Scale and bark injuries (a condition called epidermal browning) have been shown to occur on stem surfaces of 21 species of tall, long-lived cactus species throughout the Americas. The amounts of surface injuries are highly correlated with amounts of sunlight exposures. Epidermal browning causes cactus morbidity and mortality of saguaro cacti. For this study, rates of morbidity and mortality of saguaro cacti (Carnegiea gigantea (Engelm.) Britt and Rose) were determined for 1061 plants from 50 desert field plots. From 2002 to 2010, 16.5% of all saguaros died (increased mortality). During the same period, the percentage of cacti with few epidermal injuries decreased from 25 to 9%, with a 16% increase in initial morbidity. Over this same period, about 10% of the population had a rapid increase of epidermal browning (rapid morbidity). Thus, marked increases in morbidity and mortality occurred over the eight-year period. WEKA, a standard machine-learning program, was used to determine the characteristics of saguaros that predispose them to (1) die, (2) initiate morbidity (start of epidermal browning symptoms) and (3) initiate rapid morbidity (a rapid increase in epidermal browning). Decision trees were created using the C4.5 algorithm, a top-down method that repeatedly chooses the question that produces a maximal information gain. Amounts of epidermal browning on north-facing troughs alone predicted cactus mortality with 89.1% accuracy. Initial morbidity was predicted with 80.2% accuracy with data of south-facing troughs and crests only. Rapid morbidity (less than 20% epidermal browning in 2002 and more than 50% by 2010) was predicted with a very large tree with 79.6% accuracy. For example, rapid morbidity is predicted by more than 3.1% epidermal browning on left troughs on south-facing surfaces in 2002 with 80% accuracy. Moreover, rapid morbidity was also predicted if east-facing crests had more than 3.9% epidermal browning in 2002. Overall, decision tree analysis was very effective at predicting changes in mortality and morbidity for the population of 1061 saguaros from 2002 to 2010. Data from trough surfaces were more effective than data from crests for predicting mortality and morbidity.
You have requested a machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Neither BioOne nor the owners and publishers of the content make, and they explicitly disclaim, any express or implied representations or warranties of any kind, including, without limitation, representations and warranties as to the functionality of the translation feature or the accuracy or completeness of the translations.
Translations are not retained in our system. Your use of this feature and the translations is subject to all use restrictions contained in the Terms and Conditions of Use of the BioOne website.