Oikos OIK-00962 Byrom, A. E., Craft, M. E., Durant, S. M., Nkwabi, A. J. K., Metzger, K., Hampson, K., Mduma, S. A. R., Forrester, G. J., Ruscoe, W. A., Reed, D. N., Bukombe, J., Mchetto, J. and Sinclair, A. R. E. 2014. Episodic outbreaks of small influence predator community dynamics in an east African savanna ecosystem. – Oikos doi: 10.1111/oik.00962

Appendix1 Details of transect surveys for -eating birds Data on the abundance of rodent-eating birds were obtained using transects by vehicle along tracks and roads. Transects were driven slowly (30-50 kph), using two observers and a recorder. Transects were run on a regular basis twice a year, at the end of the short rains in December–January and at the end of the long rains in May–June (Jankowski et al. in press). One transect crossed the plains from Seronera to Naabi (50 km), and another sampled the savanna of the western corridor from Seronera to Kirawira at the western boundary of SNP (100 km). A third transect sampled the Acacia woodland from Seronera North to Ikoma gate, and a fourth sampled the eastern savanna from Seronera northeast towards Loliondo (Fig. A1). A fifth transect (75 km) ran north outside SNP through villages and agricultural areas northwest of SNP (Fig. A1).

Details of transect surveys for small mammalian carnivores Fixed-width transect methods and distance-based methods were used to estimate carnivore density annually (1993–2006) for daytime surveys (Durant et al. 2011). Night-time surveys (2003–2010) were conducted monthly using spotlights throughout the main road network across the greater Serengeti ecosystem (Craft et al. in press). Transects were driven along roads within SNP and outside SNP in agricultural areas to the north-west with two observers inside a vehicle each holding a spotlight directed to either side (Craft et al. in press and Fig 1A). Transects covered 20 km stretches in areas outside of the national park and 40 km stretches of road within the national park.

References Craft, M. E. et al. Carnivore communities in the greater Serengeti ecosystem. – In: Sinclair, A. R. E. et al. (eds), Serengeti IV: sustaining biodiversity in a coupled human–natural system. Univ. Chicago Press, in press. Durant, S. M. et al. 2011. Long-term trends in carnivore abundance using distance sampling in Serengeti National Park, . – J. Appl. Ecol. 48: 1490–1500. Jankowski, J. E. et al. Bird diversity of the greater Serengeti ecosystem: spatial patterns of taxonomic and functional richness and turnover. – In: Sinclair, A. R. E. (eds), Serengeti IV: sustaining biodiversity in a coupled human–natural system. Univ. Chicago Press, in press.

Table A1. Latin and common names of of 37 small in the greater Serengeti ecosystem, 1962–2010. Includes published records from Misonne and Verschuren (1966), Senzota (1982, 1984), Packer (1983), and Magige and Senzota (2006), and unpublished data from A. R. E. Sinclair and D. Reed.

Latin name Common name Suncus shrew Tachyoryctes Acomys spiny mouse African mole-rat daemon kaiseri bush rat Aethomys nyikae bush rat paedulcus niloticus grass rat thicket rat Crocidura shrew Zelotomys broad-headed mouse Crocidura 2 shrew hildegardeae Dasymys incomptus shaggy swamp rat Dendromus tree mouse melanotis Elephantulus long-eared elephant rufescens shrew large naked-soled Gerbilliscus robusta gerbil Gerbillus pygmy gerbil Gerbillus pusillus pygmy gerbil thicket rat Grammomys thicket rat dolichurus Graphiurus murinus African dormouse Lemniscomys striped grass mouse barbarus Lemniscomys striped grass mouse striatus Lophuromys brush-furred mouse sikapusi Mastomys multi-mammate rat natalensis Latin name Common name Mastomys pernanus multi-mammate rat Mus (Nannomys) mouse musculoides Mus (Nannomys) mouse tenellus African groove- Myomys yemeni toothed rat rufous-nosed rat hypoxanthus Otomys angoniensis African swamp rat Pedetes capensis spring-hare groove-toothed creek fallax rat Praomys soft-furred rat Praomys jacksoni soft-furred rat Rattus rattus black rat Saccostomus African pouched rat mearnsi Steatomys parvus fat mouse Table A2. Linear regression models and model selection to determine the the influence of the amount of rainfall (in mm) in each season on rodent abundance in the greater Serengeti Ecosystem, 1968–2010. ‘West’ refers to rainfall data from a weather station in the Western Corridor and ‘Center’ refers to data from a station in the center of SNP (see main text for further explanation). ‘Short’ = short rains; ‘Long’ = long rains.

2 Model t31 r p AIC Δ-AIC Weight Support n Wet Season West 3.11 0.24 0.003 21.510 0.000 1.00 33 Annual Rainfall Center ( 12-month lag) -3.11 0.11 0.004 22.347 0.837 0.66 1.52 33 Short West 3.10 0.24 0.004 22.453 0.943 0.62 1.60 33 Wet Season Center 2.91 0.21 0.007 23.371 1.861 0.39 2.54 33 Short Center 2.88 0.21 0.007 23.541 2.031 0.36 2.76 33 Wet Season Center ( 12-month lag) -2.38 0.15 0.024 25.819 4.309 0.12 8.62 33 Long West 2.36 0.15 0.025 25.919 4.409 0.11 9.07 33 Dry Season Center -2.35 0.10 0.025 25.931 4.421 0.11 9.12 33 Long Center ( 12-month lag) -2.08 0.12 0.046 27.048 5.538 0.06 15.94 33 Long Center 1.97 0.11 0.058 27.462 5.952 0.05 19.61 33 Annual Rainfall Center 1.94 0.11 0.061 27.555 6.045 0.05 20.54 33 Annual Rainfall West 1.84 0.10 0.075 27.929 6.419 0.04 24.77 33 Dry Season Center ( 12-month lag) -1.83 0.10 0.077 27.961 6.451 0.04 25.17 33 Dry Season West -1.83 0.10 0.077 27.970 6.460 0.04 25.28 33 Dry Season West ( 12-month lag) -1.78 0.093 0.085 28.138 6.628 0.04 27.49 33 Short Center ( 12-month lag) -1.74 0.089 0.092 28.287 6.777 0.03 29.62 33 Annual Rainfall West ( 12-month lag) -1.38 0.098 0.177 29.383 7.873 0.02 51.24 33 Wet Season West ( 12-month lag) -0.83 0.022 0.414 30.631 9.121 0.01 95.63 33 Short West ( 12-month lag) -0.83 0.021 0.416 30.637 9.127 0.01 95.92 33 Long West ( 12-month lag) -0.58 0.011 0.569 31.004 9.494 0.01 115.24 33

Table A3. Logistic regression models and model selection to determine the influence of the amount of rainfall (in mm) in each season on the probability of an outbreak of in the greater Serengeti Ecosystem, 1968–2010. ‘West’ refers to rainfall data from a weather station in the Western Corridor; ‘Center’ refers to data from a station in the center of SNP (see main text for further explanation). ‘Short’ = short rains; ‘Long’ = long rains.

Model z % Deviance explained p AIC Δ-AIC Weight Support n Wet Season West 2.32 22 0.020 35.691 0 1.00 33 Dry Season Center -2.38 20 0.018 36.504 0.813 0.67 1.50 33 Short West 2.13 18 0.033 37.296 1.605 0.45 2.23 33 Wet Season Center 2.23 17 0.026 37.648 1.957 0.38 2.66 33 Dry Season West -2.12 14 0.034 38.768 3.077 0.21 4.66 33 Long West 1.97 13 0.049 39.179 3.488 0.17 5.72 33 Short Center 2.02 13 0.043 39.393 3.702 0.16 6.37 33 Long Center 1.88 11 0.061 40.037 4.346 0.11 8.78 33 Annual Rainfall Center (12-month lag) -1.82 9 0.069 40.672 4.981 0.08 12.07 33 Annual Rainfall Center 1.57 7 0.116 41.637 5.946 0.05 19.55 33 Annual Rainfall West 1.58 7 0.115 41.706 6.015 0.05 20.24 33 Wet Season Center (12-month lag) -1.53 7 0.124 41.775 6.084 0.05 20.95 33 Short Center (12-month lag) -1.33 5 0.180 42.418 6.727 0.03 28.89 33 Long Center (12-month lag) -1.21 4 0.224 42.915 7.224 0.03 37.04 33 Dry Season Center (12-month lag) -1.06 3 0.289 43.281 7.59 0.02 44.48 33 Short West (12-month lag) -0.98 3 0.330 43.353 7.662 0.02 46.11 33 Dry Season West (12-month lag) -0.97 2 0.331 43.507 7.816 0.02 49.80 33 Annual Rainfall West (12-month lag) -0.70 1 0.485 43.985 8.294 0.02 63.24 33 Wet Season West (12-month lag) -0.41 0 0.680 44.311 8.62 0.01 74.44 33 Long West (12-month lag) -0.10 1 0.921 44.475 8.784 0.01 80.80 33

Figure A1. Map of the greater Serengeti ecosystem showing the major habitat types surveyed for small mammals (end of each pointer line indicates approximate locations of small mammal trapping grids), locations of survey transects for rodent-eating birds and small mammalian carnivores, and locations of agricultural villages used for trapping small mammals. Gray outline indicates boundary of SNP.

1