80cm Mean Fire Temp. ( C ) Control Imazapyr Sulfometuron Sulfo + Hexa Hexazinone Control Imazapyr Sulfometuron Sulfo + Hexa Hexazinone ANOVA 276 (57) 275 (32) 306 (29) 260 (26) 252 (30) 181 (36) 190 (27) 197 (21) 166 (22) 257 (30) 276 (57) 275 (32) 306 (29) 260 (26) 252 (30) 276 (57) 275 (32) 306 (29) 260 (26) 252 (30) 181 (36) 190 (27) 197 (21) 166 (22) 257 (30) 181 (36) 190 (27) 197 (21) 166 (22) 257 (30) p = 0.420p =0.546 P-values for randomized block ANOVA are significant at = 0.05. Standard error in parentheses. 40

PAGE 41

0 5 10 15 20 25 30 35 ControlImazapyrSulfoSulfo + HexaHexazinone TreatmentMean Ericaceae Cover Pre-Harvest 4 yrs post-treatmentFigure 2-1. Pre-harvest (June 2001) and June 2006 percent cover of Eric aceae by treatment. Pre-harvest Ericaceae cove r was not uniform among treatment plots (ANOVA, p = 0.009). 4 years post-treatment, variance in Ericaceae cover by treatment approached significance (ANCOVA, p = 0.060). 41

PAGE 42

0 10 20 30 40 50 ControlImazapyrSulfoSulfo+HexaHexa TreatmentMean % Shrub Cover Ericaceae Serenoa Quercus Ilex 0 5 10 15 20 25 30 35 40 45 ControlImazapyrSulfoSulfo+HexaHexa TreatmentMean % Shrub Cover Ericaceae Serenoa Quercus IlexJune 2006 (4 years after treatment) June 2007 (post-fire)a) b) Figure 2-2. a) Mean percent shrub cover f our years after trea tment (June 2006) for Quercus spp., I. glabra, S. repens and Ericaceous shrubs (Vaccinium spp. K. hirsuta Gaylussacia spp., and L. lucida). b) Mean percent shrub cover in J une 2007, four mont hs after fire. 42

PAGE 43

0 5 10 15 20 25 30 35 40 ControlImazapyrSulfoSulfo+HexaHexa TreatmentMean % Cover Pre-Fire Post-FireNon-Ericaceaea) b) 0 5 10 15 20 25 30 35 ControlImazapyrSulfoSulfo + HexaHexa TreatmentMean % Cover Pre-Fire Post-FireEricaceaeFigure 2-3. Mean percent shrub cover before (June 2006) and after fire (June 2007) for a) Quercus spp., I. glabra, and S. repens and b) Ericaceae. Differences between and within treatments are not significant on either graph. 43

PAGE 44

0 10 20 30 40 50 60 ControlImazapyrSulfoSulfo+HexaHexazinone TreatmentMean % Herbaceous Cover Pre-Fire Post-Firec a bc abc ab* *Figure 2-4. Mean % herbaceous cover (including both forbs and graminoids) 4 years after treatment (Pre-fire) and 5 year s after treatment (Post-fire). Pre-fire means not sharing the same letter are significantly different (Tukey-Kramer HSD, = .05). (*) indicates significant pre-fire / post-fire differences within treatments (Tukey-Kramer HSD, =.05). 44

PAGE 45

0 10 20 30 40 50 60 70 80 ControlImazapyrSulfoSulfo + HexaHexazinoneMean % Cover Herbaceous Shrubs Ericaceous Shrubs 0 10 20 30 40 50 60 70 80 ControlImazapyrSulfoSulfo+HexaHexazinone TreatmentMean % Cover 0 10 20 30 40 50 60 70 80 ControlImazapyrSulfoSulfo + HexaHexazinoneMean % Cover Herbaceous Shrubs Ericaceous Shrubs 0 10 20 30 40 50 60 70 80 ControlImazapyrSulfoSulfo+HexaHexazinone TreatmentMean % Covera) b)June 2006 (4 years after treatment) June 2007 (Following Feb. 2007 fire)Figure 2-5. Overall vegetation cove r in a) June 2006 (4 years afte r treatment) and b) June 2007 (4 months after fire). Cover is separated into herbaceous, shrub, and Ericaceous shrub components. 45

PAGE 46

0 5 10 15 20 25 30 35 40 ControlImazapyrSulfoSulfo+HexaHexa TreatmentMean % Cover Wiregrass Pre-Fire Post-Firea) b) 0 5 10 15 20 25 30 35 40 Pre-harvest3 Mos 9 Mos 4 yrsPost-fire TimeMean % cover Wiregrass Control Imazapyr Sulfo Sulfo+Hexa Hexab b a ab ab* *Figure 2-6. a) Mean % cover of wiregrass four years after initial treatment (pre-fire) and five years after treatment (post-fire). Pre-fi re means not sharing the same letter are significantly different (Hsus MCB, = 0.05). (*) indicates a significant prefire/post-fire differen ce (Tukey-Kramer HSD, = 0.05). b) Mean % wiregrass cover within treatments over the course of the 5year experiment. First-year results from Ranasinghe (2003 and unpublished data), and Ranasinghe et al. (2005). 46

PAGE 47

0 10 20 30 40 50 60 70 80 ControlImazapyrSulfoSulfo+HexaHexa TreatmentMean Shrub Stem Density Pre-Fire Post-Fire* a) 0 20 40 60 80 100 120 Pre-harvest3 Mos9 Mos 4 yrsPost-fire TreatmentMean Shrub Stem Density Control Imazapyr Sulfo Sulfo+Hexa Hexab) 0 20 40 60 80 100 120 Pre-harvest3 Mos9 Mos 4 yrsPost-fire TreatmentMean Shrub Stem Density Control Imazapyr Sulfo Sulfo+Hexa Hexab)ab b a ab aFigure 2-7. a) Mean shrub stem density (stems/m2) before (June 2006) and after (June 2007) fire for Quercus spp., I. glabra, and S. repens Pre-fire means not sharing the same letter are significantly different. (*) indicates a significant change following fire (TukeyKramer HSD, = 0.05). b) Changes in shrub stem density within each treatment over the course of the 5-year experiment. Fi rst-year results from Ranasinghe (2003 and unpublished data), and Rana singhe et al. (2005). 47

PAGE 48

1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 Pre-treat3 Mos9 Mos4 yrsPost-Fire TimeMean Shannon Diversity Control Imazapyr Sulfo Sulfo+Hexa Hexa 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 ControlImazapyrSulfoSulfo+HexaHexa TreatmentMean Shannon Diversity Pre-Fire Post-Fire* *a) b)Figure 2-8. a) Mean species diversity (Shannon Index) by tr eatment four years after treatment (Pre-fire) and five years after treatment (Post-fire). Pre-fire differences among treatments are not significant (ANOVA, p = 0.13). (*) indicates a significant prefire/post-fire differen ce (Tukey-Kramer HSD, = 0.05). b) Changes in species richness within each treatment over the cour se of the 5-year experiment. First-year results from Ranasinghe (2003 and unpublishe d data), and Ranasi nghe et al. (2005). 48

PAGE 49

Control Imazapyr Sulfo+ Hexa Sulfo Hexazinone Control Imazapyr Sulfo+ Hexa Sulfo Hexazinone Axis 1 Axis 2Figure 2-9. NMDS ordination of plant communities Each point represents one plant survey quadrat. Chance-corre cted within-group ag reement (A) = 0.02 ( = 0.05, p = 0.01). 49

PAGE 50

Figure 2-10. a) relationship between mean Quercus cover and mean fire temperature (C) (r2 = 0.06, p = 0.03) in treatment subplots, and b) relationship between mean Ilex cover and mean fire temperature (C) (r2 = 0.13, p = 0.002) in treatment subplots. 0 100 200 300 400 500 600Fire Temperature at 30cm 0 10 20 30 40 50 Quercus % Cover 0 100 200 300 400 500 600Fire Temperature at 30cm 0 5 10 15 20 25 30 Ilex % Cover 50

PAGE 51

CHAPTER 3 HERBICIDE AND SUSTAINABLE L ONGLEAF PINE SILVICULTURE Introduction Perhaps more than any other forested ecosy stem in the United States, the longleaf pine ( Pinus palustris Mill.) savanna lends its elf to simultaneous management for timber and biodiversity. The understory communities asso ciated with longleaf pine are dominated by herbaceous plants adapted to frequent ground layer disturbance in the form of fire, and regular overstory disturbance by lightning, windthrow, an d hurricanes (Platt and Rathbun 1993, Platt et al. 2002). In longleaf pine systems with a freque nt fire regime and an intact seedbank, most native herbaceous species respond positively to the increased light levels in newly-opened canopy gaps (Keddy et al. 2006, Platt et al. 2006), and it has been hypothe sized that many of these species are such weak competitors for light that they could be considered fugitive or ephemeral (Keddy et al. 2006). In one case, remova l of the overstory resulted in a 20% increase in the number of understory species in gaps within five years, and surveys of older gaps indicate that similar increases have been maintained ove r several decades (Platt et al. 2006). Wiregrass ( Aristida stricta Michx.), the dominant bunchgrass in most longleaf pine systems and a keystone species with regard to pyrogenic ity, is also released in response to overstory removal (McGuire 2001). Since the renowned plant biodiversity of th e longleaf pine ecosystem resides largely in the herbaceous component (Walker and Peet 1983, Kirkman et al. 2001), the removal of large overstory trees for timber harvest has the potential to benefit the ecosystem. However, there are several important caveat s to the analogy between silviculture and natural disturbance in longleaf pine systems (Mitchell et al. 2006). One such caveat is that the removal of overstory pines across a large area also releases shrubs (Brockway et al. 1998, Kush et al. 1999, McGuire et al 2001, Kirkman et al. 2007, Pecot et al 2007), which, in the absence of 51

PAGE 52

adequate control, eventually leads to the exclusion of herbaceous plants (Gilliam and Platt 1999, Glitzenstein et al. 2003). The release of shrubs following oversto ry harvest is likely to be more extensive on a site with a past history of fire suppression, due to the preponderance of underground hardwood rootstocks established during periods wit hout fire (Olson and Platt 1995; Drewa et al. 2006). Shrub invasion excludes herbaceous specie s not only through competition for light and resources, but through the alteration of fire regimes away from the high-frequency fire intervals perpetuated by wi regrass and pine needles. Hardwoods produce litter of lower flammability than pines or wiregrass, which can l ead to longer fire return intervals in gaps, ultimately exerting a negative feedback on the systems pyrogenicity and perpetuating a hardwood midstory (Williamson and Black 1981, Rebe rtus et al. 1993, Jacq main et al. 1999). When a large portion of the overstory is removed, th is negative feedback cycle is likely to be exacerbated by the subsequent lack of needlefall and the potential loss of wiregrass due to harvest-related soil disturbance. These are the two fine fuel inputs essential to maintaining a frequent fire return interval in a longleaf pine ecosystem, and w ithout them the keystone species in the system will not regenerate (Williamson and Black 1981, Clewell 1989). The ability of longleaf pine seedlings to surv ive and emerge from the grass stage will also be impacted if there are not adequate fine fuels to carry fire through the system and keep hardwoods suppressed. Longleaf seedlings are hi ghly intolerant of competition and will not initiate height growth if an excessive amount of competing vegetation is present (Boyer 1963, Haywood 2000, Ramsey et al. 2003). One approach to these problems, which has re ceived increasing attention in recent years, is to develop an uneven-aged silvicultural system that replicates natural smaller-scale patterns of disturbance and succession so as not to disrupt th e overall continuity of fine fuels and understory 52

PAGE 53

dynamics (Palik et al. 2002, Mitchell et al. 2006, Brockway et al. 2006). It has long been observed that longleaf pine seed lings regenerate only where di sturbances have created canopy openings (Boyer 1963). Several rece nt studies have looked at patte rns of regeneration in natural gaps as well as gaps created by groupor single-tree selection. Naturally-regenerating longleaf pine seedlings on a xeric sandhill were clustered near the center of gaps but were absent around the outer edge of each gap, an observation whic h led Brockway and Outcalt (1998) to propose the existence of a 12-16m seedling exclusion zone resulting from 1) higher fire intensity due to high litter accumulation at the bases of trees ar ound the gap edge, and 2) competition with the root systems of adult pines around the gap edge. Some studies of longl eaf pine regeneration have supported this hypothesis. As basal area in a mature longl eaf pine stand increased, firerelated seedling mortality also increased and seedling size decr eased (Grace and Platt 1995), while higher light intensity and N availability in gap centers were correlated with seedling survival and growth (Palik et al. 1997). Though subsequent research has not found evidence for a seedling exclusion zone per se, suppressed growth with increasing proximity to adults has been shown in all studies. Container-g rown seedlings planted in gaps exhibited maximum growth in response to high light intensity at the center of gaps as small as .1-ha, but no strong response to nutrient availability (McGuire et al. 2001). Others have reported that seedling survival was actually higher toward the outer edges of gaps, though the seedlings able to survive in the center were significantly larger than t hose closer to the edges (Gagnon et al. 2003, Pecot et al. 2007). Results from a trenching experiment suggest that this tradeoff is due to underground facilitation by adult longleaf pines at gap edge s vs. greater light availability at the gap center (Pecot et al. 2007). Trenching also revealed that hardwoods were immediately released when belowground competition from mature canopy trees was remove d (Pecot et al. 2007). The upshot of these 53

PAGE 54

recent studies is that small group selection a nd single-tree selecti on are recommended as silvicultural treatments for l ongleaf pine systems in which bot h biodiversity conservation and timber yield are desired. These methods enable new seedlings to establish at acceptably high survival rates without a major disruption in the supply of fine fuels or excessive hardwood release. But what of a system in which there are no longleaf pines in the overstory, and the desired outcome is a comple te conversion of the overstory from slash pine (Pinus elliottii Engelm.) to longleaf pine? The same issues with hardwood control and fine fuel supply must be considered. It has recently been shown that a gr adual replacement approach is also an effective method for making the conversion from slash to longleaf. The single-tree selection method was used to create plots with variable basal area densities, into whic h longleaf pine seedlings were introduced in conjunction with various mechanical and herbicidal hardwood control treatments. High rates of seedling survival were achieved at all densities, though rates of grass stage emergence were low and were not expected to in crease until the next harvest cycle (Kirkman et al. 2007). The second approach to slash-longleaf convers ion, and the one employed in this study, is to remove the entire slash pine canopy, replant with longleaf pine seedlin gs, and use mechanical and herbicidal treatments to control competing ve getation. The natural analog of this type of clearcut is a very larg e-scale disturbance, such as a hurri cane (Mitchell et al. 2006). Since both longleaf pine seedlings and native herbaceous sp ecies respond positively to full sunlight, this method, too, has the potential to restore a long leaf pine canopy and a diverse, herbaceous understory. Some of the potential benefits of this approach, as compared to single-tree and group-selection methods, are: 1) less logging dama ge to the site over time because only one 54

PAGE 55

harvest is required, 2) lower management costs, 3) lower personnel skill level and inventory information needs, 4) higher short-term timber output, and 5) better equipment access and maneuverability for effective site preparation in an area with severe understory competition, such as a saw palmetto-dominated flatwoods (Palik et al. 2002, Brockway et al. 2006). If desired, subsequent management of the new longleaf pi ne stand can follow a single-tree or groupselection prescription in order to perpetuate the structural and functional diversity of the system over the long-term. The biggest challenges facing this method will be 1) to achieve control of competing shrubs without negatively impacting native understory vegetation, 2) to maintain or enhance wiregrass cover, and 3) to successf ully apply fire to the system in spite of the loss of overstory fine fuel inputs. The studies described above indicate that supplemental hardwood control in the form of herbicide or mechanical treatment will almost always be necessary after clearcutting a pine overstory, because of the rele ase of hardwoods and the lack of fine fuels for prescribed fire. Even in a groupor single-tree selection syst em, additional hardwood control may be necessary if pyrogenicity is already comp romised due to a past histor y of fire suppression and an entrenched shrub component. In the single-tr ee selection study described above, the herbicide treatment was the only one in wh ich hardwood stem density did not increase over the course of the study (Kirkman et al. 2007). Following overstory harvest, longleaf pine pl antings must be carefully managed in order to achieve survival and growth rates adequate for stand replacement. A reasonable range of survival rates for containerized pl antings is 60-75%, but in order to achieve this range of survival rates, site preparation is necessa ry before seedlings are planted (Brockway et al. 2006). Grasses and other herbaceous species are the most seriou s short-term competitors for newly established 55

PAGE 56

longleaf seedlings, while woody species control is im portant for the development of fine fuels to perpetuate the ecosystem in the longer-term (Haywood 2005). On a frequently-burned site dominated by herbaceous species, prescribed fire may provide adequate site preparation, but on sites with heavy woody plant competition, mechan ical or herbicide treatments are necessary (Brockway et al. 2006). Intens ive mechanical site preparatio n techniques such as bedding, disking, and harrowing are effective for longleaf pine establishment (Johnson and Gjerstad 2006), but are very detrimental to wiregrass (Cle well 1989) and are therefor e not options on sites where understory restoration is a goal. Ther efore, herbicide, which removes competing vegetation without disturbi ng the soil, has potential utility as a component of a sustainable forest management strategy. In order to gain a better unders tanding of the role of herbicid e in sustainable longleaf pine silviculture, we compared the effects of five post-planting, banded herbicide treatments on the growth and survival of containeri zed longleaf pine seedlings in a coastal flatwoods in the Florida panhandle: 1) Sulfometuron methyl (Oust, E.I. du Pont de Ne mours and Company, Wilmington, Del.), 2) Imazapyr (Arsenal, BASF Corporation, Research Triangle Park, NC), 3) Hexazinone (Velpar L, E.I. du Pont de Nemours and Co mpany, Wilmington, Del.), 4) A tank mix of Oust and Velpar L, and 5) a no-herbicide control. These herbicide treatments were chosen because they are commonly used for loblolly ( Pinus taeda L.) and slash pine release in the Southeast, but their effects on longleaf pine growth and flatwoods community composition are not well studied. A recent survey by Shepard et al. (2004) reported that the forestry herbicide most commonly used in the Southeast was Arsenal, followed by tank mixes of Arsenal + Oust and Oust + Velpar. These tank mixes are used to achieve broader spectrum control, but as a result they may also have more negative implications for native plant communities (Keyser 56

PAGE 57

and Ford 2006). Sulfometuron methyl and hexazinone are effective as ove r-the-top pine release applications on longleaf pi ne seedlings (Haywood 2000, Ramsey and Jose 2004, Haywood 2005), but effects of a mixed applic ation on longleaf seedlings or on the understory of a natural flatwoods system have not been studied. Imazapyr mixed with triclopyr was effective as a preplanting application to release longleaf pine seedlings on a poor ly-drained site (Knapp et al. 2006), but information on imazapyr as an over-thetop release treatment for longleaf is also limited. We examined the longer-term effects of these herbicide treatments on longleaf pine seedlings five growing seasons after initial application (see Ranasinghe 2003 for first year results). We then conducted a dormant-season pres cribed fire on all treatment plots and assessed post-fire survival at the beginning of the sixth growing season. Methods Study Area The study was conducted at the Point Washi ngton State Forest, which is located on the gulf coast of Walton County, Florida (30 20 N, 86 04 W). The 20-acre study site is part of a larger mesic flatwoods/wet flat woods matrix, dominated by saw palmetto ( Serenoa repens W. Bartram), gallberry ( Ilex glabra A. Gray), rhizomatous oaks ( Quercus minima Small and Q. elliottii Wilbur), blueberries ( Vaccinium myrsinites Lam. and V. darrowii Camp.), huckleberries ( Gaylussacia dumosa Torr. and A. Gray and G. frondosa L. Torr and A. Gray ex. Torr), bluestem grasses ( Andropogon virginicus var. glaucus (L.) Heck and Andropogon virginicus L.) and wiregrass ( Aristida stricta ). The majority of the site can be classified as mesic, but due to subtle topographi cal variation across the site some areas are at the wetter end of the mesic flatwoods spectrum and some are at the drier end. For th is reason, we used a randomized complete block design, with three bl ocks laid out in mesic-wet areas and three 57

PAGE 58

blocks laid out in mesic-dry areas. A pre-trea tment vegetation survey identified wetter areas by the presence of indicator species such as sedge s (Cyperaceae). The soils of the study site are sandy, siliceous, thermic, aeric alaquods of the Leon series (Spodosols consisting of deep, poorly to very poorly drained soils derived from marine parent material (Ranasinghe 2003)). Yearly weather patterns during the study period were va riable. Total annual precipitation during the study period was 103cm in 2002, 200cm in 2003, 135cm in 2004, 146cm in 2005, and 106cm in 2006 (NOAA 2007). Rainfall in the months following the prescribed fire was very low, with totals of .5cm in March 2007, 4cm in April 2007, and .3cm in May 2007 (NOAA 2007). Prior to August 2001, the 20-acre study site was a planted slash pine stand with a mean stand age of about 26 years, which had been burned on a 3-year rotation by the Division of Forestry since its acquisition of Point Washington State Forest in 1992. A pre-harvest, baseline data survey was conducted in the areas slated to become study blocks in June 2001. In August 2001, the entire overstory was harves ted and the site was prepared for the study with a single, light roller chopping followed by a prescribed burn in October 2001. In December 2001, six study blocks, each divided into five 36.6m x 24.4m treatment plots, were established according to the aforementioned randomized complete block design. Each treatment plot was hand-planted with 100 containerized longleaf pine seedlings laid out in 10 rows, with 3.1m between rows and 1.8m between the seedlings in each row. Treatments Within each block, each 100-seedling plot was randomly selected to receive one of five herbicide treatments in March of 2002: 1) Sulfometuron methyl (Oust) at 0.26 kg a.i./ha. This is a moderate application rate for sulfometuron methyl; other studies have us ed 0.16 kg a.i./ha (Lauer and Glover, 1998) and 0.21 kg ai/ha (Shiver and Mart in 2002, Keyser and Ford 2006), while the manufacturer58

PAGE 59

recommended rate for herbaceous weed control in longleaf pine stands is 0.10 to 0.42 kg a.i./ha. Sulfometuron methyl is a selective herbicide primarily effective for controlling herbaceous species (Lauer and Glover 1998), and mixing with hexazinon e is recommended if broader spectrum control is desired (DuPont 2007). 2) Hexazinone: (Velpar L) at 0.56 kg a.i./ha. Recommended application rates for hexazinone range from 2.0 to 6.7 kg a.i./ha (E.I. Du Pont De Nemours), while rates successfully used to control hardwoods in sandhills ecosystems have ranged from 1.1 kg a.i./ha to 2.4 kg a.i /ha (Brockway et al. 1998, Provencher et al. 2001) The hexazinone application rate in our study was therefore very low. This rate was chosen because hexazinone had been deemed effective at the aforementioned rates, and we were interested in whether an even lower application rate would provide vegeta tion control. 3) Sulfometuron + Hexazinone mix (0.26 kg a.i./ha Sulfometuron methyl + 0.56 kg a.i./ha Hexazinone). This is a common tank mix used thr oughout the Southeast (Shepard et al. 2004) with the potential to control a broad sp ectrum of woody and herbaceous plants. 4) Imazapyr (Arsenal) at 0.21 kg a.i./ha. For a restoration project, this was a moderate application rate for imazapyr. Signifi cant hardwood control and habitat improvements have been obtained using imazapyr at 0.42 kg a. i./ha (Jones and Chamberlain 2004), and at all rates from 0.08 kg a.i./ha to 0.24 kg a.i..ha (t hough higher rates did not offer significant improvements over lower rates) (K eyser and Ford 2006). Arsenal is not labeled for use over longleaf pine at the recommended rates of 0.32 to 0.42 kg a.i./ha until after the end of the second growing season (BASF 2007), and has caused stunted growth in loblolly pi ne seedlings (Barber 1991). We chose the rate of 0.21 kg a.i./ha in order to determine whether a lower-thanrecommended application rate would prove less injurious to pine seedlings. 59

PAGE 60

The reasons for using a banded post-plan ting herbicide application on unshielded seedlings were both operational and environmenta l. Prior to harvest, the study site already contained many desirable understory species (including wiregrass) and only a few weedy species, none of which were exotic (Ranasinghe 2003). The banded application allowed us to pinpoint the delivery of herbicide to an area directly around planted seedlings, while native vegetation remained intact in the strips between rows. This application method also lowered the cost of herbicide treatment relative to a broadcast application. The herbicides were applied in a 1.2m ba nd over the seedling rows using a backpack sprayer. Within each 10-row treatment plot, three rows received a second herbicide application in March of 2003, three rows received a second app lication in April of 2003, and four rows were left with only the first-year treatment. The results from these subplots were pooled with the first year data for most parameters because of a l ack of significant differences between one and twoyear applications. Data Collection and Analysis In November 2006, the root collar diameter (RCD) and height to the top of bud were measured on every seedling. Following a prescribed fire in February of 2007 (Chapter 2), postfire seedling mortality was assess ed in June 2006. Survival rate was calculated as a percentage of each 10-tree, 56m2 treatment row. Rate of grass-stage release was calculated as a percentage of surviving seedlings in each treatment row with a bud height greater than 12cm, following Haywood (2000). Height and RCD co mparisons were made separately for seedlings in the grass stage (GS) and out of the grass stage (OOGS). The study addressed only the main effects of herbicide treatment, and tests of these eff ects were not dependent on the assumption of no treatment x block interaction. Block effects were therefore tr eated as random effects in a univariate ANOVA model with two independent variables: treat ment with Block&Random as 60

PAGE 61

a covariate. Data were log-transformed where necessary to meet the assumptions of ANOVA. Significant differences between treatments were separated with the Tukey-Kramer HSD test. Post-fire survival was analyzed with ANCOVA, us ing pre-fire survival as a covariate. All parameters were analyzed in JMP IN version 5 (SAS Institute, Inc.). Results Seedling Survival Five growing seasons after treatment (Novemb er 2006), survival rate varied significantly among herbicide treatments ( = 0.05, p < 0.0001) (Figure 3-1). The highest survival rates occurred in the hexazinone and control treatme nts (80% and 78.5%, respectively) (Figure 3-1). Survival rates were significantly lower in th e sulfometuron + hexazinone (sulfo + hexa) and imazapyr treatments (64.2% and 63.2%, respectively). It should be noted that the mortality in the imazapyr and sulfo + hexa treatments o ccurred shortly after herbicide application (Ranasinghe 2003), and there was little subsequent mort ality in any of the treatments. None of the treatment plots experienced significant mortality following fire (Figure 3-2). However, there were some slight differences in fire-related mortal ity among the different treatments, and as a result, surv ival rate no longer varied signifi cantly by treatment following fire (p = 0.238). In the imazapyr treatment, the surv ival rate dropped by only .7% following fire, as compared to decreases of 3% to 3.5% in the other treatments. Percent Out of Grass Stage Five growing seasons after herbicide application (November 2006), the mean grass-stage release rate varied significantly among treatments (p < 0.0001). The highest level of grass-stage release after four years was obs erved in the imazapyr treatment, though both sulfo + hexa and hexazinone also significantly ra ised the release rate relative to the control (Figure 3-3). 61

PAGE 62

Growth: Out of Grass Stage Seedlings The mean RCD, height, and SVI of seedlin gs out of the grass stage (OOGS) varied significantly among treatments (p = 0.0013, p < 0.0001, and p < 0.0001, respectively). In all parameters, imazapyr showed the greatest improveme nt relative to the co ntrol, though sulfo + hexa also significantly improved height relative to the contro l (Table 3-1, Figure 3-4, Figure 35). RCD of OOGS seedlings varied less among treat ments than height or SVI, which is to be expected due to the preferential investment in height growth characteri stic of longleaf pine seedlings in the bolting stage. Growth: Grass Stage Seedlings Root collar diameter (RCD), height, and SVI of grass-stage seedlings varied significantly among treatments (p < 0.0001, p = 0.003, and p = 0.0015, respectively) (Table 3-1, Figure 3-5), with the smallest seedlings in the sulfometuron treatment. Among s eedlings that were still in the grass stage after five growing seasons, none of the herbicide treatments showed significant improvements relative to the control. This pa ttern has been observed since the first growing season after treatment (R anasinghe 2003), though it was unclear whether the lower seedling size was due to seedling injury fr om sulfometuron or the poor sh rub control evident in the sulfometuron treatment since the fi rst growing season (Chapter 2). Discussion Five growing seasons after treatment, imazapyr plots had a grass stag e release rate more than twice that of the control, and the mean stem volume index of OOGS seedlings in the imazapyr treatment was also more than double that in the control. Although direct application of imazapyr to recently-planted seedlings caused significant mortality in the first year after planting, the overall post-fire surviv al rate of 62.5% in this treatmen t is still within the acceptable range of 60-75% recommended for adequate over story restocking (Brockway et al. 2006). 62

PAGE 63

Seedlings in this treatment experienced a fire-r elated mortality rate of only .06%, whereas fire mortality was around 3% in all of the other treatments. As a resu lt, the differences in survival rate between imazapyr and the other treatments were no longer significant following fire. Though the imazapyr treatment provided the be st overall pine re lease results, both hexazinone and the sulfometuron + hexazinone ta nk mix also showed promise as over-the-top longleaf pine release treatments for flatwoods ecosystems. Significant improvements in the grass-stage release rate were ach ieved with both of these treatments, and the OOGS seedlings in plots treated with the tank mix were significantly ta ller than those in the control. An absolute comparison between the effectiveness of thes e two formulations versus imazapyr is not warranted because we used a very low rate of hexazinone. However, based on success others have had using hexazinone for longleaf pi ne release (Haywood 2000, Ramsey and Jose 2004, Haywood 2005), we can expect that at a high er rateboth alone and in the tank mixthis herbicide would have compared more favorably with imazapyr. Contrary to the results previously observed with longleaf pine seedlings planted in an old field (Ramsey and Jose 2004), sulfometuron alone did not provide any pine release benefits over the control. This was most likely due to the higher degree of woody compe tition on this site, against which sulfometuron was significantly less effective than the other treatments (Chapter 2). For grass stage seedlings, variation among treatments was much less pronounced. Grass stage seedlings in the sulfometuron treatment were significantly smaller than those in the control, but none of the other treatments di ffered significantly from the cont rol in any of the grass stage parameters. A possible explanation for this patte rn is that as vegetati on regrew in the years following application, any seedlings not released relatively early in the process were suppressed and remained in the grass stage. As differen ces in understory vegeta tion between treatments 63

PAGE 64

became less pronounced over the years, differences in biomass accumulation rates between suppressed seedlings probably al so decreased. In contrast, th e earlier in the process a pine seedling was released, the better its chances ma y have been of gaining a competitive advantage for light, water, and soil resources. The early gains made by seedlings in the imazapyr plots (Ranasinghe 2003) apparently translated into an increasing competitive advantage over the years, leading to much larger seedlings in these plot s even as vegetation differences between plots became less pronounced. In a scenario such as this, where the dual goals of silviculture and ecosystem restoration dictate light site prepara tion and banded herbicide a pplication in spite of the heavy flatwoods competition, early release fr om the grass stage may be critical to the ultimate success of the planted pines. The site preparation burn conducted followi ng overstory harvest in 2001 consumed the last of the accumulated slash pine needles, and, though none of the herbicide treatments decreased wiregrass cover, distri bution of wiregrass across the site was patchy (Chapter 2). Five years after initial herbicide treatment, we had di fficulty applying a prescribed fire of sufficient intensity to control shrubs a phenomenon which was undoubtedly exacerbated by length of time since the application of herbicide and the need to apply fire during the dormant season (Chapter 2). The lack of fine fuels was also probably a major factor limiting our ability to apply fire effectively, given the key ro le played by dead pine needles in the pyrogenicity of a longleaf pine system (Williamson and Black 1981, Noss 1989). Our results support the theory proposed by Kirkman et al. (2007) that cl earcutting, even when the goal is a total canopy conversion from slash pine to longleaf pine, may ultimately hi nder the success of a restoration project. The overstory and understory effects of imazapyr, hexazinone, and the sulfometuron + hexazinone tank mix lined up well with the restora tion goals for this site. These treatments, 64

PAGE 65

especially imazapyr, released pines while also maintaining or improving wiregrass cover and overall cover of native herbaceous species. While tank mixes have been widely studied in silvicultural applications, they have understandably been avoided as habitat treatments thus far, out of concern that a broad spectrum mix will nega tively impact native species (Keyser and Ford 2006). However, our results indicate that this tank mix not only released pines, but had positive impacts on the understory, which might have been even more significant if a higher level of hexazinone had been included in th e mix (Chapter 2). An important caveat to this conclusion is that the use of a banded applica tion method may have been a key to the positive result. On pine plantations in the Virginia piedmont, broad cast application of an imazapyr + sulfometuron methyl tank mix significantly lowered herbaceous speci es diversity relative to the control and to imazapyr-only treatments, and banded application wa s suggested as an alternative for sites with wildlife habitat concerns (Keyser and Ford 2006). Recent studies of sustainable silviculture in longleaf pine systems have cited the potential negative impacts of herbicide as a justification for the use of gr oup or single-tree selection, since these harvesting methods should enhance the effec tiveness of prescribed fire as a management alternative to herbicide (Mitchell et al. 2006, Peco t et al. 2007, Kirkman et al. 2007). However, we observed only beneficial impacts on understo ry species following chemical treatment, indicating that herbicide can in fact play a positive supporting role in conservation-oriented longleaf pine silviculture. Based on the result s of this study, we recommend banded application of imazapyr as an ecological re storation tool for flatwoods sites with aggressive understory shrubs. 65

PAGE 66

Table 3-1. Root collar diameter (mm), height (mm), and stem volume index (basal area2 x height) of grass stage and OOG S seedlings five years after initial herbicide treatment. Root Collar Diameter (mm)Height (mm) Stem Volume Index Out of Grass Out of Grass Out of Grass Stage 25.4a23.6ab22.3b25.0a25.1a25.4a23.6ab22.3b25.0a25.1a Co *Letters denote significant differences between means in the same column at = 0.05 based on Tukey-Kramer HSD test P-values are for randomized block ANOVA within columns (significant at = 0.05). ntrol Imazapyr Sulfometuron Sulfo + Hexa Hexazinone Control Imazapyr Sulfometuron Sulfo + Hexa Hexazinone Treatment ANOVA Grass Stage 39.5abc41.8a38.6c39.4bc40.6ab39.5abc41.8a38.6c39.4bc40.6abp < 0.0001 Stage Grass Stage Stage Grass Stage 63.1ab70.2a62.2b71.0a65.4ab63.1ab70.2a62.2b71.0a65.4ab492ab467ab388b519a499a 492ab467ab388b519a499a492ab467ab388b519a499a389.1c663.6a490.8bc513.1b484.1bc389.1c663.6a490.8bc513.1b484.1bcp < 0.0001 p = 0.0015 6332b14,542a9437b10,042b6332b6332b14,542a9437b10,042b6332bp < 0.0001 p = 0.003 p = 0.0013 66

PAGE 67

20% 30% 40% 50% 60% 70% 80% 90% ControlImazapyrSulfoSulfo + HexaHexa Treatment% Surviving (trees/row )a b a b a Figure 3-1. Mean survival rate by treatment after five growing seasons, calculated as # surving / # planted in each treatment row. Means not sharing the same letter are significantly different (Tukey-Kramer HSD). 67

PAGE 68

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% ControlImazapyrSulfoSulfo+HexaHexa Treatment% Surviving (trees/row) Pre-Fire Post-Fire Figure 3-2. Percent seedling survival (trees/row), before and after fire. None of the treatments experienced significant mortality du e to fire (Tukey-Kramer HSD, = 0.05). 68

PAGE 69

0% 10% 20% 30% 40% 50% 60% 70% 80% ControlImazapyrSulfoSulfo + HexaHexa Treatment% out of grass stagec a c b ab Figure 3-3. Mean grass-stage releas e rate by treatment after five gr owing seasons, calculated as # released / # surviving in each treatment row. Means not sharing the same letter are significantly different (Tukey-Kramer HSD). 69

PAGE 70

100 200 300 400 500 600 700 800 ControlImazapyrSulfoSulfo + HexaHexazinone TreatmentMean height (mm) c a bc b bc Figure 3-4. Mean height (mm) of out of grass stage seedlings by treatment. Means not sharing the same letter are significantly different (Tukey-Kramer HSD, = 0.05) 70

PAGE 71

100 150 200 250 300 350 400 450 500 550 600 ControlImazapyrSulfoSulfo + HexaHexazinone TreatmentMean SVI 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 ControlImazapyrSulfoSulfo + HexaHexazinone TreatmentMean SVIab b ab a aa)b a b b bb)Grass Stage Out of Grass Stage Figure 3-5. Mean stem-volume index (SVI) of a) grass stage seedlings and b) OOGS seedlings by treatment. Means not sharing the same letter are significantly different (TukeyKramer HSD, = 0.05). 71

PAGE 72

CHAPTER 4 SUMMARY AND CONCLUSIONS One of the biggest challenges to succe ssful restoration of the longleaf pine ( Pinus palustris Mill.) ecosystem is the persis tence of shrubs in the unde rstory, which suppress longleaf pine seedlings as well as native herbaceous plants. Herbicide can be used as a supplement to fire in order to enhance shrub control, but must be studied carefully because of the potential for negative impacts on native plants. Information is lacking about the effects of herbicide on natural longleaf pine flatwoods communities. We used a banded a pplication of three herbicides and one tank mix as shrub cont rol treatments following harves t of a slash pine stand and replanting with containerized longleaf pine seed lings in a mesic-wet flatwoods. The herbicides tested were Arsenal (imazapyr), Oust (sulfometuron methyl), Velpar L (hexazinone), and an Oust + Velpar L tank mix. Imazapyr herbicide significantly improved longleaf pine seedling growth, due to its effectiveness at controlling competing shrubs during the first growing seasons after planting. Though these shrub control effects were shortlived, the short term shrub control allowed herbaceous species to gain more of a foothold in the understor y, and significant increases in herbaceous cover relative to the c ontrol were recorded in imazapyr -treated plots four years after initial treatment. In a longleaf pine silviculture project with both timber and conservation goals, this was a very desirable result, because of the high level of biodive rsity in the herbacous component of longleaf pine ecosystems (Walker and Peet 1983, Kirkman et al. 2001). No negative impacts on understory species richness, diversity, evenness, or community composition resulted from the application of this herbic ide. Though over-the-top application of imazapyr caused significant longle af pine seedling mortality, the mean survival rate of seedlings in imazapyr-treated plots was 62.5% after five gr owing seasons and 61.8% after a fifth-year 72

PAGE 73

prescribed fire, which is stil l within the acceptable range of 60-75% recommended for adequate overstory restocking (Brockway et al. 2006). We therefore recommend imazapyr for use in sustainable longleaf pine silvic ulture, as it enhanced both long leaf pine seedling growth and herbaceous cover in the understory. Hexazinone, both alone and in a tank mix with sulfometuron, also showed promise as both a longleaf pine rele ase treatment and as an understory restoration tool. Four years after treatment, hexazinone plots had significantly higher herbaceous cover than the control, and plots treated with a sulfometuron + hexazinone tank mix had significantly higher wiregrass cover than the control. Significant improvements in the gras s stage release rate were achieved with both of these treatments, and released seedlings in plots treated with the tank mix were significantly taller than those in the control. An absolute comparison between these two formulations versus imazapyr is not warranted because we used a ve ry low rate of hexazinone. However, based on success others have had using hexazinone for longleaf pine release (Haywood 2000, Ramsey and Jose 2004, Haywood 2005) and rest oration (Brockway et al. 1998, Brockway and Outcalt 2000, Provencher et al. 2001) we can expe ct that at a higher rateboth alone and in the tank mixthis herbicide would have compared more favorably with imazapyr. Sulfometuron methyl did not improve pine rel ease relative to the cont rol, and instead had a releasing effect on the tw o most common oak species ( Quercus minima and Quercus elliottii ), indicating that this is not an appropriate herbicide for flatwoods with a high density of oaks. However, in spite of having higher shrub cove r and lower herbaceous cover than the other herbicide treatments, sulfometuron plots also had the highest herbaceous species richness, diversity, and evenness four years af ter initial treatment. One reas on for this may have been that wiregrass cover in the sulfom eturon plots was somewhat lowe r than the other herbicide 73

PAGE 74

treatments, leaving more space for other small herbaceous species. These gains in species richness, diversity, and evenness are therefore un likely to persist through future fire cycles because it will be more difficult to apply prescribed fire to this treatment. In support of this prediction, we observed that su lfometuron was the only treatment in which species richness, diversity, and evenness did not increase following the prescr ibed fire in 2007. The long-term sustainability of any longleaf pine ecosystem restoration project depends on the managers ability to apply prescribed fire with enough frequency a nd intensity to control shrubs. The herbicides applied in this study pr ovided short-term shr ub control and had positive impacts on longleaf pine growth, wiregrass cove r, and overall herbaceous cover. However, complete removal of the slash pine overstory at the start of the pr oject resulted in a lack of fine fuels and presumably also encouraged aggressi ve regrowth of shrubs due to the loss of competition from mature pines (Brockway et al. 1998, Kush et al. 1999, McGuire et al. 2001, Kirkman et al. 2007, Pecot et al. 2007). These hindrances were most likely exacerbated by the length of time since herbicide a pplication (five years since first application and four years since second application), as well as the use of dormant -season fire due to the sites proximity to a highway and residential areas (Waldrop et al. 1992, Drewa et al. 2002, Drewa et al. 2006). Though three of the four herbicid es we tested were successful at releasing longleaf pine seedlings, the herbicide-related gains in wiregr ass and overall herbaceous cover may be lost in the coming years due to the aggressive resprouti ng of shrubs. We therefore make the following recommendation to those managing flatwoods si tes for both timber and biodiversity: when possible, some overstory pines shou ld be retained following harvest to serve as a source of fine fuels. Banded application of imazapyr is r ecommended in order to improve longleaf pine 74

PAGE 75

seedling growth and enhance herbaceous cover in the understory, espe cially on sites where growing-season fire is not an option. 75

PAGE 76

APPENDIX SPECIES LIST SpTable A-1. Species lis t. ecies name Common name Family Life Form Andropogon virginicus var. glaucus Chalky Bluestem PoaceaeGrass Andropogon virginicus var. virginicus Broomsedge BluestemPoaceaeGrass Aristida spiciformis Bottlebrush ThreeawnPoaceaeGrass Aristida stricta Wiregrass PoaceaeGrass Asclepias cinerea Carolina Milkweed ApocynaceaeForb Balduina uniflora Oneflower HoneycombheadAsteraceaeForb Baptisia lanceolata Gopherweed FabaceaeLegume Calamovilfa curtissii Curtiss's Sandgrass PoaceaeGrass Carphephorus odoratissimus Deer's Tongue AsteraceaeForb Chrysopsis gossypina subsp. cruseana Cruise's Goldenaster AsteraceaeForb Chrysopsis mariana Maryland GoldenasterAsteraceaeForb Cladoniaceae Deer Moss CladoniaceaeMoss Ctenium aromaticum Toothache Grass PoaceaeGrass Cynanchum angustifolium Gulf Coast SwallowwortApocynaceaeForb Cyrilla racemiflora Swamp Titi CyrillaceaeShrub Desmodium lineatum Sand Ticktrefoil FabaceaeLegume Desmodium strictum Pinebarren TicktrefoilFabaceaeLegume Dichanthelium aciculare Needleleaf WitchgrassPoaceaeGrass Dichanthelium ensifolium Cypress Witchgrass PoaceaeGrass Dichanthelium ovale Eggleaf Witchgrass PoaceaeGrass Dichanthelium erectifolium Erectleaf Witchgrass PoaceaeGrass Dichanthelium strigosum Roughhair WitchgrassPoaceaeGrass Drosera capillaris Pink Sundew DroseraceaeCarnivorous Eryngium yuccifolium Button RattlesnakemasterApiaceaeForb Eupatorium capillifolium Dog Fennel AsteraceaeForb Eupatorium compostifolium Yankeeweed AsteraceaeForb Eupatorium mohrii Mohr's ThoroughwortAsteraceaeForb Eupatorium pilosum Rough Boneset AsteraceaeForb Eupatorium rotundifolium Roundleaf ThoroughwortAsteraceaeForb Euphorbia inundata Florida Pineland SpurgeEuphorbiaceaeForb Eurybia eryngiifolia Thistleleaf Aster AsteraceaeForb Euthamia graminifolia Flattop Goldenrod AsteraceaeForb Gaylussacia dumosa Dwarf Huckleberry EricaceaeShrub Gaylussacia frondosa Dangleberry EricaceaeShrub Gelsemium sempervirens Yellow Jessamine GelsemiaceaeVine Gratiola hispida Rough Hedgehyssop VeronicaceaeForb Houstonia procumbens Roundleaf Bluet RubiaceaeForb Hypericum brachyphyllum Coastalplain St. John's WortClusiaceaeShrub Hypericum hypericoides St. Andrew's Cross ClusiaceaeShrub Hypoxis sessilus Yellow Star Grass HypoxidaceaeMonocot Nomenclature follows Wunderlin (2004). Species ar e divided into the following life forms: Grass = member of the Poaceae family, Sedge = member of the Cyperaceae family, Monocot = other monocot, Legume = member of the Fabaceae family, Carnivorous = carnivorous forb, Forb = other forb, Fern = member of the order Filicophyta, Shrub = woody un derstory species, Tree = woody overstory species. 76

PAGE 77

Table A-1. Continued. Species name Common name Family Life Form Ilex coriacea Giant Gallberry AquifoliaceaeShrub Ilex glabra Gallberry AquifoliaceaeShrub Ilex opaca American Holly AquifoliaceaeShrub Kalmia hirsuta Hairy Wicky EricaceaeShrub Lachnanthes caroliniana Carolina Redroot HaemodoraceaeMonocot Lachnocaulon auceps Whitehead Bog ButtonEriocaulaceaeMonocot Liatris gracilis Slender Blazing Star AsteraceaeForb Liatris tenuifolia Shortleaf Blazing StarAsteraceaeForb Lycania michauxii Gopher Apple ChrysobalanaceaeForb Lyonia lucida Fetterbush EricaceaeShrub Mimosa quadrivalvis Sensitive Vine FabaceaeLegume Mitreola sessilifolia Swamp Hornpod LoganiaceaeForb Muhlenbergia capillaris Cutover Muhly PoaceaeGrass Opuntia humifusa Prickly Pear CactaceaeForb Orchidaceae Orchid sp. OrchidaceaeMonocot Osmunda cinnamomea Cinnamon Fern OsmundaceaeFern Photinia pyrifolia Red Chokeberry RosaceaeShrub Pinus elliottii Slash Pine PinaceaeTree Piteopsis graminifolia Grassleaf GoldenasterAsteraceaeForb Polygala lutea Orange Milkwort PolygalaceaeForb Polygala nana Candyroot PolygalaceaeForb Polygonella polygama October Weed PolygonaceaeForb Pteridium aquilinum Bracken Fern DennstaedtiaceaeFern Pterocaulon pychnostachium Blackroot AsteraceaeForb Quercus elliottii Runner Oak FagaceaeShrub Quercus incana Bluejack Oak FagaceaeTree Quercus minima Dwarf Live Oak FagaceaeShrub Rhexia alifanus Savannah Meadow-BeautyMelastomataceaeForb Rhexia petiolata Fringed Meadow-BeautyMelastomataceaeForb Rhynchospora filifolia Threadleaf BeaksedgeCyperaceaeSedge Rhynchospora plumosa Plumed Beaksedge CyperaceaeSedge Sabatia brevifolia Shortleaf RosegentianGentianaceaeForb Scleria ciliata Fringed Nutrush CyperaceaeSedge Serenoa repens Saw Palmetto AracaceaeShrub Seriocarpus tortifolius Whitetop Aster AsteraceaeForb Smilax auriculata Earleaf Greenbrier SmilacaceaeMonocot Smilax glauca Wild Sarsparilla SmilacaceaeMonocot Smilax laurifolia Laurel-leaf GreenbrierSmilacaceaeMonocot Smilax pumila Sarsparilla Vine SmilacaceaeMonocot Solidago odora Sweet Goldenrod AsteraceaeForb Sporobolus junceus Pineywoods DropseedPoaceaeGrass Stillingia sylvatica Queen's Pleasure EuphorbiaceaeForb Stylisma patens Coastalplain DawnflowerConvolvulaceaeForb Stylisma villosa Hairy Dawnflower ConvolvulaceaeForb Symphiotrichum adnatum Scaleleaf Aster AsteraceaeForb Tephrosia hispidula Sprawling Hoarypea FabaceaeLegume Tragia urens Wavyleaf Noseburn EuphorbiaceaeForb Vaccinium darrowii Darrow's Blueberry EricaceaeShrub Vaccinium elliottii Elliott's Blueberry EricaceaeShrub Vaccinium myrsinites Shiny Blueberry EricaceaeShrub Vernonia angustifolia Tall Ironweed AsteraceaeForb Vitis rotundifolia Muscadine Grape VitaceaeVine Xyris caroliniana Carolina Yelloweyed GrassXyridaceaeMonocot 77

PAGE 78

LIST OF REFERENCES BASF Corporation. 2007. Arsenal AC label. http://www.cdms.net/LDat/ld542007.pdf BASF Agricultural Products, Research Triangle Park, NC. Behm, A.L., Duryea, M.L., Long, A.J., and W.C. Zipperer. 2004. Flammability of native understory species in pine flatwood and hardwood hammock ecosystems and implications for the wildland-urban interface. Inte rnational Journal of Wildland Fire 13:355-365. Boyd, R.S., Freeman, J.D., Miller, J.H., and M.B. Edwards. 1995. Forest herbicide influences on floristic diversity seven years after broadcast pine release treatments in central Georgia, U.S.A. New Forests 10: 17-37. Boyer, W.D. 1963. Development of longleaf pi ne seedlings under parent trees. USDA Forest Service, Southern Forest Experiment Stati on, Research Paper SO-4, New Orleans, LA. Brockway, D.G. and C.E. Lewis. 1997. Long-term effects of dormant season prescribed fire on plant community diversity, structure, and productivity in a longleaf pine-wiregrass ecosystem. Forest Ecology and Management 96: 167-183. Brockway, D.G. and K.W. Outca lt. 1998. Gap-phase regenerati on in longleaf pine wiregrass ecosystems. Forest Ecology and Management 106: 125-139. Brockway, D.G., Outcalt, K.W., and R.N. Wilkin s. 1998. Restoring longleaf pine wiregrass ecosystems: plant cover, diversity, and bioma ss following low-rate hexazinone application on Florida sandhills. Forest Ecology and Management 103: 159-175. Brockway, D.G. and K.W. Outcalt. 2000. Rest oring longleaf pine wi regrass ecosystems: hexazinone application enhances effects of pres cribed fire. Forest Ecology and Management 137: 121-138 Brockway, D.G., Outcalt, K.W., and W.D. Boye r. 2006. Pages 95-133 in Jose, S., Jokela, E., and Miller, D.L., editors. The Longleaf Pine Ecosystem: Ecology, Silviculture, and Restoration. Springer Science + Bu siness Media, LLC., New York. Clewell, A.F. 1989. Natura l history of wiregrass ( Aristida stricta Michx., Graminae). Natural Areas Journal 9: 234-245. Croker, T.C. Jr. 1975. Seedbed preparation aids natural regeneration of longleaf pine. USDA Forest Service Research Paper. Southe rn Forest Experiment Station No. SO-112. Daubenmire, R.F. 1959. Canopy coverage method of vegetation analysis. Northwest Scientist 33: 43-64. Drewa, P.B., Platt, W.J., and E.B. Moser. 2002. Fire effects on resprouting of shrubs in headwaters of southeastern long leaf pine savannas. Ecology 83(3): 755-767. 78

PAGE 79

Drewa, P.B., Thaxton, J.M., and W.J. Platt. 2006 Responses of root-crown bearing shrubs to differences in fire regimes in Pinus palustris (longleaf pine) savannas: Exploring oldgrowth questions in second growth system s. Applied Vegetation Science 9: 27-36. Dufrene, M. & P. Legendre. 1997. Species asse mblages and indicator species: the need for a flexible asymmetrical appr oach. Ecological Monographs 67: 345-366. DuPontTM. 2007. Velpar L Label. http://www.dupont.com/ag/vm/products/velparl.html E. I. du Pont de Nemours and Company,Wilmington, DE. DuPontTM. 2007. Oust XP Label. http://www.dupont.com/ag/us/prodi nfo/prodsearch/information/H65144.pdf E. I. du Pont de Nemours and Company,Wilmington, DE. Frost, C.C. 2006. History and Future of the L ongleaf Pine Ecosystem. Pages 9-47 in Jose, S., Jokela, E., and Miller, D.L., editors. The Longleaf Pine Ecosystem: Ecology, Silviculture, and Restoration. Springer Science + Business Media, LLC., New York. Gagnon, J.L., Jokela, E.J., Moser, W.K., and D.A. Huber. 2003 Dynamics of artificial regeneration in gaps within a longleaf pine flatwoods ecosystem. Forest Ecology and Management 172: 133-144. Gilliam, F.S. and W.J. Platt. 1999. Effects of long-term fire exclusion on tree species composition and stand structure in an old-growth Pinus palustris forest. Plant Ecology 140: 15-26 Glitzenstein, J.S., Streng, D.R ., and D.D. Wade. 2003. Fire frequency effects on longleaf pine ( Pinus palustris Mill.) vegetation in South Carolina a nd northeast Florida, USA. Natural Areas Journal 23(1): 22-37. Grace, S.L., and W.J. Platt. 1995. Effects of a dult tree density and fire on the demography off pregrass stage juvenile longleaf pine ( Pinus palustris Mill.) Journal of Ecology 83: 75-86. Hay-Smith, L. and G.W. Tanner. 1994. Responses of wildlife food plants to a forest herbicide. University of Florida Cooperative Extension Fact Sheet WRS-9. Haywood, J.D. 2000. Mulch and hexazinone herbicid e shorten the time longleaf pine seedlings are in the grass stage and increase height growth. New Forests 19: 279-290. Haywood, J.D. 2005. Effects of herbaceous and woody plant control on Pinus palustris growth and foliar nutrients through six growing seasons. Forest Ecology and Management 214: 384397. Harcombe, P.A., Glitzenstein, J.S., Knox, R.G., Orze ll, S.L., E.L. Bridges. 1993. Longleaf pine vegetation of the Southern Atlantic and Eastern Gulf Coast regions: a preliminary classification. Proceedings of the Tall Timbers Fire Ecology Conference 18: 83-104 79

PAGE 80

Hiers, J.K., Wyatt, R., and R.J. Mitchell. 2000. The effects of fire regime on legume reproduction in longleaf pine savannas: is a season selective? Oecologia 125: 521-530. Hurst, G.A. and W.E. Palmer. 1987. Vegetati on responses to imazapyr for pine release. Proceedings of the Southern Weed Science Society 40: 247. Hobbs, R.J., Currall, J.E.P., and C.H. Gimingham. 1984. The use of thermocolor pyrometers in the study of heath fire be havior. Journal of Ecology 72: 241-250. Iverson, L.R., Yaussy, D.A., Rebbeck, J., Hutchi nson, T.F., Long, R.P., and A.M. Prasad. 2004. A comparison of thermocouples and temperatur e paints to monitor spatial and temporal characteristics of landscape-scale prescribed fi res. International Journal of Wildland Fire. 13(3): 311-322. Jacqmain, E.I., Jones, R.H., and R.J. Mitchell. 1999. Influences of frequent cool-season burning across a soil moisture gradient on oak commun ity structure in longleaf pine ecosystems. American Midland Naturalist 141: 85-100. Jones, J.D., and M.J. Chamberlain. 2004. Efficacy of herbicides and fire to improve vegetative conditions for northern bobwhites in mature pine forests. Wildlife Society Bulletin 32(4): 1077-1084. Johnson, R. and D.J. Gjerstad. 2006. Restoring the overstory of longleaf pine ecosystems. Pages 271-295 in Jose, S., Jokela, E., and Miller, D.L., editors. The Longleaf Pine Ecosystem: Ecology, Silviculture, and Restor ation. Springer Science + Business Media, LLC., New York. Keyser, P.D., and V.L. Ford. 2006. Wildlife habitat and herbicides: an evaluation of a widely applied tank mix. Southern Journal of Applied Forestry 30(1): 46-51. Keddy, P.A., Smith, L., Campbell, D.R., Clark, M., and G. Montz. 2006. Patterns of herbaceous plant diversity in a Louisiana pine savanna. Applied Vegetation Science 9: 17-26. Kirkman, L.K., Mitchell, R.J., Helton, R.C., a nd M.B. Drew. 2001. Productivity and species richness across and environmental gradient in a fire-dependent ecosystem. American Journal of Botany 88(11): 2119-2128. Kirkman, L.K., Mitchell, R.J., Kaiser, M.J., Peco t, S.D., and K.L. Coffey. 2007. The perpetual forest: using undesirable species to bridge restoration. Journal of Applied Ecology 44: 604614. Knapp, B.O., Wang, G.G., Walker, J.L., and S. Cohen. 2006. Effects of site preparation treatments on early growth and surv ival of planted longleaf pine ( Pinus palustris Mill.) seedlings in North Carolina. Forest Ecology and Management 226: 122-128. 80

PAGE 81

Kush, J.S., R.S. Meldahl, and W.D. Boyer. 1999. Understory plant co mmunity response after 23 years of hardwood control treatm ents in natural longleaf pine (Pinus palustris ) forests. Canadian Journal of Forest Research 29: 1047-1054. Lauer, D.K. and G.R. Glover. 1998. Early pine response to control of herbaceous and shrub vegetation in flatwoods. Southern Journal of Applied Forestry 15(4): 201-208. Leach, M.K. and T.J. Givnish. 1996. Ecologica l determinants of species loss in remnant prairies. Science 273: 1555-1558. Litt, A.R., Herring, B.J. and L. Provencher. 2001. Herbicide effects on ground-layer vegetation in southern pinelands, USA: a review. Natural Areas Journal 21(2): 177-188. McNease. L.L. and G.A. Hurst. 1991. Preh arvest treatment with imazapyr for seed tree regeneration and habitat improvement. Proceed ings of the Southern Weed Science Society 44: 265. Mehlman, D.W. 1992. Effects of fire on plant community composition of North Florida second growth pineland. Bulletin of the Torrey Botanical Club 119: 376-383. McCune, B. and Grace, J. 2002. Analysis of Ecological Communities. MJM Software Design. Gleneden Beach, OR. McGuire, J.P., Mitchell, R.J., Moser, E.B., Pecot, S.D., Gjerstad, D.H., and C.W. Hedman. 2001. Gaps in a gappy forest: plant resources, longleaf pine regeneration, and understory response to tree removal in longleaf pine savann as. Canadian Journal of Forest Research 31: 765-778. Miller, J.H., R.S. Boyd, and M.B. Edwards. 1999. Floristic diversity, stand structure, and composition 11 years after herbicide site preparat ion. Canadian Journal of Forest Research 29: 1073-1083. Mitchell, R.J., Hiers, J.K., OBrien, J.J., Jack, S. B., and R.T. Engstrom. 2006. Silviculture that sustains: the nexus between silviculture, fre quent prescribed fire, and conservation of biodiversity in longleaf pine forests of the sout heastern United States. Canadian Journal of Forest Research 36: 2724-2736. Neary, D.G., J.E. Smith, B.F. Swindel, and K.V. Miller. 1991. Effects of forestry herbicides on plant species diversity. Proceedings of the Southern Weed Science Society 44: 266-272. NOAA Satellite and Information Service: National Climatic Data Center. 2007. Panama City, FL Station 5N. http://www4.ncdc.noaa.gov/cgiwin/ wwcgi.dll?wwDI~StnSrch~StnID~20004593 Noss, R.F. 1989. Longleaf pine and wiregras s: keystone components of an endangered ecosystem. Natural Areas Journal 9: 211-213. 81

PAGE 82

Noss, R.F., LaRoc, III, E.T., and J.M. Scott. 1995. Endangered ecosystems of the United States: a preliminary assessment of loss and degrada tion. USDI National Biological Service, Washington, D.C., Biological Report No. 28. Olson, M.S. and W.J. Platt. 1995. Effects of ha bitat and growing season fires on resprouting of shrubs in longleaf pine savannas. Vegetatio 119: 101-118. Outcalt, K.W., Williams, M.E., and O. Onokpise. 1999. Restoring Aristida stricta to Pinus palustris ecosystems on the Atlantic Coastal Plain, U.S.A. Restoration Ecology 7(3): 262270. Peet, R.K., Wentworth, T.R., and P.S. White 1998. Method for recording vegetation composition and structure. Castanea 63: 262-274. Palik, B.J., Mitchell, R.J., Houseal, G., and N. Pederson. 1997. Effects of canopy structure on resource availability and seed ling responses in a longleaf pine ecosystem. Canadian Journal of Forest Research 27: 1458-1464. Palik, B.J., Mitchell, R.J., J.K. Hiers. 2002. M odeling silviculture after natural disturbance to sustain biodiversity in the longleaf pine ( Pinus palustris ) ecosystem: balancing complexity and implementation. Forest Ecology and Management 155: 347-356. Pecot, S.D., Mitchell, R.J., Palik, B.J., Moser, E.B., and J.K. Hiers. 2007. Competitive responses of seedlings and understory plants in longleaf pine w oodlands: separating canopy influences above and below ground. Canadian Journal of Forest Research 37: 634-648. Peet, R.K. 2006. Ecological clas sification of longleaf pine woodl ands. Pages 51-93 in Jose, S., Jokela, E., and Miller, D.L., editors. The Longleaf Pine Ecosystem: Ecology, Silviculture, and Restoration. Springer Science + Business Media, LLC., New York. Peet, R.K. and D.J. Allard. 1993. Longleaf pine-dominated vegetation of the southern Atlantic and eastern Gulf Coast region, USA. Pro ceedings of the Tall Timbers Fire Ecology Conference 18: 45-81. Platt, W.J. and S.L. Rathbun. 1993. Dynamics of an old-growth longleaf pine population. Proceedings of the Tall Timb ers Fire Ecology Conference 18: 275-297. Platt, W.J., Beckage, B., Doren, R.F. & H.H. Slater. 2002. Interactions of large-scale disturbances: prior fire regimes and hurrica ne mortality of savanna pines. Ecology 83: 15661572. Platt, W.J., Carr, S.M., Reilly, M., and J. Fahr 2006. Pine savanna overstorey influences on ground-cover biodiversity. A pplied Vegetation Science 9: 37-50. 82

PAGE 83

Provencher, L., Herring, B.J., Gordon, D.R., R ogers, H.L., Galley, K.E.M., Tanner, G.W., Hardesty, J.L., and L.A. Brennan. 2001. Effects of hardwood reduction techniques on longleaf pine sandhill vegetation in nort hwest Florida. Restoration Ecology 9: 13-27. Ramsey, C.L., S. Jose, B.J. Brecke, and S. Merritt. 2003. Growth response of longleaf pine ( Pinus palustris Mill.) seedlings to fertilization and herbaceous weed control in an old field in southern USA. Fore st Ecology and Management 172(2-3): 281-289. Ramsey, C.L., and S. Jose. 2004. Growth, surviv al, and physiological eff ects of hexazinone and sulfometuron methyl applied overtop of longleaf pine seed lings. Southern Journal of Applied Forestry 28(1): 48-54. Ranasinghe, S. 2003. Role of herbicides in longl eaf pine flatwoods restoration: pine growth, understory vegetation response, and fate of applie d herbicides. Masters thesis. University of Florida, Gainesville. Ranasinghe, S., Jose, S., and C. Ramsey. 2005. Understory species dynamics following herbicidal release treatments on a longleaf pine site in the Southern U.S.A. In New Research on Forest Ecosystems, A.R. Burk, ed. Nova Science Publishers, Inc. Rebertus, A.J., Williamson, G.B., and W.J. Platt. 1993. Impacts of temporal variation in fire regime on savanna oaks and pines. Pro ceedings of the Tall Timbers Fire Ecology Conference 18: 215-225. Shepard, J.P., Creighton, J., and H. Duzan. 2004. Forestry herbicides in the United States: an overview. Wildlife Society Bulletin 32(4): 1020-1027. Varner III, J.M., and J.S. Kush. 2004. Remnant ol d-growth longleaf pine (pinus palustris Mill.) savannas and forests of the Southeastern USA; status and threats. Natural Areas Journal 24: 141-149. Waldrop, T.A., White, D.L., and S.M. Jones. 1992. Fire regimes for pine-grassland communities in the southeastern United States. Forest Ecology and Management 47: 195210. Walker, J., and R.K. Peet. 1983. Composition and species diversity of pine-wiregrass savannas of the Green Swamp, Nort h Carolina. Vegetatio 55: 163-179. Walker, J. and A.M. Silletti. 2006. Restoring the Ground Layer of Longleaf Pine Ecosystems. Pages 297-333 in Jose, S., Jokela, E., and Miller, D.L., editors. The Longleaf Pine Ecosystem: Ecology, Silviculture, and Restoration. Springer Science + Business Media, LLC., New York. Welch, J.R., Miller, K.V., Palmer, W.E., and T.B. Harrington. 2004. Response of understory vegetation important to the northern bobwh ite following imazapyr and mechanical treatments. Wildlife Society Bulletin 32(4): 1071-1076. 83

PAGE 84

Wilkins, R.N., W.R. Marion, D.G. Neary, and G. W. Tanner. 1993. Vascular plant community dynamics following hexazinone site preparati on in the lower Coastal Plain. Canadian Journal of Forest Research 23: 2216-2229 Williamson, G.B. and E.M. Black. 1981. High te mperature of forest fires under pines as a selective advantage over oaks. Nature 293: 643-644. 84

PAGE 85

85 BIOGRAPHICAL SKETCH Johanna Freeman is a native of western New Yo rk state. She received a B.S. in natural resources from Cornell University in 2001, where her studies focused on animal behavior and environmental policy. She worked as a land-use planner at the New York City Department of Parks and Recreation for three years prior to en tering the masters program in interdisciplinary ecology at the University of Florida.