492 FAHRIGAnnu. Rev. Ecol. Evol. Syst. <str<strong>on</strong>g>2003.</str<strong>on</strong>g>34:487-515. Downloaded from arjournals.annualreviews.orgby University <str<strong>on</strong>g>of</str<strong>on</strong>g> Nevada - Reno <strong>on</strong> 01/25/06. For pers<strong>on</strong>al use <strong>on</strong>ly.reas<strong>on</strong> is that when ecologists think <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g>, the word invokes more than<str<strong>on</strong>g>habitat</str<strong>on</strong>g> removal: “<str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> ... not <strong>on</strong>ly causes loss <str<strong>on</strong>g>of</str<strong>on</strong>g> the amount <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>habitat</str<strong>on</strong>g>,but by creating small, isolated patches it also changes the properties <str<strong>on</strong>g>of</str<strong>on</strong>g> theremaining <str<strong>on</strong>g>habitat</str<strong>on</strong>g>” (van den Berg et al. 2001).Habitat can be removed from a landscape in many different ways, resulting inmany different spatial patterns (Figure 2). Do some patterns represent a higherdegree <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> than others, and does this have implicati<strong>on</strong>s for <strong>biodiversity</strong>?If the answer to either <str<strong>on</strong>g>of</str<strong>on</strong>g> these questi<strong>on</strong>s is “no,” then the c<strong>on</strong>cept <str<strong>on</strong>g>of</str<strong>on</strong>g><str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> is redundant with <str<strong>on</strong>g>habitat</str<strong>on</strong>g> loss. The asserti<strong>on</strong> that <str<strong>on</strong>g>habitat</str<strong>on</strong>g> <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g>means something more than <str<strong>on</strong>g>habitat</str<strong>on</strong>g> loss depends <strong>on</strong> the existence <str<strong>on</strong>g>of</str<strong>on</strong>g> effects<str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> <strong>on</strong> <strong>biodiversity</strong> that can be attributed to changes in the pattern<str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>habitat</str<strong>on</strong>g> that are independent <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>habitat</str<strong>on</strong>g> loss. Therefore, many researchers define<str<strong>on</strong>g>habitat</str<strong>on</strong>g> <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> as an aspect <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>habitat</str<strong>on</strong>g> c<strong>on</strong>figurati<strong>on</strong>.FRAGMENTATION AS A CHANGE IN HABITAT CONFIGURATION In additi<strong>on</strong> to loss<str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>habitat</str<strong>on</strong>g>, the process <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>habitat</str<strong>on</strong>g> <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> results in three other effects: increasein number <str<strong>on</strong>g>of</str<strong>on</strong>g> patches, decrease in patch sizes, and increase in isolati<strong>on</strong><str<strong>on</strong>g>of</str<strong>on</strong>g> patches. Measures <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> that go bey<strong>on</strong>d simply <str<strong>on</strong>g>habitat</str<strong>on</strong>g> amount aregenerally derived from these or other str<strong>on</strong>gly related measures (e.g., amount <str<strong>on</strong>g>of</str<strong>on</strong>g>edge). There are at least 40 such measures <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> (McGarigal et al.2002), many <str<strong>on</strong>g>of</str<strong>on</strong>g> which typically have str<strong>on</strong>g relati<strong>on</strong>ships with the amount <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>habitat</str<strong>on</strong>g>as well as with each other (Bélisle et al. 2001, Boulinier et al. 2001, Droletet al. 1999, Gustafs<strong>on</strong> 1998, Haines-Young & Chopping 1996, Hargis et al. 1998,Robins<strong>on</strong> et al. 1995, Schumaker 1996, Trzcinski et al. 1999, Wickham et al. 1999)(Figure 3).The interrelati<strong>on</strong>ships am<strong>on</strong>g measures <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> are not widely recognizedin the current <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> literature. Most researchers do not separate theeffects <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>habitat</str<strong>on</strong>g> loss from the c<strong>on</strong>figurati<strong>on</strong>al effects <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g>. This leadsto ambiguous c<strong>on</strong>clusi<strong>on</strong>s regarding the effects <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>habitat</str<strong>on</strong>g> c<strong>on</strong>figurati<strong>on</strong> <strong>on</strong> <strong>biodiversity</strong>(e.g., Summerville & Crist 2001, Swens<strong>on</strong> & Franklin 2000). It is alsocomm<strong>on</strong> for <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> studies to report individual effects <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g>measures without reporting the relati<strong>on</strong>ships am<strong>on</strong>g them, which again makes theresults difficult to interpret.THE PATCH-SCALE PROBLEM Similar problems arise when <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> is measuredat the patch scale rather than the landscape scale. Because <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> isa landscape-scale process (Figure 1), <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> measurements are correctlymade at the landscape scale (McGarigal & Cushman 2002). As pointed out byDelin & Andrén (1999), when a study is at the patch scale, the sample size atthe landscape scale is <strong>on</strong>ly <strong>on</strong>e, which means that landscape-scale inference isnot possible (Figure 4; see Brennan et al. 2002, Tischendorf & <str<strong>on</strong>g>Fahrig</str<strong>on</strong>g> 2000).However, in approximately 42% <str<strong>on</strong>g>of</str<strong>on</strong>g> recent <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> studies, individual datapoints represent measurements <strong>on</strong> individual patches, not landscapes (Table 1).Similarly, using a different sample <str<strong>on</strong>g>of</str<strong>on</strong>g> the literature, McGarigal & Cushman (2002)
Annu. Rev. Ecol. Evol. Syst. <str<strong>on</strong>g>2003.</str<strong>on</strong>g>34:487-515. Downloaded from arjournals.annualreviews.orgby University <str<strong>on</strong>g>of</str<strong>on</strong>g> Nevada - Reno <strong>on</strong> 01/25/06. For pers<strong>on</strong>al use <strong>on</strong>ly.EFFECTS OF HABITAT FRAGMENTATION 493Figure 2 Illustrati<strong>on</strong> <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>habitat</str<strong>on</strong>g> loss resulting in some, but not all, <str<strong>on</strong>g>of</str<strong>on</strong>g> the other threeexpected effects <str<strong>on</strong>g>of</str<strong>on</strong>g> <str<strong>on</strong>g>habitat</str<strong>on</strong>g> <str<strong>on</strong>g>fragmentati<strong>on</strong></str<strong>on</strong>g> <strong>on</strong> landscape pattern. Expected effects are(a) an increase in the number <str<strong>on</strong>g>of</str<strong>on</strong>g> patches, (b) a decrease in mean patch size, and(c) an increase in mean patch isolati<strong>on</strong> (nearest neighbor distance). Actual changes areindicated by arrows.